ISPs Lie About Broadband "Up To" Speeds

Haffner writes "Ars Technica has an article detailing the difference between ISP advertised 'up to x Mbps' speeds and the actual speeds, in addition to some possible solutions. They find that on average, the advertised speeds were 'up to 6.7 Mbps' while the real median was 3 Mbps and the mean was 4 Mbps. This implies that ISPs were falsely advertising by at least 50%."

Comments

ISP's want your money...

[Anonymous Coward]

News at 11

Re:ISP's want your money...

[JWSmythe]

    It wouldn't be news, it would be a rerun. They should have shortened the title to "ISPs Lie", which directly relates to "Companies Lie".

Re:ISP's want your money...

[causality]

It wouldn't be news, it would be a rerun. They should have shortened the title to "ISPs Lie", which directly relates to "Companies Lie".

... which directly relates to "people lie".

Re:ISP's want your money...

[JWSmythe]

    Well, people lie. People representing companies lie even more.

    On the other hand, most folks lie all the time, don't they. They caught a fish THIS big, or their penis is SO long. Oh well, there goes the last bit of faith that I had in humanity. So long, and thanks for all the fish.

Re:ISP's want your money...

[Surt]

This was the last piece of faith you had in humanity? I'd have expected this to be one of the first pieces to go.

Re:ISP's want your money...

[causality]

Well, people lie. People representing companies lie even more.

On the other hand, most folks lie all the time, don't they. They caught a fish THIS big, or their penis is SO long. Oh well, there goes the last bit of faith that I had in humanity. So long, and thanks for all the fish.

I'd say the faith belongs in sound principles that humans can acknowledge and appreciate or fail to acknowledge and appreciate. It's a good start, anyway.

Re:ISP's want your money...

[Morth]

Even the lawmakers have managed to catch up to this one in Sweden. ISPs now have to say 5-10 Mbps or similar.

Re:ISP's want your money...

[Hognoxious]

News at up to 12

FTFY.

Re:ISP's want your money...

[BiggestPOS]

I pay $120/US a month for 107 megabits down, and 5 megabits of upload in East Texas.
I never have any problems downloading from the "right places" at 12.5 megabytes/sec. Amazon Web Services being one of the few places that can actually saturate my pipe with an HTTP download. Use your imagination for the other places I'm able to max it out :)

Re:ISP's want your money...

[ilec_geek]

This story from Ars is misleading. Chances are you really are getting the bandwidth you are paying for. I am amazed at how many "technical" people have no idea how computer networks actually work. The "Internet" is not much more than a vast mesh of many many interconnected networks. My ISP, for example, has a speed test link right on their website that links to a server that physically resides on their network. When using this tool, I always get 100% and sometimes a little more than the package I am paying for. Do you people realize that "speedtest.net" or any other service like BitTorrent is NOT capable of showing accurate results of your connection speed? The Internet is built on a "best effort" design and once your traffic egresses your ISPs network, they have absolutely no control over how that traffic gets routed to its ultimate destination. It's like trying to make a phone call from Boise Idaho to New York City. All the lines are busy in New York and your call can't get through. How is that your local telephone company's fault? And what exactly do you expect them to do to fix it? Your ISP can only guarantee your connection speed from your house to their Central Office. Once it leaves their domain, it is totally out of their hands. Any perception of "slow speeds" are 99.999% of the time related to network issues "somewhere else." Your ISP has a reputation to uphold and customers to keep happy. They most likely have built their infrastructure as robust as they possibly can or they wouldn't even be able to open their doors and offer services in the first place. Do you have any idea what it actually takes to build a wireless broadband network, fiber-to-the-home infrastructure, or even re-condition a copper plant to get DSL performing at its optimum level? Trust me, their profit magin is actually pretty slim. Most likely your monthly bill is just paying for infrastructure for the first year before they ever even see a profit. So don't go blaming ISPs for being money hungry liars when all you need is a little education on how networks actually work and the economics involved in building infrastructure and selling services.

On the subject of lying

[Anonymous Coward]

I'd say they were lying by up to 100%

Sneaky, yes. Lies, not quite.

[Rary]

They find that on average, the advertised speeds were 'up to 6.7 Mbps' while the real median was 3 Mbps and the mean was 4 Mbps. This implies that ISPs were falsely advertising by at least 50%.

"Up to" doesn't mean "median" or "mean". "Up to" means "up to", as in "maximum".

That being said, it is rather sneaky to advertise a product by focusing on a theoretical maximum that you may (or may not) experience on the rarest of occasions. It's kind of like selling a limited service as "unlimited". But no one would ever do that, right?

Re:Sneaky, yes. Lies, not quite.

[MBCook]

Do you think Doritos would be allowed to sell bags as "up to a pound" when they averaged 9oz and some had quite a bit less? The big problem is it's one way. When you are promised Xmbps, you get some number, Y, where Y<= X. I would be amazed if more than 1% of the broadband population got higher than their rated speed. If it was a real normal distribution, or when you called to sign up they told you "you can expect to get X most of the time".

But my parents have 12 or 15mbps cable internet. During normal hours (even early afternoon) it is almost never faster than about 8mbps, and that's with multiple downloads coming from what I assume to be a CDN, because most sites aren't anywhere near that. Over the last 5-6 years, the top speed you could reach on their cable line has dropped as more people have signed on, but the advertised speed (and the price) have both increased. They have a medium package since there is no point trying to get more on an oversubscribed line.

I, on the other hand, pay for 6mbps DSL, and get almost exactly 6. I like getting what I pay for, and if I could only get 3, I'd pay for that service level.

If your "up to" only applies to 5% of your customers, you're scamming them. If it was 30%, I think we'd all be a lot more forgiving.

Re:Sneaky, yes. Lies, not quite.

[Jeff DeMaagd]

Do you think Doritos would be allowed to sell bags as "up to a pound" when they averaged 9oz and some had quite a bit less? The big problem is it's one way. When you are promised Xmbps, you get some number, Y, where Y

If they say "up to" they can provide any speed between and including zero and the "up to" figure and still be correct.

I don't think anyone expects any industry to deliver more than promised, seems to be a lost cause.

This problem will not go away so easily when the average person isn't paying attention and even supposedly technically knowledgeable people do not understand the language enough to see the weaseling before they sign up. Then there's the problem where a /. submitter and editor do not understand basic statistical terms.

Re:Sneaky, yes. Lies, not quite.

[gzipped_tar]

It is not clear from TFA whether the histogram displayed there was drawn from the sample of experimentally measured _maximum_ speeds or just the "daily usage" speeds.

If it was the former, then it gives us a snapshot of the underlying distribution of the maximum speed, and we can estimate the probability of "ISP lying about the speed", along with the variance of this estimator, directly from it.

If it was the latter, the distribution of the maximum can still be estimated. However, this is usually difficult to be done in a model-independent way.

Re:Sneaky, yes. Lies, not quite.

[sjames]

It would be a lot more honest to advertise the "at least" speed, but they don't want to do that because according to the fine print they are actually offering at least zero Mbps.

I wonder how they would feel if their customers offered to pay them up to $40/month.

Re:Sneaky, yes. Lies, not quite.

[Kitkoan]

Diet/exercise companies do this all the time. They show you ads of people with amazing figures stating they used their products and look at them. For the diet ads its the rare person that comes out looking like that but by US law they can use those people as their examples because they did use the product, they just had a rare outcome (why the fine print tends to state that these aren't normal outcomes, but most people don't read the fine print). Similar for exercise equipment, but while also "forgetting" to mention that the person had to diet a lot and do much, much more then the "minimum amount of time needed" and had to use more then the equipment shown. Its all in what it takes to stay just legal enough to push your product for the most amount to money.

Re:Sneaky, yes. Lies, not quite.

[westlake]

It's kind of like selling a limited service as "unlimited".

"Unlimited Internet" is phrase which took hold in the 14K dial-up days when AOL began offering unmetered service for a flat monthly rate of $19.95.

The perfect compliment to your unmetered local calling plan.
 

Re:Sneaky, yes. Lies, not quite.

[DaveAtFraud]

As a guess, the "up to" speed is what you would see if you download something from one of their servers (e.g., retrieved your e-mail). They can't control how fast anyone else's server serves data nor how fast the other server's provider is. You might see something served that fast from across the Internet but don't count on it; too many variables.

Besides, the "up to" speed is a lot faster than any speed estimate that involves realistic usage. Go figure that they quote it.

Cheers,
Dave

Re:Sneaky, yes. Lies, not quite.

[masshuu]

My DSL is rated at 3mbps, ive hit 3.2-3.6mbps in the past. Don't normally hit that though. Mean speed is ~350KBps, with steam downloading at ~400MBps now(its almost 10PM)

When i had FIOS, i could usually max out my speed with a good large download(ubuntu torrent for example), but when i was on cable, i don't think i ever maxed it out, or came close.

-Irving, TX

Re:Sneaky, yes. Lies, not quite.

[Anaerin]

No, typically the "Up to" is the maximum possible raw bandwidth, before any kind of packetization is performed. So if the line is capable of carrying 10,000,000 raw bits of data per second, they'll advertise it as "Up to 10Mb/s". Despite the fact that, even in a perfect situation the most you would get is 7.15Mb/s (That's 10,000,000/1,048,576 (or 1024*1024)*0.75 (To allow for packet framing overhead)), or a transfer speed of 915.5KByte/sec from the ISP's servers. That's without any packet losses, signal attenuation issues, or noisy transfer media. And probably even less than that, even in a perfect situation, as their outgoing bandwidth is likely to be highly contested.

Re:Sneaky, yes. Lies, not quite.

[gandhi_2]

Would be nice if they advertised something more like a CIR.

Re:Sneaky, yes. Lies, not quite.

[mysidia]

I believe they are looking at 'median' and 'mean' maximum speeds.

If the maximum speed your line achieve is less than the advertised speed, and they didn't inform you that the advertised speed was unavailable and inform you would get XXX instead, then the ISP lied to you.

Re:Sneaky, yes. Lies, not quite.

[DeadPixels]

I actually like the parent's Doritos analogy - it's true when you think of it that way; there would be all sorts of uproar if physical goods were advertised and sold the way broadband is.
"Up to" a dozen bagels in your order, or "up to" two patties on your burger would never fly. And who would work for pay on an "up to" scale? I'm sure companies would be happy to pay someone "up to" four hundred dollars an hour.

Part of the problem, in my opinion, is the fact that a sizable portion of the population is not terribly computer-literate or technically savvy. They want "an Internet" or "a Google" or (my new favorite) "the Facebook" and don't really care much about how they get it.
The average end user, in my experience, has difficulty distinguishing between a slow computer and a slow connection. To many, they might as well be one and the same. I get asked for help all the time with people saying "my computer is slow" and it turns out they actually have connectivity problems. ISPs not only take advantage of that mentality, they count on it. I'm sure many of you have seen the commercials for those sites like "FinallyFast" or "MaxMySpeed" or whatever they're called, where they advertise a "free scan" to tell if you're "infected" or "experiencing registry errors", and by purchasing their product, you can avoid having to buy a new computer. That is basically the same demographic ISPs are targeting; the population that knows they want a computer and an internet connection but doesn't know much beyond that. I would honestly describe it as predatory.

I know I'll probably get modded down for not taking a more pro-capitalistic stance, but in my opinion this is a case where consumers are being taken advantage of - and there simply are no better options. It's very easy to say "vote with your dollar and don't buy their services", but an internet connection is critical for many people nowadays. I know several people who run businesses out of their homes using websites, VOIP lines, etc. For them, canceling their internet connection is just not an option. If there were an ISP that actually provided good service and had consumer-friendly policies, I would be more than happy to switch to their service and recommend all of my friends. The problem is that my options right now are "bad", "worse", and "even worse yet". Comcast blocks all torrent upload data in my area (disclaimer: I don't pirate content, but I do use torrents for FOSS/Linux downloads and similar uses); Verizon has declared that they plan to test a 150GB (if I remember correctly) monthly cap on FIOS in this area; and there's basically no one else around because they've been driven out of business or out of the area. Again, with the nature of the internet and the role it plays in communication and commerce, I would almost consider supporting it being regulated like a public utility, or at least with more oversight. It's all well and good to say "don't give them your money", but when I need the internet to obtain that money, I don't have many options.

Re:Sneaky, yes. Lies, not quite.

[Anonymous Coward]

Do you think Doritos would be allowed to sell bags as "up to a pound" when they averaged 9oz and some had quite a bit less?

Stores often advertise sales where items are "up to 50% off". Sometimes only 1 or 2% of the items in-store are actually 50% off. How is this any different?

Re:Sneaky, yes. Lies, not quite.

[poetmatt]

only if they weren't the ones who built it up in the first place. If it's hops along an ISP's own backbone you better believe they can control the speeds. It's one more reason for packet inspection (whether deep or otherwise), one of the few legitimate reasons even.

Meanwhile, I guess it depends on population in a given area. I pretty much see the speeds as promised by comcrap in my area, as much as I hate them. I am also but a few miles outside of a major metro city.

Re:Sneaky, yes. Lies, not quite.

[russotto]

It is not clear from TFA whether the histogram displayed there was drawn from the sample of experimentally measured _maximum_ speeds or just the "daily usage" speeds.

The "actual speeds" are averages. The "advertised speeds" were based on maximums. TFA isn't clear, but the study it references goes into detail. There does not appear to be a chart detailing the actual maximums versus the nominal maximums derived from rounding the maximums up to the nearest tier.

Re:Sneaky, yes. Lies, not quite.

[causality]

the average person isn't paying attention and even supposedly technically knowledgeable people do not understand the language enough to see the weaseling before they sign up

None of this is difficult to understand. None! If it seems as though no one understands, it's just because it's accepted as normal.

It's something like apathy. That isn't a problem you can solve by throwing more IQ points at it.

Re:Sneaky, yes. Lies, not quite.

[quadelirus]

'Do you think Doritos would be allowed to sell bags as "up to a pound" when they averaged 9oz and some had quite a bit less?'

I don't see why not. Isn't that exactly what w00t does with its bag of random stuff. At least, I don't think it is false advertising. They said "up to" not "averaging" so anyone who speaks english and has an iota of mathematical sense should know that they aren't going to see those speeds.

That said, I hate these companies. I just think we should complain about all the legitimate things they give us to complain about rather than engineer a problem using a bad understanding of pseudo-mathematical terms.

Re:Sneaky, yes. Lies, not quite.

[drsmithy]

It's kind of like selling a limited service as "unlimited". But no one would ever do that, right?

"Unlimited internet" has, since its inception, pretty much universally referred to how how much you could download, not how fast you could do it.

Re:Sneaky, yes. Lies, not quite.

[Totenglocke]

Well, we do sell cars that are capable of speeds "up to" 155 mph, 200mph, 250 mph, etc. Just because the odds of you ever hitting that high are slim to none doesn't make their statement false - it just means that SOME people will hit that speed and others won't, but it is possible to hit that speed.

Re:Sneaky, yes. Lies, not quite.

[daveime]

If by very little control, you mean oversubscribing the line by a factor of N, then okay, whatever.

The days of people only using the DSL in bursts to checkmail or grab a webpage are gone. In my house, we've got 3 towers and 2 laptops running off a wifi'd 3MB DSL.

At any moment, one or more of us will be streaming video or running torrents etc, so we're usually running at capacity 24/7 - which I have no qualms about whatsoever, as we were sold a 3MB package, not one of those silly "up to 3MB packages" - and for the most part we DO get what we paid for.

I guess my point is, they cannot oversubscribe the lines anymore, because it will mean an instant slowdown for everyone else.

Re:Sneaky, yes. Lies, not quite.

[nschubach]

I have an advertised "up to" 7Mbps (Insight RR) and I just tested my connection to a server here in town:
6.7Mbps

If you are losing 3Mbps (from 10Mbps) you have a serious problem.

Re:Sneaky, yes. Lies, not quite.

[Newander]

You're not promised Xmbps. You're promised up to Xmbps. I would expect exactly 0% of customers to get above their rated speed.

As far as your DSL goes, I wouldn't be at all surprised that you get better service from a phone company than a cable company.

Re:Sneaky, yes. Lies, not quite.

[DaveAtFraud]

But who says that the ISP can't include packet framing overhead in their speed claim?

I remember when Detroit used to spec engine output by the horsepower and torque at the back of the engine; not after any losses due to the things like power steering, air conditioning, etc. At some point the FTC made the manufacturers use net horsepower. Since everybody made the same shift, the relationship between different products was unchanged; just the absolute number changed. The same applies here. If all ISPs factor in the same overhead, nothing changes except the absolute numbers.

My recommendation is that people look at a site like DSL Reports (http://www.dslreports.com/) to get an idea of what they'll actually see.

Cheers,
Dave

Re:Sneaky, yes. Lies, not quite.

[Anonymous Coward]

Yes, the Doritos analogy is good, but it's flawed in the way it's analysed. You get 9oz of dorito chips, not 100 chips. You get an aggregate that adds up to what you ordered, which may vary depending on the size of chips and other factors. Show me someone who orders chips by number instead of weight, and I'll show you a crazy person. 'Fast internet' is, in reality, an qualitative thing, not a quantitative thing. There really isn't much difference between 6Mb/s and 8Mb/s, but there is between 256kb/s and 8Mb/s. Not to mention that the ISPs where I am also provide graphs to give you an idea of expected speed given your distance from the exchange, hardly obfuscatory.

If you really want to be told specifically what speed you'll get, feel free to offer your ISP some $$$ to send out a technician and find out the combination of umpteen factors that will give you your result. It may take quite some $$$ but you'll be able to convince them - and you will get your precious accurate values. The way things currently are is an approximation that assists the whole process, not greedy anti-consumer capitalism at work.

Re:Sneaky, yes. Lies, not quite.

[armanox]

Depends on what you have and how you drive. I can get the 21MPG (Estimated 20 Highway) out of my Chevy Blazer when I drive to work at times that I'm averaging 2000RPM on the engine speed (~65MPH), which was usually the case before I started working 9-6 (or 8:30-5:30 on Fridays).

Re:Sneaky, yes. Lies, not quite.

[PitaBred]

The difference is that the "sale" favors the consumers in theory (well, it's sometimes cheaper than normal price...), and is not controlled by a one-sided contract with few to no other options.

Re:Sneaky, yes. Lies, not quite.

[Nemyst]

Seems to vary quite a bit by ISP. I've surpassed my ISP's rated speed once (by a few Mbps).

Re:Sneaky, yes. Lies, not quite.

[Jafafa Hots]

They're lucky. On my "up to" something like 3-6 mbps (I've so long since given up paying attention that I don't remember) I get a maximum of about 760kbps.

In my experience, in most towns you have at best a choice between two broadband providers and they both usually suck in different ways... and both are vastly overpriced, though one is usually about 20% less expensive than the other, so that's who you go with.

Re:Sneaky, yes. Lies, not quite.

[halltk1983]

I do. Try letting off the loud pedal every once in a while, sparky. For the record, I'm a 26 year old male, who likes to drive fast and corner hard, and I still manage to get the suggested mileage out of my car.

Re:Sneaky, yes. Lies, not quite.

[Surt]

Doritos wouldn't be allowed to, but only because food is an industry with special regulations designed to keep people from dying from consumption of the product.

Re:Sneaky, yes. Lies, not quite.

[Anonymous Coward]

What if I were to say "As a bartender you can make up to $400 of tips in a night."

Are there nights that you can make $400 in tips? Certainly, on some extremely busy nights that you get lucky.
Would you expect the mean/median to be $400 in tips from this sentence? No.

Let's be honest, they are certainly not lying to you, and most people would agree "up to" (=) is not the same as "equal to" (==).

Now I think as a consumer we should be given the mean/median speed of a service we are considering (kind of like how food has a nutrition label) in the form of a federal mandate requiring any advertising or sales materials including the min, mean, median, max connection speeds averaged from a statistically valid sample base of their customers.

Re:Sneaky, yes. Lies, not quite.

[djlowe]

"Up to" doesn't mean "median" or "mean". "Up to" means "up to", as in "maximum".

Yes, thanks for pointing out the obvious, for those with modpoints who don't understand such terms. They've already modded you up for stating the obvious: Good for you!.

That being said, it is rather sneaky to advertise a product by focusing on a theoretical maximum that you may (or may not) experience on the rarest of occasions.

Yeah, 'cause "up to", as plainly stated, is apparently so difficult to comprehend, and so misleading, for so many of you, when stated so in plain English.

It's almost like saying "up to", you know?

It's kind of like selling a limited service as "unlimited". But no one would ever do that, right?

Yeah, kind of like nobody would put limited, and "unlimited", in the same sentence, without defining either, to invoke and cultivate the prejudices of their audience while doing so?

Nice Karma whoring :)

Regards,

dj

Re:Sneaky, yes. Lies, not quite.

[mjwx]

Seems to vary quite a bit by ISP. I've surpassed my ISP's rated speed once (by a few Mbps).

That is often because they advertise an accepted or throttled maximum speed. In Australia ISP's are allowed to advertise the maximum possible speed on a DSL line, so long as it's clear that it's an "up to" thing (this is common knowledge by now in Australia, so it's your problem if you don't understand it by now). ISP's advertise up to 24 Mbit, which is the theoretical maximum for ADSL 2 RE. Some people get 23 Mbit, others get 2 Mbit, it depends on your distance from the exchange.

Re:Sneaky, yes. Lies, not quite.

[Splab]

Errr, most companies sell you close to advertised, a lot of companies down right cheat.

Examples:
A crate of tomatoes advertised to 600g - do you seriously expect them to hit 600g with 5 tomatoes each time?
Bacon, ham - and often meat in general - is pumpes so full of water you will often end up losing some 7-10%
Newspapers claim to do investigation while they often verbatim copy.
Cars claimed to do umpteen miles per gallon (the car I got does 22 km/l according to specs, usually won't go above 18)

The list goes on...

Re:Sneaky, yes. Lies, not quite.

[arth1]

Do you think Doritos would be allowed to sell bags as "up to a pound" when they averaged 9oz and some had quite a bit less?

Why wouldn't they? It would be a stupid thing to do, but "up to" is synonymous with "no more than".

A Berkshire Hathaway company here in the US has advertised with "can save you up to 15% or more on car insurance", which is a completely meaningless statement -- it translates to "a possibility of saving either 15% or less, or more than 15%". Which of course is true.

So, yes, "up to" speeds are probably just as legal. But only idiots would buy based on that. Granted, there are a lot of idiots out there.

I buy my service with a CIR, and have a guaranteed minimum speed as well as a maximum speed.

Re:Sneaky, yes. Lies, not quite.

[Aliotroph]

This is true of Doritos, but the last can of Pringles I opened said on the side that there are at least 100 chips in it.

Re:Sneaky, yes. Lies, not quite.

[IrquiM]

"Up to" doesn't mean "median" or "mean". "Up to" means "up to", as in "maximum".

That being said, it is rather sneaky to advertise a product by focusing on a theoretical maximum that you may (or may not) experience on the rarest of occasions. It's kind of like selling a limited service as "unlimited". But no one would ever do that, right?

I'm glad my ISP says "at least" instead of "up to". So, when I pay for "at least" 10/10, I'm close to 13/14. Life is good on fibre!

Re:Sneaky, yes. Lies, not quite.

[mjwx]

I actually like the parent's Doritos analogy - it's true when you think of it that way

If the weight of a packet of crisps diminished the further away you got from the shop, it would not only be perfectly legal, but prudent to advertise as "up to 200 grams". So the Analogy is terrible

With DSL your speed is entirely dependent on environmental conditions. With Cable it's entirely dependent on load. With Fibre, then you have a point as it should sync at whatever speed your paying for but with 2 Mbit Fibre coming in at A$500 a month, forget that. This is how Australian ISP's market their broadband offerings (ADSL2+). Some are quite unscrupulous but still legal whilst some actually publish information on expected speeds and at what rate they drop. So there is nothing wrong or misleading about the "Up To" advertising because they cannot physically guarantee a certain speed at any one time (thus expecting them to is unreasonable).

Re:Sneaky, yes. Lies, not quite.

[RobinEggs]

Verizon has declared that they plan to test a 150GB (if I remember correctly) monthly cap on FIOS in this area; and there's basically no one else around because they've been driven out of business or out of the area.

Trust me, 9 out of 10 people who can't get FIOS at all (which is like 80% of the country) would *joyously* accept a 150 GB cap if they could actually have fiber. Most people, even most hardcore nerds, don't download anywhere near that much.

Re:Sneaky, yes. Lies, not quite.

[Neil Boekend]

Well, my ADSL provider may sell me an 20 Mb/s connection. That does not mean I will get it. The max at my distance from the little house with the phone stuff in it (and the ADSL netcards) means I will get up to 6 (with default netcards). It wouldn't be a lie, for the speed of that connection package is up to 20.

BTW: they would probably test it and advise against the expensive high speed connection. That's common courtesy, and if they didn't it would create a bad name.

By the way MEAN != max, so it's a bit offtopic.

Re:Sneaky, yes. Lies, not quite.

[Patch86]

Anecdote time.

On my fiber connection, I'm advertised 10MBps. I usually get 9.something, so am generally pretty happy.

Some family of mine (over copper) were paying for 8MBps, but always complaining about how "their computer was so slow". I tested the speed- 0.04MBps. I've tested it since, and never ever ever seen it reach 1MBps. We called their provider, they poppedtheir post code into their database, and sure enough "oh, you're too far away from the exchange, etc.".

Now I wouldn't expect them to magically have a beter connection, laws of physics being what they are, but it raises this point- if the ISP can tell roughly by post code how fast your connection will be (that is, they have method of predicting your maximum possible connection speed) how can they get away with selling the internet with a vastly inflated "up to" value?

Surely they are not only admitting that they cannot give you "up to" 8MBps, but also demonstrating that they knew this, in principal, all along?

Re:Sneaky, yes. Lies, not quite.

[Meneth]

Do you think Doritos would be allowed to sell bags as "up to a pound" when they averaged 9oz and some had quite a bit less?

Yes. Anyone stupid enough to buy it will get what they deserve.

Assuming sufficent intelligence in the majority of consumers, Doritos would be quickly driven out of business (or forced to change) by competitors who use exact wheights.

Re:Sneaky, yes. Lies, not quite.

[teh kurisu]

In this day and age, I don't know why ISPs don't just advertise their speed as "the maximum that your line will support" instead of showing the actual number. 9 times out of 10 this is what they're providing anyway, in the UK at least, with different levels of data caps being used to differentiate their products.

Re:Sneaky, yes. Lies, not quite.

[Nerdfest]

You don't have to buy the item before you find out that it's not on sale.

Re:Sneaky, yes. Lies, not quite.

[Cimexus]

Precisely.

For xDSL technologies I don't see a problem with advertising an 'up to' speed. There's no way in hell an ISP can break the laws of physics. ADSL2+ is capable of 24 Mbps over a short line, but the attainable sync speed drops off rapidly once you surpass 1 km or so.

But since everyone's phone line is different (in length, copper gauge, number of bridge taps and joins etc.), an ISP can't psychically predict what your modem will be able to sync at before they connect you. All they can do is say "you'll get the maximum speed your line will support, up to the theoretical maximum of 24 Mbps". xDSL is a 'best effort' service - you'll get whatever speed your modem can manage to squeeze out of your line and no more or less.

Indeed, how ELSE could they possibly advertise it?

(Fibre and cable on the other hand should be able to deliver the advertised speed regardless of your geographical position, and in my experience, mostly do. Used to be on a FTTN network that advertised 8Mbps and that's ~exactly~ what I got. But on my currently 'up to 24 Mbps' ADSL2+ connection, I only get 6.4 Mbps on a good day, 5.9 Mbps when it rains, due to my 4km-long phone line)

Re:Sneaky, yes. Lies, not quite.

[Dachannien]

A Berkshire Hathaway company here in the US has advertised with "can save you up to 15% or more on car insurance"

Are you saying that ISPs are advertising to cavemen because they're stupid? I find that offensive!

Re:Sneaky, yes. Lies, not quite.

[Cimexus]

How can an ISP possibly tell the condition (and hence maximum speed) of a potential customers phone line?

They can't even get anything more than a VERY rough estimate of line length (since phone lines can be of radically different lengths even just a few houses apart, due to differing routes to the exchange). Add to that other factors such as the quality, age and gauge of the copper, the EM interference environment a particular line passes through on its path to the DSLAM, random uncontrollable interference or damage to the line inside the customer's premises itself etc, and you have no real way of predicting the speed possible over that line, without physically connecting them up and testing it.

All they can realistically do is offer an 'up to' speed. It would be a nightmare trying to assess each potential customer individually and 'advertise' a particular speed that could be offered to them.

Re:Sneaky, yes. Lies, not quite.

[Internetuser1248]

there are many other examples where it does happen though. The term "from only [low price]" is frequently used and rarely means you will be paying anywhere near that if you are serious about buying said product. Drug advertising is bad to "excitocal(R)(C)(TM) can make you stronger, fitter, healthier, live longer, have more friends, your wife will love you again and that bully who beat you up in high school might meet with an unfortunate accident" but rarely do people experience a significant percentage of these results. I would like to start a petition that the topic title be changed to "advertisers tell white lies about products and services"

Re:Sneaky, yes. Lies, not quite.

[jez9999]

They said "up to" not "averaging" so anyone who speaks english and has an iota of mathematical sense should know that they aren't going to see those speeds.

We're talking about Americans, so unfortunately neither of those is likely to be true.

Re:Sneaky, yes. Lies, not quite.

[HereIAmJH]

Most people, even most hardcore nerds, don't download anywhere near that much.

Yet.

I launched my first BBS at 1200 bps and at the time it was a pretty good speed. A year later the big boards were installing 9600s. And a couple years later it was 56k. Early BBSs were text based message boards and didn't need a lot of speed. I had friends that were perfectly happy at 300 bps because that was as fast as they read. (no need for paging) As modems got faster BBSs added file downloads and graphics.

With ISPs we are working from the other side of equation. Speed has outpaced the providers ability to send content because we've gone to a shared medium. (modems provided a dedicated link from provider to consumer) But just like with BBSs, as speed increases so does content demand. Content demand will just continue to increase as new services are created. On a 1200 bps link you were happy with your 100k gifs of nekkid women. With broadband you want 100m mpgs. When you get into streaming video you're looking at 1-2g per hour for standard def TV. If you switched from broadcast TV to entirely online providers you would be looking at 75-150 hours of TV per month, assuming you don't degrade picture quality. That won't be acceptable in a year or two when everyone wants Netflix to send them 3 or 4 HD movies a week and HD content from the networks. So 150g per month is quickly going to limit your average couch potato consumer. Hardcore nerds sharing their Ubuntu DVDs are the lightweights in this environment.

Look at it another way. AT&T and Verizon have limited their data plans to 2g and 5g per month, respectively, saying that it is more than sufficient for their current definition of the service's intended use. The 2g is kind of tight if you're a regular internet user tethering your laptop, 5g isn't much of a problem if you also have occasional wifi access for large downloads and software updates. Neither limit is a problem for a regular user who is only consuming with their smart phone. My average session is about 10m reading all my news sites, unless I get into articles with linked photo essays and videos. And then it's pretty easy to hit 150m in a session. But I know people who frequent youtube that are going to have problems with 5g monthly limits.

Re:Sneaky, yes. Lies, not quite.

[HereIAmJH]

And who would work for pay on an "up to" scale? I'm sure companies would be happy to pay someone "up to" four hundred dollars an hour.

It happens every day for people who work on a commission. It's kind of opposite of bandwidth demand, since increased traffic generally equates to higher sales and higher commissions. But regardless, the employer is not guaranteeing a maximum pay rate. At best they are guaranteeing a minimum pay rate at an absurdly low rate and a potentially much higher overall rate. And while ISPs don't guarantee a minimum contractual rate they do give a minimum in practice. Although it may be an absurdly low rate just like in the commission pay rate example.

Another way to look at it is the guy running the cash register down at 7-11 gets paid $8 an hour regardless of whether he has 1 customer per hour or 10. Is he getting $8 per hour or $.80 to $8 per customer?

Re:Sneaky, yes. Lies, not quite.

[delinear]

The issue of course is that if you pay for 6mbps and only get 3 and so decide to downgrade to their 3mbps plan, you'll then likely only get 1.5. This is where the massive unfairness comes in, since they've already proved they can deliver 3mbps to my home but they refuse to do so unless I pay for the next tier of service. The law for offering these variable services should be based on the minimum service you can expect. That way the people getting 3mbps are happy because that's what was advertised (and it makes it much easier to do comparison shopping for a guaranteed minimum speed), the people getting 6 are over the moon as they're getting double what was advertised, but of course that doesn't allow them to fudge their marketing to sound bigger and better than the competition.

Re:Sneaky, yes. Lies, not quite.

[delinear]

65% of people paying for 20Mbps in the UK are getting less than 8Mbps or less. That's not like saying this pack of Doritos is approximately 9oz, it's more like saying this pack of Doritos is anything between 0 and 9oz when in fact the average is less than 4.5oz. If ISPs really cared about selling an average, they could monitor the speeds customers were getting at various intervals and award them a discount based on how much lower than advertised their speed is. Of course, they don't care because the status quo means they can advertise far more than they're actually providing - although even they are starting to see that people aren't going to swallow this much longer (here for instance, a few ISPs will now give you a speed reading for the area you live in before you sign up so you can see what you will likely get - still not ideal but a step in the right direction).

Re:Sneaky, yes. Lies, not quite.

[jedidiah]

> A crate of tomatoes advertised to 600g - do you seriously expect them to hit 600g with 5 tomatoes each time?

This is one of those situations where it's actually handy to have cable and all of those many channels that may not seem immediately relevant to you. If you had ever seen any footage of a modern food processing plant, you would have absolutely no problem expecting a high degree of consistency and predictability out of a crate of tomatoes. The relevant machinery is actually quite sophisticated.

Yes, I seriously expect them to get n grams of product in every package.

In the "worst" case, the error should be in the consumer's favor.

Re:Sneaky, yes. Lies, not quite.

[taucross]

In my case, the packet of crisps most certainly diminishes the further away I get from the shop.

Re:Sneaky, yes. Lies, not quite.

[FrankieBaby1986]

They want "an Internet" or "a Google" or (my new favorite) "the Facebook" ...

On a side note, just FYI, "theFacebook" was facebook's original name.

from wikipedia: "On February 4, 2004, Zuckerberg launched "Thefacebook", originally located at thefacebook.com. "

Re:Sneaky, yes. Lies, not quite.

[jedidiah]

Chances are, that you've never bought a car that's advertised to go that fast and never will.

OTOH, people who do buy those kinds of cars have infact gotten them to those sorts of speeds.

Re:Sneaky, yes. Lies, not quite.

[tburkhol]

Do you think Doritos would be allowed to sell bags as "up to a pound" when they averaged 9oz and some had quite a bit less? The big problem is it's one way.

Businesses do "Free while supplies last" or "All you can eat while supplies last" deals pretty regularly. Bandwidth is more like all-you-can-eat than a bad of Doritos. Your access to bandwidth may be transiently limited by other people sharing your wires anywhere between you and and the server. Your ISP can't reasonably be expected to control any of their upstream providers, much less the ISP of the server, or the quality of wiring inside your home. If the guy in front of you takes the last egg roll from the all-you-can-eat buffet, then you don't get an egg roll. If someone's Brittney Spears torrent fills the upstream pipe, then you don't get your 6.7 Mbps.

If you want guaranteed bandwidth, pay for guaranteed bandwidth. If you're only willing to pay for as-much-as-you-get, then take what you get. Personally, if an ISP says "up to," I expect that means "under extremely improbable, optimal conditions".

Re:Sneaky, yes. Lies, not quite.

[xaxa]

A crate of tomatoes advertised to 600g - do you seriously expect them to hit 600g with 5 tomatoes each time?

Your country probably has rules specifying what 600g means on a pack of tomatoes. For instance, stuff sold in the EU has the estimated sign after the weight, usually, and there are rules about whether 599g, 580g, or whatever, is acceptable.

We don't (yet) seem to have these rules for internet service.

Re:Sneaky, yes. Lies, not quite.

[hedwards]

Actually, you're not right. People pay on the basis that you get that up to number most of the time, because the advertisements from the ISPs imply that to be the case. And given the difficulties of actually measure bandwidth for most people, the ISPs just expect to get away with it. The up to figure is there because they can't provide completely uniform speed at all times and some things that the end user does will impact performance. It is not however a license to opt not to provide the promised bandwidth.

Re:Sneaky, yes. Lies, not quite.

[hedwards]

If you don't have the expertise and technology to keep an eye on the status of the network then how on earth can you maintain it? I think the point is that you can't. The biggest difference between when my folks switched to DirecTV from Comcrap was that DirecTV has the luxury of being able to monitor the connections that people have on a regular basis and often times step in before the customer calls to complain about it. Comcast on the other hand, wasn't able to even ensure that we got a signal most of the time.

If it's so unreasonable to expect that ISPs make a reasonable guarantee of speed, then how on earth is it that some manage to actually do so reliably? I don't think anybody expects ISPs to provide a completely steady connection at precisely the advertised rate, but I do think that it's reasonable to expect that if they're going to charge us for say 5mbps that we get at least 4 mbps on a routine basis.

Re:Sneaky, yes. Lies, not quite.

[cgenman]

As you mention, there are ways of estimating the speed a user can expect to experience with DSL. At BARE MINIMUM DSL providers should state their numbers as a range, I.E. "speeds between 500kbps and 5mbps." By nature of the product, they know they cannot deliver speeds they are promising.

And while DSL has seen a surge of quality in recent years, traditionally DSL providers have been the smarmiest of the lot. Promised speeds that are just way outside the bounds of reality. Line noise that is blamed on the customer, but is really due to outdated switching equipment still on the line. Charging customers for leaving when the promised 5mbps was actually about the speed of dial-up. Blaming the underlying technology when really the ISP was really just criminally oversubscribed.

Re:Sneaky, yes. Lies, not quite.

[kcitren]

The problem with the Doritos analogy is that there is a physical good whose contents remain basically unchanged based on environmental conditions [fires, floods, etc. excluded]. And the contents are fixed and determined by the manufacturer. With the case of ISPs, they cannot guarantee anything outside of their control, which consists of most of the internet, and hence the places your traffic will go. Can you blame your ISP when you get horrible d/l speeds from someone serving up a website from their homes 56k modem? when site gets slashdotted? when a site goes down? Is it the ISPs responsibility to cache the sites for you? You have to pin down the bottleneck before assigning blame and/or claiming false advertising. Now, I dislike the ISPs as much as the next guy, and I know they oversubscribe. I'm willing to make that trade in order to keep my cost down. Things are expensive, generally a lot more expensive then people think. I know that if I wanted a fixed, high bandwidth connection, I'd have to pay a lot more for it, and I'm happy with the price/performance I get from my current connection.

Re:Sneaky, yes. Lies, not quite.

[poetmatt]

the main point is up to doesn't advertise an actual speed. It only advertises a maximum.

If we say "your connection is 6mb/s" then you should be able to expect that. If we say "your connection is up to 20mb/s" then that doesn't tell you squat as far as what your actual speed is.

Re:Sneaky, yes. Lies, not quite.

[Tim C]

I'm sure companies would be happy to pay someone "up to" four hundred dollars an hour.

I often see salaries in job adverts quoted as "up to X" - at the interview you negotiate the actual rate. The only real difference here is that by the time you negotiate the actual bandwidth you're already under contract.

But I do agree with you; I think it should be fine for companies to advertise a service as "up to XMbps", as long as you get to back out of the contract if the actual speed is not high enough (where "high enough" would obviously be open to interpretation and arbitration). That's unlikely to happen however, given that provision of the service incurs a cost that someone would still have to swallow.

Re:Sneaky, yes. Lies, not quite.

[Inda]

I think nothing of downloading 5gb on Xbox Live. Those demos are so helpful.

You don't need to be into Warez these day in order to download a metric shitload of files.

Re:Sneaky, yes. Lies, not quite.

[Pharmboy]

It happens every day for people who work on a commission.

Bad analogy. Working on commission means your pay is directly influenced by your efforts, and there is no direct comparison for "up to" with internet connections. Same with the second, which is flawed at too many levels to cover in a single paragraph.

We are talking about consumer goods, which should be advertised accurately. A closer analogy is saying "We are hiring people for $8 to $80 an hour", except in this case, no one ever has been hired for $80, and no one ever will, AND the highest they have ever paid an hour was $40. THAT is a bit closer to the analogy you are looking for. Regardless, the cable companies never deliver the promised "max" to any customer, ever, or if they do, it is to less than 1% of the people, and rarely even to them. That is misleading, at the very least.

Re:Sneaky, yes. Lies, not quite.

[southlander]

Plenty of products are sold with "fine print" disclaimers. Of course there is no cheating with bagels and burgers in those analogies and getting by. But there are better examples. Life insurance plans where the advertised "price" is 19.99/Month, then in fine print "Per Unit". Then you find out later it takes 3 "units" to make the plan you wanted. Then the REAL price becomes 59.97/Mo. Lots of selling relies on this type of trick.

Re:Sneaky, yes. Lies, not quite.

[sorak]

I also live in a small town and get good speeds from Comcast. I still wish there was an alternative, so that I could get television service elsewhere without being penalized for not being a ComCast tv customer. (How is the $15 surcharge not bundling, again?)

Re:Sneaky, yes. Lies, not quite.

[natehoy]

True, but line speeds are variable, and can't be controlled as well as bags of chips.

A more amusing, though probably no more enlightening, analogy is a crowded all-you-can-eat restaurant with very small plates.

  - Those sitting right near the buffet line get the best throughput. They can see when interesting new things come out, and it's a shorter walk, so they can make more trips in less time.
  - Those sitting a little bit away can make fewer trips, because the trips are longer. There's also a chance they'll be delayed by the crowd between them and the buffet line.
  - Those sitting the furthest away can make the fewest trips and will be blocked more frequently by closer patrons.
  - The restaurant owner realizes that the average person eats 2 plates, so he charges for three in his "all you can eat" plan.

If everyone eats 4 plates of food, he loses money. So he's got to come up with a way to get people to like his restaurant enough to pay for the 3 plates but only eat 2. If everyone decides they want 4 plates, he's going to have to charge for 5. So he'll take the people who are more likely to eat more than 2 plates and try to find a way to get them to not like the restaurant, or eat less while they are there, or charge them a premium for access to larger plates, or something.

So, of course, he takes the larger folks and places them as far away from the buffet line as possible, to reduce their throughput of food. OK, that's really not a good part of the analogy, but I used to weigh a lot more than I do now, and don't think for a second that I'm making that part up.

Re:Sneaky, yes. Lies, not quite.

[tophermeyer]

the last can of Pringles I opened said on the side that there are at least 100 chips in it.

I'd be fine with that. I would love ISP to be straight up about the lowest possible speeds. Advertising or guaranteeing the minimum possible bound would be fantastic.

Re:Sneaky, yes. Lies, not quite.

[tophermeyer]

In the "worst" case, the error should be in the consumer's favor.

Bingo. I have no problem with producers or service providers giving estimates about the quantity or quality. But that estimate should be the lowest typical, or even the average under typical conditions. That estimate should be qualified e.g. pre-cooked weight or MPG Highway. That estimate should not be an unqualified maximum theoretically possible.

Think of it like an SLA. When I buy a pack of tomatoes advertised as 600g, I should expect to receive at least 600g of tomato. I might receive faster service or more product, but so long as the advertised expectation is met that should be sufficient.

Re:Sneaky, yes. Lies, not quite.

[Totenglocke]

You'd be surprised - even most $20,000 family sedans today can hit around 130mph. Just because they don't advertise it on TV doesn't mean that the car can't do it.

As for your assertion that the people who buy those cars max them out - I'd say, in general, that I'd have to disagree. In my experience it's much more common for people to buy expensive and fast cars and then never drive them fast because they're scared of hurting the car.

Re:Sneaky, yes. Lies, not quite.

[Rary]

there would be all sorts of uproar if physical goods were advertised and sold the way broadband is.

You mean like having a sale and advertising that all items are "up to 50% off their regular price"?

It happens all the time. When a key selling feature is variable, "up to" is pretty much the standard way to advertise it.

Re:Sneaky, yes. Lies, not quite.

[drinkypoo]

The reality is that you cannot give the same speed to all users because there is simply not enough available bandwidth for a broad variety of reasons not limited to oversubscription or marginal wires that have reduced bandwidth due to some form of damage. The only way to guarantee the same speed to all users is to cap all users artificially and to cancel the subscriptions of all the users who cannot achieve even that speed... which is what happened with DSL. Regulation meant that Pac Bell and later SBC got slammed hard any time anyone complained about poor DSL speeds, so they cut off everyone over 14,500 feet that was having problems, and stopped selling over that distance. So sure, more people got advertised speeds, but meanwhile a lot of us got NOTHING and some people even had their service terminated. This is what you are asking for if you ask for guaranteed speeds. Asking for the right to break the contract if you don't like it is also asking for them to terminate your contract if you don't achieve basic speeds, because having you as a customer is a potential liability.

Re:Sneaky, yes. Lies, not quite.

[Muad'Dave]

ISPs aren't the only industry that uses this sales technique. I hear ads for "discounts up to 90% off!" for cars, furniture, discount shopping stores/web sites, etc all the time.

As long as at least one single item is in fact 90% off, they are not misrepresenting anything.

According to the '10 Mbps claimed' graph in TFA, 6% of users got 10+ Mbps - that's actually higher than I thought it would be.

Re:Sneaky, yes. Lies, not quite.

[clone53421]

Do you think Doritos would be allowed to sell bags as "up to a pound" when they averaged 9oz and some had quite a bit less?

No, but that’s because the laws are written so as to require Frito-Lay to guarantee their claims of what is in the bag. They have to sell bags that are “at least a pound” if they want to label them 16oz.

The law makes no such requirements of ISPs.

Re:Sneaky, yes. Lies, not quite.

[CAIMLAS]

It really depends on where the limit is being imposed.

If I subscribe to a 10Mbit pipe, this should mean that my modem/router device and their network is able to sustain those rates. That's all. If I'm not able to pull content down at those rates, it doesn't mean they're in breach of the service terms - and that's likely why most ISPs put those terms in there. They can't control anything outside their network.

Now, if you can't get the advertised speeds on their network, that's another story.

What I think has happened is ISPs have upgraded some of their equipment to DOCSIS 3 to get the higher throughput, but DOCSIS 1 and DOCSIS 2 equipment is still out there - in homes as well as near the ends of nodes. This will bring the overall averages down - even if 90% of people get what they pay for, those people with 1Gbit FTTH pipes are going to be limited pretty close to home.

Combine that with companies over subscribing nodes, and you've got the same problem we had 10 years ago with early adoption. (Oh, and remember all the promises they were making and never kept?) Hardly new for ISPs.

Sometimes, I miss the small mom and pop dialup ISPs. You may have been limited to 9600/14400/36600/etc. baud but the people running things took pride in their work, it seemed.

Re:Sneaky, yes. Lies, not quite.

[CAIMLAS]

I actually like the parent's Doritos analogy - it's true when you think of it that way; there would be all sorts of uproar if physical goods were advertised and sold the way broadband is.
"Up to" a dozen bagels in your order, or "up to" two patties on your burger would never fly. And who would work for pay on an "up to" scale? I'm sure companies would be happy to pay someone "up to" four hundred dollars an hour.

Been to a fast food restaurant lately?

No, they don't typically advertise things using words, but use pictures instead. But when and where have you been to a food establishment with pictures in the menus where the food looks anything like what you actually get? Usually, it's as little as 1/3 as large or a substantially smaller portion. The meat and veggies are of a significantly lower quality/grade and so on...

False advertising is everywhere. Most people just consider it "the eye of the beholder" though.

Re:Sneaky, yes. Lies, not quite.

[pruss]

Suppose you play one of those fast-food restaurant scratch-off games, and you're told that every ticket is a winner, with the smallest prize being a coupon, and that you can win up to $100,000. I don't see any innate deception here. Yes, you can win up to $100,000, but every sensible person knows that only a very small minority of the players will do that. Likewise, I don't see anything all that wrong in principle with a grocer having some weird promotion like: "For $3, I'll sell you a random bag of nuts. You'll get up to five pounds." If you get five pounds, you're lucky and you've won the nut lottery. But you should expect to get less. There really isn't any innate deception. There may be some non-innate deception, though, in both cases due to the fact that people don't deal with probabilities very well.

Re:Sneaky, yes. Lies, not quite.

[Shakrai]

I see the speeds promised by TWC all summer.... until the SUNY college brats return to town. Then my peak usage (9pm to 12am) speeds drop from 7-8mbit/s to 3mbit/s or less. It's bad enough that I can't stream Netflix without it pausing every five minutes to buffer or dropping down to an unacceptably low quality level.

Incidentally, I'm paying for 10mbit/s. Used to pay for Turbo @ 15mbit/s but what's the point if you can only achieve that speed at 4am?

Re:Sneaky, yes. Lies, not quite.

[stealth_finger]

I actually like the parent's Doritos analogy - it's true when you think of it that way; there would be all sorts of uproar if physical goods were advertised and sold the way broadband is.

broadband isn't a physical good though is it. The fairest way would be for them to say to consumers something along the lines of "your local network has a peak capacity of XMbps and that is shared bewteen n households. The actual speed you get will vary". Which in a way they do, they just dont tell people and most people don't know that if you have a real net hog living down the street he can adversely affect everyones speed.

Re:Sneaky, yes. Lies, not quite.

[TrentTheThief]

Nine out of ten would be extremely happy to have FiOS over any of the cable offers I'm aware of; cable's best being 15Mpbs (shared throughout the neighborhood, so when the kids get out of school your BW drops dramatically).

I have a 25/25Mbps FiOS connection. I'm definitely not in favor of the cap. Three household users and a very active webserver run comfortably. My server averages about 8Mbps outbound with peaks to 28Mbps (as reported by the local bw app).

FiOS is a life changing experience for many people. Especially older adults who never used a 300bps modem, by that I mean the people who don't remember a download taking a couple hours to complete. For a new middle-aged or older adult, FiOS is perfect to start with. They generally lack the patience required to wait for a graphics-heavy website to load, or to wait more than a few minutes for a file download.

I think nothing at all of grabbing few linux ISO's to play with. I wouldn't have done that before FiOS. If I needed a new ISO, I'd kick it off and go to bed. Now I just go get some coffee and check my email and it's there in 15-20 minutes. I love FiOS.

I hope they don't cap it. I've been toying with getting a BD player with netflix capability...

Re:Sneaky, yes. Lies, not quite.

[Muad'Dave]

I wondered what that meant. Why did they choose to create a new symbol instead of using the existing symbol '±' or the abbreviations 'approx' or 'est'?

Re:Sneaky, yes. Lies, not quite.

[DarthVain]

Companies do this all the time. Its sneaky, but not new, or news.

"SAVE UP TO 80%" Translation: We have one crappy product that is 80% off. There are 10 physical copies of them across the entire country. Everything else will be regular price or 5-10% off. We will likely inflate our prices a week before anyway to make it seem like a better deal. "BUY BUY BUY"

I don't pay any attention to those ads, some stupid people do and get angry. Likewise is the speed issues, its a bit different because it is technical and not everyone knows that, so even more sneaky. You can get 10mb/s if all the planets are aligned just right, its just not common or really practical. You can say the same thing about Ethernet speeds, or Hard drive speed, or any speed not actually set or driven by an actual clock. They are all theoretical. Last time I transferred files across my network I only got 40mb/s not 100mb/s as advertised. The mb/s on my Hard drive is even higher, but that doesn't happen either. It is however about par for the course that are the dirt bags that run our telecommunications around here. They have no competition, have no enforced regulation other that to obtain a monopoly so it is no surprise they do the stupid things they do. We let them screw us, so logically, they do.

Re:Sneaky, yes. Lies, not quite.

[DarthVain]

Which is why in recent times they also put caps on everything.

Re:Sneaky, yes. Lies, not quite.

[networkBoy]

Actually,
That's not quite true.
They tend to look for people that naturally are ideal looking, but through injury or other reason gained weight. They then provide their product to these people while on the mend and once healed up the models look good again.

Total sham.
-nB

Re:Sneaky, yes. Lies, not quite.

[xaxa]

± isn't appropriate, since the tolerance isn't symmetric (1kg e means not less than 985g, I don't think going over 1000g matters...).

I don't know why they used e, but some guesses:
- Approx and est are English, the stylised 'e' isn't.
- Approx and est may have had an existing legal meaning, or existing use with no legal meaning

Re:Sneaky, yes. Lies, not quite.

[networkBoy]

I can get a consumer DSL line "up to" or I can buy a commercial DSL line that is "between X and Y" from my carrier. If I want "X" on the spot I have to buy something else (I inquired about DS3 and they thought I was joking, once I convinced them I was serious they said I can't have it :( ).
-nB

Re:Sneaky, yes. Lies, not quite.

[fat4eyes]

The main problem about advertising "truthful" data rates is that the internet is best-effort. It's impossible to guarantee any data rate on the internet when congestion is involved. Car analogy: it's like buying a Lamborghini because it can go up to 200mph, but then complaining that you can't go above 30 during rush hour traffic. When your ISP gives you a 10mbps router when you're paying for a 20mbps connection then you can complain. Otherwise, just deal with it, it's how the internet works. Or use DSL instead of cable, as many others here have mentioned.

Re:Sneaky, yes. Lies, not quite.

[xaxa]

But that estimate should be the lowest typical, or even the average under typical conditions.[...] When I buy a pack of tomatoes advertised as 600g, I should expect to receive at least 600g of tomato.

In the EU, your worst case is 585g of tomatoes, but more statistics than I'm familiar with means you'll probably get 600g.

(See the formulae in the relevant law -- last two pages.)

"As regards the criterion for the mean calculated by the standard deviation method, a sampling plan used by a Member State shall be regarded as comparable with that recommended in Annex II if, taking into account the operating characteristic curves of the two plans having as the abscissa axis (Qn - m)/s the abscissa of the 0.10 ordinate point of the curve of the first plan (acceptance probability of the batch = 0.10) deviates by less than 0.05 from the abscissa of the corresponding point of the curve of the sampling plan recommended in Annex II.". Um... what?

Re:Sneaky, yes. Lies, not quite.

[Quirkz]

I actually like the parent's Doritos analogy - it's true when you think of it that way

If the weight of a packet of crisps diminished the further away you got from the shop, it would not only be perfectly legal, but prudent to advertise as "up to 200 grams". So the Analogy is terrible

Agreed. You just can't compare the delivery of a physical product that the manufacturer has complete control over to the delivery of a service that can be affected by a multitude of outside influences. Bad logic all around. It's like asking, "If Doritos can put the right weight of chips in the bag, why does my shower have to get hot when someone flushes the toilet?" The situations aren't even close.

(Realizing, of course, that many modern houses don't have that problem, but clearly the internet still does.)

Re:Sneaky, yes. Lies, not quite.

[Orestesx]

Facebook originally was thefacebook.com. I have an email from 3/15/05 from thefacebook.com.

Re:Sneaky, yes. Lies, not quite.

[Rising Ape]

If ISPs really cared about selling an average, they could monitor the speeds customers were getting at various intervals and award them a discount based on how much lower than advertised their speed is.

Given that the lower speed is usually a technical limit of ADSL, it costs the same for the ISP whether the end user is getting 20 MBps or 2 MBps. So I can't see that idea going down well.

Re:Sneaky, yes. Lies, not quite.

[xenapan]

The problem in both cases is the average. Using the chips example, if they advertise 18oz, they are required to at least average a good % of that. probably like 95%. On the other hand, scratch to wins with everyone being a winner, you can expect 99% "free small drink with purchase of your next full meal" the average "prize" won't exceed a % of the profit for the minimum purchase required to get one of the scratch and wins.

The difference between the two however is the fact that one is the actual product you are buying (like whatever your ISP offers. cable or DSL) and the other is a "bonus" like those speed boost things. You are really comparing apples and oranges. Getting 22mpg hwy from what they claim 25mpg is roughly 10%+ difference. on the other hand, averaging 50% on a 6.7mbs connection is a huge difference?

Oh.

I accept the fact that they cant control the speed to the server you are attempting to reach due to limitations out of their control. But when they have full control and people are CAPPED at a less than 50% fraction of what is advertised. Thats just false advertising. The biggest problem is the fact they charge an arm and a leg to "come diagnose the problem" I've been suffering from intermittent connection issues at home and I haven't changed any router settings in months etc. The chances are theres a problem with their hardware (either their modem or switch box) when the modem has to reconnect to the ISP 30+ times a day on average. I'd switch except there aren't any other real choices.

Re:Sneaky, yes. Lies, not quite.

[Xyrus]

Then why just stop at "up to 6.7 Mbps"? Why not say "up to 100 Mbps"?

At what point exactly does it become an outright lie? When the claim is 3 standard deviations away from the mean? Or 6?

Is it a lie if that is the HYPOTHETICAL maximum speed but the only way you can get it is by sitting at their center plugged into their rack? Or by having no one else online?

What if it's impossible for anyone to ever attain that speed under anything but the most ridiculous and unlikely scenarios? Where can we draw the line at being optimistic and being dishonest?

If in practice no can or does achieve the "up to" limit, then it is misleading (at best).

Re:Sneaky, yes. Lies, not quite.

[CCarrot]

I do. Try letting off the loud pedal every once in a while, sparky. For the record, I'm a 26 year old male, who likes to drive fast and corner hard, and I still manage to get the suggested mileage out of my car.

True that. Unfortunately, MPG is something which you have some control over (drive like a dick, you pay for it in gas), whereas broadband speeds are not. Broadband speeds depend on subscription rates and traffic over existing infrastructure, which you have practically no control over (unless you want to run around and start snipping your neighbour's connections).

Perhaps a better practice would be for the provider to provide 'up to' and 'average' posted speeds? The average speeds could be made available by neighborhood and calculated based on subscriber base, infrastructure and recorded actual speeds from residential modems. That way people would be able to make informed choices based on real world expectations, rather than grossly inflated perfect conditions.

Re:Sneaky, yes. Lies, not quite.

[greed]

What they could do is, actually advertise the bit you said: '"you'll get the maximum speed your line will support, up to the theoretical maximum of 24 Mbps". xDSL is a 'best effort' service - you'll get whatever speed your modem can manage to squeeze out of your line and no more or less.'

Since they actually advertise the maximum speed as if everyone will get it, the "up to" part is in the fine print. Or when you phone up to order, they go on and on about how fast the lines are... the colour glossy brochures that they dump in your mailbox talk about the top speed... every time you hear from the ILEC ISPs you never hear or see "up to"... unless you get a stethoscope and magnifying glass out.

They also, or at least Bell Canada, charge you for the "up to" speed regardless of what you actually get. If I can't get 5/800, I can only get 3/512, why am I paying the 5/800 rates?

Well, actually, I'm not paying for any of that now, I'm paying the cable company. Only partly because of the crappy uplink on DSL; the other part is every time a truck rolled through my neighbourhood with Bell's logo on it, either the phone or the DSL would fail. Fixing one would, 1/2 the time, break the other.

Re:Sneaky, yes. Lies, not quite.

[jjk3]

I completely 100% agree that what they advertised is a scam, but if I step back a couple of years, it's completely amazing that we are complaining about only getting 8Mbps.

Re:Sneaky, yes. Lies, not quite.

[Aliotroph]

My ISP tries to do this without really saying it. They say we get up to 15 Mb/s, but if it's nowhere near then we can usually get them to put some effort into fixing it. Most of the time we really get speeds between 15 and 20 Mb/s.

Re:Sneaky, yes. Lies, not quite.

[HereIAmJH]

It's not an analogy. Parent post said And who would work for pay on an "up to" scale?, and that is exactly what commission work entails. Another example, go to CareerBuilder.com and search for "work from home". Nearly 20k hits and the first page is all Work From Home and Earn $500 - $1000/Week - Easy!. They are offering to pay on a scale with no definition of the job or the scale. But even if you were selling cars at the local Chevy dealer, your maximum pay rate isn't necessarily achievable because it entirely depends on traffic that is out of your control. If your employer 'oversubscribes' sales staff, your commissions will plummet because just like bandwidth, the size of the pie is limited.

Re:Sneaky, yes. Lies, not quite.

[Pharmboy]

I'm guessing you haven't worked commission at a legitimate company. No company worth working at tells you "Earn up to $x per week!" in my experience. If a company has to inflate their pay potential, then they are not legitimate. Most of the commission jobs I have had (and most of my long life has been on commission), they might tell you what others are making, and obviously tell you what percentage you are getting, but they don't advertise "Make as much as $x" because any legitimate company knows that this isn't cricket, and in most cases illegal unless they can demonstrate this. Of course, on the internet, who can police work at home jobs? No one. That doesn't make the companies legitimate, and in fact demonstrates they are not. The vast majority of real commission jobs (ie: you can walk into the actual business) are not advertised that way, primarily because it would likely violate labor laws in the same state.

Re:Sneaky, yes. Lies, not quite.

[MokuMokuRyoushi]

The difference being that maxing out your internet connection is lawful, safe and attempted frequently. This analogy is garbage.

Re:Sneaky, yes. Lies, not quite.

[dotlin]

If you really want to be told specifically what speed you'll get, feel free to offer your ISP some $$$ to send out a technician and find out the combination of umpteen factors that will give you your result. It may take quite some $$$ but you'll be able to convince them - and you will get your precious accurate values.

Or just click on: http://www.youtube.com/my_speed/

Re:Sneaky, yes. Lies, not quite.

[cwsumner]

DSL is a dedicated line at constant speed. Oversubscribing doesn't apply. Of course if you have several of your own computers on your line, then they are sharing. But your speed is not being shared with your neighbors.

Cable, however, is shared with your neighbors. So on cable you are not going to get the max speed except in the middle of the night. Cable is not faster than DSL except every now and then. The ad's lie!

Re:Sneaky, yes. Lies, not quite.

[Jaime2]

Not sneaky, just lazy consumers. People should really read statements like "Up to 6.7Mb/s" as "I have no idea how fast it will be, but it won't be higher than 6.7Mb/s". Unfortunately, people read it as "Pretty close to 6.7Mb/s". A more interesting variant is "Up to 15% off or more", which really means absolutely nothing, as in, any number satisfies the given constraint. If people read these statements literally, they wouldn't buy services based on them and marketers would have to come up with new statements. Unfortunately, we gobble these things up. People give me crap all the time for being too literal, but I think the whole world has gone way overboard the other way.

Technically correct

[fluffy99]

Yes, some customers are getting "up to" the advertised speed. Since all the advertising says "up to" this isn't lying. Where's the story in this?

Re:Technically correct

[garcia]

I'm one of those people and since moving to Minnesota I've always been one of those people. My DSL was "up to 2mbps" and I received 4.7mbs and with cable I'm "up to 7mbps" and receive 20mbps (18.85mbps according to my most recent aptitude safe-upgrade).

This is a non-story in my most recent experience but I'm guessing it's not for others. In a previous life, back when I had DSL in college from VZW, we would routinely see 400k/64k speeds on a connection that was supposed to be 768/256. YMMV.

Re:Technically correct

[MBCook]

If I sold toilet bowl cleaner tablets that hang in the tank, and say they are good for "up to 1000 flushes", would it be OK if they worked for only 500 flushes for the majority of people, and the rated amount for less than 5%? No one would accept that.

When other industries advertise something (the weight in a bag of food, or of some raw material) they are advertising mean, and they have a lot of quality control to keep close to that number. Too much and they lose money, too little and people stop buying or they get sued for false advertising.

But that doesn't happen in broadband. They think it's OK for the speed to be way less than the rated, but it is almost never higher (let alone by 50%). But I have two choices right now. I have DSL that maxes out at 6mbps, and cable that is supposed to go to 24mbps. But if the top cable tier delivers 8, what am I supposed to do? It's the fastest available.

When bags of concrete mix turn out to be light, contractors stop buying because they are being ripped off and can buy another brand. The free market works there. Broadband has so little competition in most places (the majority of americans only have 2 choices, many only have one) that the options are usually "pay and suck up the false advertising" or "have no broadband at all".

They aren't selling 24 and delivering 21, they are selling 24 and delivering 12. That's not a "not always quite there", that's "complete exaggeration."

Re:Technically correct

[fermion]

It is not lying,but it may be deceptive,and deceptive advertising is an actionable offense. Two examples of not lying that got retailers in trouble.

Many years a US discount retailer would make up prices and discount off these prices. The ad copy basically said they were made up, but in such a way that the consumer would think there were based in fact. This enabled the retailer to offer 50% savings on almost everything, though the prices were comparable with any other discount retailer. The company, whose name slips my mind, is out of business.

A department store, maybe Foley's, also got in trouble due to a tactic that many would think was legitimate. They would offer clothing at a rather high price,then advertise a sale discounting off the high price. Now, these products were actually offer for sale, so the retail price was legitimate, but it was still seen as deceptive as there was no intention by the retailer to actually make a sale at this price, just to set a price for advertising a discount. There might have been some sales at the high price, but that was not an issue. This practice is not illegal, but one will see ad copy that states no sales may have occurred at the advertised high price.

So really, on one hand this is not a big deal. The 'up to' might be enough. But given these two cases, and the fact that so few people get the 'up to' amount(much les than 10%), I would say additional ad copy would be required to make this legit. At minimum I would think a note saying that nearly no one achieves this speed. Ideally I would like to see a listing of the speed that the second and third quartile gets, in this case 0.5-2 mb. This would be most useful for the consumer as it would at least help the consumer know the kind of speeds they are likely to get. The fact that this is not done clearly indicate the 'up to' numbers are meant to be deceptive.

Re:Technically correct

[c6gunner]

It's dishonest and intended to rip off the customer. They don't promise to give you "up to 500 TV channels" - they structure it so that you pay for what you want. Bandwidth should be held to the same standards.

Re:Technically correct

[Sancho]

Actually, other industries do use this kind of math:

http://online.wsj.com/article/NA_WSJ_PUB:SB126092257189692937.html

In fact, those are worse, because they don't use the phrase "up to."

Re:Technically correct

[lopaka1998]

Once a time warner ad stated "8 megabyte connection". It was pulled quickly - but still. Apparently the reader was arrogant of internet terminology and incorrectly substituted mbps with megabytes instead of megabits. Technically speaking they were advertising a 1,024 megabit connection. Have you ever noticed that ISP's always state that your game will lag or whatever without a 15mbps connection? Makes me wonder why they don't ever mention the effects of ping speed between two computers.

Re:Technically correct

[sjames]

At one time, bandwidth was sold based on committed rate and burstable. The burstable rate was the technical limit of the circuit. Where only one rate was mentioned, it was understood to be the committed rate.

Based on that, all ISPs and many cheapie colo facilities offer you a big fat ZERO Mbps burstable to some figure you'll never see.

The problem these days is that there is so much deceptive advertising going on that honest players can't compete at all in many areas. They must either try to tell themselves that big fat lies are the new truth or shut the business down. The FTC has let us all down.

Re:Technically correct

[daveime]

My cable promised me about 60 channels ... unfortunately they failed to mention that half of the programming is either commercials or trailers for other shows. So essentially I only have 30 channels to choose from at any given point.

If they said "up to 60 channels" it's possibly MORE honest in this case.

Re:Technically correct

[smellotron]

Light bulbs tend to be rated by "up to X hours" lifetimes. Nobody seems to complain about that. But then, it's very tangible when a light bulb goes out.

Re:Technically correct

[Anonymous Coward]

Forget all this download speed, where the fuck is my upload speed? Why the hell do I want 15 Mbps down when I still have 786k upload? WTF is that?

Re:Technically correct

[Olipro]

This is a terrible analogy; they're advertising a rate of download which you will probably not get, nonetheless, whilst whatever you want to download might well take longer than expected, you'll still get your download. So here's an analogy that works; the advertised weight of the chips/burger/whatever is exactly the same, what you're being told is that you can expect to spend 10 minutes consuming it when in reality the only people who are going to spend that long on it are small children; either way you still got your bag of chips, you're just polishing them off faster than you'd expect.

Re:Technically correct

[Dr_Barnowl]

But they don't get sold on "up to 100 watts" and turn out to be a 50.

Re:Technically correct

[boxwood]

If a retail store advertises a sale thats "up to 50% off" but many items were only 25% off, and some items weren't even on sale, no one would accept that!

oh wait people do accept that because they understand what "up to" means.

Re:Technically correct

[c6gunner]

They don't promise that the websites you visit will be free of ads, either. Likewise, your phone company doesn't say anything about how often you might get a call from a telemarketer, or how often the person you're trying to call may be unavailable. There's a huge difference between expecting your access provider to guarantee their service, and expecting them to guarantee the availability or quality (subjective) of the content transmitted over their service.

Re:Technically correct

[Inda]

It's against the law in the UK. Items must be for sale at the higher price for 30 days before any sale can be advertised.

Re:Technically correct

[blackbear]

this isn't lying. Where's the story

Are you astroturfing? Of course it's a lie.

In a strong tail-wind my car really can get 4-5 more MPG for an extended period of time. However, using that as a metric for a car's performance, while factually correct under certain conditions, denies the implication of routine repeatability, and makes the metric unreliable.

It may be technically true that I can get the maximum bandwidth at 02:00 for 2 hours, but that's of no use to me. So I toss out the idea of ever getting the theoretical maximum and instead use maximum bandwidth as a metric to set my general expectations, and to decide the relative value of the exchange (bandwidth for money.)

The ISP knows that maximum bandwidth is my metric, and attempts to deceive me by inflating that metric in an unreasonable and unexpected way. This is exactly the same as inflating MPG ratings on cars based on theoretical maximum conditions.

The lie is, knowing that routine repeatability does not exist for the numbers they quote, ISPs prominently advertise this useless metric in order to deceive the customer as to the true relative value of the product.

The news is that this report proves it.

Re:Technically correct

[Tim C]

Bulb lifetime depends upon the particulars of the installation and usage patterns - e.g. a bulb that is left on constantly with a nice steady power supply will last longer than one that is frequently switched on and off and/or has a power supply that suffers frequent large fluctuations from spec.

To an extent ADSL suffers similar problems - even if my ISP is perfectly capable of delivering a continuous 8Mpbs connection, if my phone wires are too poor I won't achieve that speed (which is exactly why I switched to cable). Of course, over-subscribing the line so it is frequently saturated isn't going to help either...

Re:Technically correct

[Kjella]

Well, if they aren't illegal in the US maybe they should be. At least here in Norway you can not claim a sale unless you have sold a reasonable volume at that price in the last 6 weeks (what volume depends on the product). You can't claim savings over a "full" price unless you can prove it's a realistic retail price in the market. You can not claim a sale on goods or services that are individually priced and there's no basis for establishing a normal price. If you've bought a batch of goods at low price, you can advertise it as a "batch good" (partivare) with limited availability but not a sale.

Another funny one they caught was price "guarantees" when the product is only available at that retailer, e.g. washing machines and stuff like that. They'd be slightly different at each chain (superficial feature changes, off by one model numbers) so you couldn't actually use it. Again, there must now be a reasonable volume of the exact model in other outlets to promise that.

Broadband here is quite heavily reviewed and essentially you need to offer the advertised bandwidth most of the time, if you took hourly speed checks on a normal day then at least half should be the offered speed. People that can't quite reach the speed are offered to either downgrade one step or keep their current subscription with partial performance. You can not claim packet overhead into it, it's delivered data speed to the customer that counts so the physical lines are slightly faster than advertised.

Of course this drives prices up somewhat, I'm paying $75/month for 25/5 Mbit. However, it's really real 25/5 Mbit where I can get 3MB/s down sustained if only the other server is fast enough. I haven't really tested my upload since the last upgrade, but back then I got the full 3.6 Mbit I had. Also we have rather strong legal consumer rights that often go beyond any warranty, if the ISPs drop a paying customer for using too much "unlimited" bandwidth they will get into trouble. Only clearly communicated limits are valid, some mobile broadband providers still have that but typically slow you down to mini-speed after your quota is up, no terminations or such.

Re:Technically correct

[MBCook]

No, my point with that was that since most American's don't really have a choice of provider, the market can't fix this since at best you can switch to someone else who is lying to you in the same way.

Re:Technically correct

[MBCook]

Ah, but each individual item is marked with the discount that applies to it, so that's OK. Now if they were all marked "up to 50% off" but you didn't find out what the discount was until after you'd bought it, that would be a problem.

And that's very similar to what's going on here.

Re:Technically correct

[snowgirl]

The weight of a bag of chips is actually supposed to be a minimum amount of chips. The reason they have so much QC on it, is because they want to over-pack as little as possible, and never underpack (as this leads to potential torts of deceptive labeling.)

A good example is the lines on German glasses at restaurants. If you order 750 ml of a drink, and it comes back not at that line, you can legally refuse to accept the drink. But how many people are going to fill it to 800 ml every time just to ensure that they never get a return? Not many.

Advertisers who present a minimum value are going to do their best to ensure that they over-provide as little as possible. Advertisers who present a maximum value are going to have a wide statistical variance on what they actually provide, because there are no legal consequences for them doing so. And it is unlikely that anyone advertises median values, but they're probably the most honest, but still have large variances.

There is no way to correct this problem without regulation dictating that internet connections cannot be sold with an "up to X speed" advertising.

Re:Technically correct

[fluffy99]

this isn't lying. Where's the story

Are you astroturfing? Of course it's a lie.

They are not lying, as the ads are factually true - that is the max you can ever expect to see. Of course its misleading and somewhat deceptive to only advertise the potential peak speeds. They simply are not giving you useful information or even agreeing to provide a minimum level of service. I applaud the FCC for trying to address this, but I suspect like many other things they'll fuck it up or put in enough loopholes to make their efforts ineffective.

Re:Technically correct

[DarthVain]

When you transfer a file along your home network does it travel at up to 100mb/s or 1000mb/s as per the standard. No it doesn't. That is the theoretical maximum. The ISP is basically saying if all the planets align on Friday the 13th, and your favorite color happens to be blue, you MIGHT get 10mb/s.

For some reason people accept that. They should be asking for, OK what is your average speed for you average user. I blame monopoly.

Re:Technically correct

[evilviper]

Since all the advertising says "up to" this isn't lying. Where's the story in this?

Why don't they say "up to" twice the speed? 4x the speed? It's still technically true, right?

Re:Technically correct

[smellotron]

But they don't get sold on "up to 100 watts" and turn out to be a 50.

To a limited extend, they do. CF bulbs usually have ratings for output brightness. However, brightness degrades over time, meaning that a house full of year-old bulbs will be less bright than the "lumens" or "equivalent wattage" would indicate. It's definitely not the same thing as cable over-subscription, but it is a case where the rated numbers are expected to be higher than actual output.

Re:Technically correct

[forceman130]

You seem to be missing an important point. Bags of cement and boxes of ceral aren't advertised and sold as "up to" 50 pounds (for cement) or 18 oz (for cereal). They are sold as that weight, which is why they normally weigh pretty darn close to that weight. Your comparison/complaint would only be valid if the ISPs were offering "8 Mbps", not "up to 8 Mbps".

Re:Technically correct

[daveime]

There's a huge difference between expecting your access provider to guarantee their service, and expecting them to guarantee the availability or quality (subjective) of the content transmitted over their service.

I think this was the main point of the thread. The guarantee means NOTHING if the actual service is crap. I can be guaranteed 3Mbit/s but if I only get 56Kbit/s for 23 hours 59 minutes, and the 3Mbit for 1 minute each a day, it's simply not true. Same as paying for HBO "movies 24/7", then discovering 8 to 12 hours are actually commercials, trailers or the movie I just watched 6 hours ago.

Re:Technically correct

[Deliveranc3]

The story is that a company might be offering 12MB and giving service worse than another company offering 3MB.

However, of course they are going to oversubscribe... it just makes sense. What we need is a fair way of figuring out how good each service actually is.

The only way I can think of to do it is Youtube, which is used fairly universally and at all times day or night, it knows when it pushes more than a down loader can receive if it correlated the information it could give a fairly accurate picture of each service.

Another way might be by doing what Google tried to do unsuccessfully in Germany, examine open wi-fi hot-spots and judge speed as they moved through them.

Barring that kind of analysis it will always be the company's word and hearsay, and we'll never know for sure who's the best and be informed consumers.

Which means we can't vote with our dollars intelligently. That seems like a situation where government should nationalize it. There are many other good reasons to do so as well.

And I want their bandwidth...

[SanityInAnarchy]

You'd think that'd be a mutually beneficial arrangement of the sort that would make Adam Smith proud...

But no, it seems they want to keep my money and their bandwidth, so fuck 'em.

Re:And I want their bandwidth...

[ebusinessmedia1]

How about I start paying AT&T "up to" $30 per month for the "up to" 3mbps that they promise me?

What we have here is pure deception. It's a manipulative deception because we're talking about communication speed that is dolled out in tiers. Communication is a human-species-defining quality. We're wired to want more of it, and this is the reality that all communications companies - ISP's included, bank on.

An analogy would be a company that sold breathable air, saying that they would provide "up to" a certain amount of oxygen per month. They would give you enough, but just enough so that you would always want more. ATT, Comcast and the rest have this all figured out, and they continue to limit the potential of America's social and intellectual capital, in the name of their tunnel-vision profits. I say this makes the senior executives of those companies charlatans and criminals of the first order, because they are stealing our future, as other countries pass us by.

Re:And I want their bandwidth...

[Eunuchswear]

What, you mean like:

"People of the same trade seldom meet together, even for merriment and diversion, but the conversation ends in a conspiracy against the public, or in some contrivance to raise prices."

Re:And I want their bandwidth...

[I_Voter]

So you just discovered that Adam Smith's "invisible hand" can pick your pocket?

Re:And I want their bandwidth...

[commodore64_love]

Maybe as part of net neutrality, Congress will require cmpanies to alter their plans from "upto XXX" to "minimum XXX".

You would be guaranteed that level of speed (or get a pro-rated refund). But most times it would be faster.

Re:And I want their bandwidth...

[AlamedaStone]

Maybe as part of net neutrality, Congress will require cmpanies to alter their plans from "upto XXX" to "minimum XXX".

You would be guaranteed that level of speed (or get a pro-rated refund). But most times it would be faster.

I think that's a perfectly rational approach, but I suspect it's a little to "extreme" a measure for the US to actually adopt. In particular, prorated billing seems like a lose-lose, as the providers would need to invest in bandwidth tracking and consumers would have the monthly burden of bill scrutiny. I'm sure I'm not the only one who has wrestled with sales/customer service reps to get money refunded for one cock up or another. (Particularly Verizon, whose board surely have a special place waiting for them in hell...)

I find myself wondering if the creep of broadband expense may be intentional as part of a larger plan to implement metered bandwidth. Is metered bandwidth better for providers? For consumers? I guess it comes down to rates, but mandating fiber to homes (as one of the reports recommends) suggests there may be a mandated rate schedule too. And if there's a mandate, there will be outcry for Federal subsidy for a net meter in every home.

Re:And I want their bandwidth...

[SanityInAnarchy]

the providers would need to invest in bandwidth tracking

Oh, they have. How else would they implement those monthly caps? Or, before that, how would Comcast have been able to consistently find and ban their top bandwidth hogs for violating an imaginary cap?

I'm sure I'm not the only one who has wrestled with sales/customer service reps to get money refunded for one cock up or another.

"Better Business Bureau" are the magic words.

Is metered bandwidth better for providers? For consumers?

Yes.

I guess it comes down to rates,

Well, yes, but metered bandwidth is better for two reasons: It makes consumers conscious of how much bandwidth they're wasting (and thus put some effort into limiting it), and it means at least we know what we're getting.

there will be outcry for Federal subsidy for a net meter in every home.

Ok, first off, why would you place that in every home, where it can be easily tampered with? Especially in the case of fiber, wouldn't it be easier to have the meter be at the ISP?

And second, how expensive is this, really? At least some "modems" have taken on the responsibilities of being a NAT router, if not also a wireless router. Most off-the-shelf Linksys routers have QoS and such built in. How hard would it be to add instrumentation to these? It'd be a firmware patch at worst, I'd think.

Re:And I want their bandwidth...

[spazdor]

" Is metered bandwidth better for providers? For consumers?"

Yes, because it will align the ISP's profit motive with their "responsibilities" toward their customers.

It's common to refer to the alternative as "all-you-can-eat" bandwidth, and the comparison is apt. Consider the restauranteur who operates a buffet, and the careful balance of marketing and psychology he must strike in order to maximize his profits. To maximize income, he wants to get as many asses into the seats as possible - but to minimize his costs, he wants as few of those as possible to be really big eaters. He can't very well put a sign on the door saying "gluttons unwelcome", so instead he'll probably try to create an environment where unlimited eating is "allowed" but is discouraged in every way he can manage without giving the customer cause to claim they're not getting what they were promised for the money. So they'll manage the thermostat and the humidistat to make people less hungry, they'll make sure the buffet layout takes you past the cheap, plentiful items first to take up plate-space (maybe even keeping these items fresher), they'll spend extra effort to make sure everyone's water glass is full, and so forth.

In both games, the objective is the same: To give prospective customers the idea that they can take all they want, and then do whatever you can get away with to prevent that from happening.
If the restauranteur was instead charging a flat per-gram rate, uniformly marked-up from the ingredient and cooking costs, they'd behave differently.

Re:And I want their bandwidth...

[sourcerror]

Just for the free-market sake:
In the EU ISPs must give their average bandwith in advertisements.

Masters of the bleeding obvious

[jrq]

in other news: pope wears hat bear shits in wood etc etc

Re:Masters of the bleeding obvious

[mysidia]

Are you sure it's not bear wears hat, pope shits in wood?

That would be more interesting....

New toilet paper brand: up to 1000 sheets (actually contains 500 sheets)

New printer ink cartridge brand: up to 10000 pages (actually only enough ink for 10 pages)

New type of fast food hamburger.... up to 3 beef patties (actually doesn't contain any meat)

Re:Masters of the bleeding obvious

[T Murphy]

The story here is the government recognizes the misleading marketing as a problem, which implies there may be changes in the near future- unlike those popes and bears which aren't going anywhere.

Re:Masters of the bleeding obvious

[EricX2]

In even more unrelated news: slashdot subscriber posts sarcastic comment, news conference at 11 AM followed by round table discussion with the inventor of sarcasm on the use of such comments on a serious news site.

Re:Masters of the bleeding obvious

[HereIAmJH]

New type of fast food hamburger.... up to 3 beef patties (actually doesn't contain any meat)

But a 1/2 lb. burger is 'up to' 1/2 lb. They always advertise the pre-cooked weight of the patties. Actual cooked weight will depend on the quality of the beef used.

Hah

[copponex]

As if companies had incentives to lie. It's a good thing they don't, or we'd need some sort of third party to make sure they didn't rip anybody off. Where the hell would we get one of those?

Re:Hah

[Culture20]

As if companies had incentives to lie. It's a good thing they don't, or we'd need some sort of third party to make sure they didn't rip anybody off. Where the hell would we get one of those?

Well, we have two parties, and I hear China has only one party, so they could be the third party.

Docsis

[dhanson865]

Tell me about it. I was intermittently losing connectivity with a docsis 1.1 modem that was giving me 4+ mbit most of the time. The tech I called recommended I get a Docsis 3 modem so I did. I no longer have connectivity issues so it was a good call by the tech but I'm still seeing 2.x to 4.x mbit downloads and getting 0.3 Mbit uploads at best.

I called back after getting the new modem to get it provisioned, then called the next day after running speedtests. They said I should expect closer to 7mbit down instead of the 2 to 4 I'm getting but DOCSIS 3 would hit my area in the next few weeks taking the advertised to 12 Mbit down. So if I'm getting half the advertised speed I'll still see my download speed double if all they do is bond 2 channels for me.

I'm running at 100 Percent

[amiga3D]

I have a small local ISP here. Comsouth.net they consistently run at 100 percent of advertised speed. I'm amazed sometimes how fast it is. No lag, no drop in speed after the kiddies get home from school. I don't know what's wrong with them.

Re:I'm running at 100 Percent

[phantomcircuit]

Clearly they need more MBAs.

Re:I'm running at 100 Percent

[Ironhandx]

I'm beginning to subscribe more and more to a friends theory that all that is wrong in first world nations can be blamed on an MBA.

Re:I'm running at 100 Percent

[mark-t]

Are you running on dialup, perchance?

Re:I'm running at 100 Percent

[jon3k]

I'm a Cox customer and I get max advertised speed every single day without exception and never have a single problem. Maybe I'm just lucky I don't know but I get well over 20mb/s all the time and we're all DOCSIS 2.0. I'm only supposed to get over 15mb/s with "speedboost" but it works far more often than I _ever_ expected.

Re:I'm running at 100 Percent

[amiga3D]

No dialup. I'm on Cable at 10Mb. I've been ridiculously pleased with it since I got it. It's been 3 years or more since I had to call about a problem with connection. I remember the dark days of dialup all too well. I have the 2400 baud modem, the 9600, the 14.4 (a USR courier) and 28.8 and 56K. Man the money I spent.

Re:I'm running at 100 Percent

[aquabat]

So, are you saying you like Cox?

Re:I'm running at 100 Percent

[jon3k]

Love it, can't get enough. I'd like to get two or three Cox at once if I could.

Re:I'm running at 100 Percent

[vidnet]

I had a national ISP (NextGenTel), though in a small country, where paying for a 4Mbps line meant that the browser's reported speed was 500kb/s. Not only was the throughput consistent, it meant that all protocol overhead was included in the estimate. I even started paying for a 1Mbps line that was upgraded at no cost to 4Mbps as competition sharpened.

My only feasible theory is that the installation guy had a crush on me. You'll always get shafted, one way or the other.

Re:I'm running at 100 Percent

[aquabat]

I guess if Cox were bigger, then one would be enough.

Re:I'm running at 100 Percent

[Richard_at_work]

I find it amusing that you list a load of dialup standards, and yet they had the same issues as this article highlights - you could never get the quoted line speed, especially with 56K, as it was all reliant on the line quality.

Re:I'm running at 100 Percent

[snowgirl]

I'm beginning to subscribe more and more to a friends theory that all that is wrong in first world nations can be blamed on an MBA.

I'm down with this theory. Tell me more, can I pay you to hear more? Where is your seminars, and when will it be in my town?

Re:I'm running at 100 Percent

[amiga3D]

The only dialup modem I had that got it's listed speed was the 14.4 USR courier. It was the best dialup modem I had. The 56K modem generally downloaded at about 44K. I remember getting 49K a few times but that was rare. I never saw 56K download speeds.

Technically..

[Anonymous Coward]

'up to 6.7 Mbps' while the real median was 3 Mbps and the mean was 4 Mbps.

Prefixed with "up to" this technically isn't false advertising, or at least you could rationalize that. 'Up to' doesn't necessarily mean you will actually get that number. Don't get me wrong, this still sucks.

Re:Technically..

[jc42]

I learned long ago that whenever an advertiser says "up to", you should always translate it as "less than" in your mind. That's what they're really saying; they're just saying it in a way that's misleading but legal.

Re:Technically..

[kevinadi]

In general, dividing the "up to" number by 3 is actually closer to reality. That way, you're happy even if you get 50% less than what they said.

Re:Technically..

[T Murphy]

I learned long ago whenever an advertiser says "up to" to just ignore that number. Of course in the case of ISPs there is no competition to run to when they don't give you a straight answer, which is what makes this such an issue.

Re:Technically..

[arth1]

No, "up to" does not translate to "less than", but "no more than". There's a subtle difference:
Getting 10 Mbps satisfies "no more than 10 Mbps", even if it's not "less than".

But if you get 12 Mbps, you have a reason to complain. (And, yes, it can be a valid concern. You may have infrastructure devices that can't handle more. Deep-packet inspecting firewalls, for example, are known to cause network outages if subjected to more data than their rated throughput. And a certain e-mail-substitute server tends to crash pretty hard if data come in too fast.)

Re:Technically..

[TractorBarry]

I'd go further than that as I've learnt that every single thing said in an advertisement is most probably a lie.

It's more remarkable when something they say is actually true !

Re:Technically..

[clone53421]

Yeah, same here, but nothing pisses me off more than when they guarantee the number too...

Lose up to 5 pounds per week – guaranteed!

I'd always assumed as much

[AaxelB]

I always taken "up to" X Mbps to mean you might get bursts up to X, but would hopefully average X/2 or so, and I'm a bit of an optimist. They've always been very careful to specify that you'll definitely get less than X, why is it surprising that you do?

Re:I'd always assumed as much

[Golden_Rider]

I am just surprised that they can get away with advertising a connection as "up to" when the majority of their users NEVER see a speed anywhere near that and the ISP knows that. Car analogy: advertising a car as "max speed up to 250 miles per hour" when it can, due to engine etc., reach a max of 120 miles/hour ("well, our car CAN reach 250 miles/hour, if you strap a rocket to the back of it, but during normal use, you may not reach that speed").

Re:I'd always assumed as much

[Anaerin]

Strapping a rocket to the back might be a bit much. This is more like:

Up To 250MPH*

* In a vacuum

Re:I'd always assumed as much

[M8e]

No its really like:

"Top speed: 210 km/h*

*@Curb weight** with original tires on paved road with 0 incline, 0m/s wind, 101.325 kPa air pressure, 20c air temperature, XX% air humidity, 30c road surface temperature, engine coolant temperature of Xc, oil temperature of Xc, E85 gasoline with 105 octane(AKI)...

**European Directive 95/48/EC"

You don't say!

[fudoniten]

This is unbelievable! Next you're going to tell me that "3.9G wireless" doesn't mean anything, or that 9 out of 10 doctors don't recommend Crest, or that most items in an "up to 90% off!" sale are not in fact 90% off!

Sounds pretty paranoid to me. If we can't trust company advertisements for unbiased information, what can we trust?

So just what am I paying for?

[SanityInAnarchy]

And how do we compare plans? If one ISP has "up to" 10 mbits, and another has "up to" 20 mbits, which one is faster?

Not lying, but not in any way honest.

Re:So just what am I paying for?

[infra-red]

I would suggest understanding the base technology of the services you are considering.

Cable for example is a shared platform. You may share 40meg of capacity with 100 other people and be individually limited to 5meg. Depending on the users, you may or may not not ever have the network congested so you will receive less then peak value, but that doesn't mean it won't happen.

You could be unlucky enough to end up on a node with dozens of heavy torrent users.

The up-to is probably generally acceptable because 95% of consumers don't understand the underlying technology and probably are not as concerned about if they get 5meg all the time, or if sometimes they only get 3meg.

The speed they are giving probably refers to the physical cap on your access path, and the up to is to cover them off when the load simply get overwhelmed. If they are good company, they either have have tons of capacity, or they fire customers who use excessive quantities of capacity, most likely a mix of both.

Re:So just what am I paying for?

[Anonymous Coward]

The answer: FIOS is faster.

Re:So just what am I paying for?

[jrumney]

Sometimes the "lying" goes both ways. My connection was advertised as up to 2Mbps. During peak times I'm lucky to get 200k, but offpeak I've seen it running at 9Mbps.

Re:So just what am I paying for?

[Anonymous Coward]

Sometimes the "lying" goes both ways. My connection was advertised as up to 2Mbps. During peak times I'm lucky to get 200k, but offpeak I've seen it running at 9Mbps.

Sorry you have to hear it from me, but that is just porn that is already in your cache. It wasn't really downloading at that rate....

Re:So just what am I paying for?

[mindwhip]

In the UK at least...

If they are both ADSL the chances are they are both going to be the same speed as they will both be using the same copper cable and the ISPs will all have very similar upstream connections...

If one is cable and one is ADSL the cable will almost always be faster (unless you are very unlucky), and will always give the specified speed during off peak hours, although very few sites will send data at 20Mbs, a popular torrent (of say a linux distribution) with good seeds will easily hit your speed cap.

I don't know how BT's rollout of fibre-to-cabinet will work out but I suspect it will be very similar to cable speeds...

Re:So just what am I paying for?

[Sycraft-fu]

Well you can't really compare ISPs based on a single number no matter what. I mean even if both were willing to offer you a CIR (which you need to pay more for) there are still different in terms of latency, peering, uptime, etc.

I see nothing wrong with advertising maximum rates, as that is a consideration. Yes it is putting their best foot forward but if you aren't used to that in advertising you are being willfully blind.

Re:So just what am I paying for?

[Totenglocke]

Then you do some research and you find out what that ISP's average speeds are in your area.

Re:So just what am I paying for?

[martin-boundary]

And how do we compare plans?

Make a website where people can run a bandwidth test, and ask each visitor for their zip code and ISP and monthly payment. Then add this data to your database and print the visitor's nearest neighbour for comparison.

Bonus points if several people set up such websites at various locations throughout the country, and combine their measurement databases.

Re:So just what am I paying for?

[deek]

Sure, they know how far away you are, but they're not advertising only for you. They're advertising for all potential customers. There exist customers that can reach that top speed.

I do agree partially though. They're not telling you what the average connection speed is, and how likely it is that you'll fall below the max. I'd like the advertisements to at least mention mean levels and standard derivations. In small writing is sufficient.

Re:So just what am I paying for?

[aquabat]

Sometimes the "lying" goes both ways. My connection was advertised as up to 2Mbps. During peak times I'm lucky to get 200k, but offpeak I've seen it running at 9Mbps.

Fuck that! When I ordered my internet service, they promised me "Up to 5Mbps". I'd better not fucking get any more than that, or there's going to be hell to pay! I want what I was promised, and if they break their promise, I'll sue their asses!

Re:So just what am I paying for?

[PitaBred]

Right up until they burn your house down

Re:So just what am I paying for?

[Cimexus]

For xDSL technologies, it depends on your phone line (which the ISP can't predict obviously).

If your phone line was long enough/in poor enough condition that it could only support 7 Mbps, both the 'up to 10' and 'up to 20' plans would give you exactly the same speed.

If your line was good enough to attain, say, 16 Mbps, then the 'up to 10' would give you 10, and the 'up to 20' would give you 16. And so on.

For other technologies where the limiting factor is congestion from other users, rather than physical limitations of your line (e.g. cable or wireless), the higher one would generally be better you'd think (in the quiet periods, you'd get more out of the 'up to 20' plan). But you'd be wise to ask around on some forums or something to see how congested that particular ISP's network is and what other users experience.

For technologies where you aren't spectrum sharing (e.g. FTTN/FTTH fibre deployments), they generally won't be advertised as 'up to'. You should attain the maximum speed on fibre (to a local, fast host at least - obviously if things are slow beyond your ISP's network then no amount of fibre is going to help that).

Re:So just what am I paying for?

[valeo.de]

I'd be careful using words like 'always' when talking about cable speeds in the UK. Take a look at Virgin Media's customer support forums some time, you'll find an abundance of people that aren't getting anywhere near the advertised speeds...

Re:So just what am I paying for?

[Monkeedude1212]

Where do you live and who is your ISP?

I've never once gotten close to breaking my advertised up to speed - I'm always about 3 or 4 MBits below it.

Re:So just what am I paying for?

[clone53421]

Yeah... you should write a driver to divert any data that exceeds the 5Mbps threshold to the line printer so that you can return it to them by post.

Re:So just what am I paying for?

[SecurityGuy]

The problem is that when the market is built around "up to", the one provider who starts advertising their median speed will get almost no customers. They might get people like you and me who understand that a median speed of 5 mbps may well be better than "up to" 10 (you really can't compare them at all), but most of the market is not me or you. Most of the market will do "10 is bigger than 5, so I'll buy the 10."

I read an interesting article a couple years back about carpet cleaning. You've seen the ads. Get n rooms cleaned for only $X, where X is low enough that it's a steal. Then they get in and tell you that yeah, they can do that, but you REALLY want $OTHERSERVICE as well, or $ADDON, and $OTHERTHING. Basically, it's a rare customer who actually gets n rooms for $X.

Some ethical business owner tried to buck that trend and tell people up front that to really get what they wanted, they needed to get the carpet cleaned, plus $OTHERSERVICE, $ADDON, and $OTHERTHING for a total price of $aGoodBitMoreThanX. In the ideal world, consumers would realize everyone's offering the same service for the same price. They might even reward the ethical business owner for being up front about it. In the actual world we live in, it didn't work at all and he was forced to abandon the practice and go back to pricing like everyone else.

Re:So just what am I paying for?

[10101001 10101001]

I see nothing wrong with advertising maximum rates, as that is a consideration. Yes it is putting their best foot forward but if you aren't used to that in advertising you are being willfully blind.

Ie, we should accept fraud in a free market because..um..it really helps people who commit fraud? Really, fraud is deception to gain value. If one knows full-well that one's maximum rate has little or no bearing on median rate or even on any real probability of seeing that maximum rate at any time ever, then describing one's product with that value is intentionally deceptive precisely to defraud. Advertisement should be to inform. Think Wikipedia's non-bias standard: qualify all statements made and include as complete a picture as one reasonably can.

One should not have to second-guess the complete truthfulness of advertisements. Such is precisely the reason why the US economy is so fucked up right now (literally, the recession is a byproduct of understating the risk of subprime mortgages during securitization and the derivatives market misrepresenting their capacity to payout on failed contracts).

Re:So just what am I paying for?

[gfreeman]

Then you do some research and you find out what that ISP's average speeds are in your area.

Excellent, I'll just go to my computer and google for ... ... wait, what?

;p

Re:So just what am I paying for?

[DragonWriter]

For xDSL technologies, it depends on your phone line (which the ISP can't predict obviously).

Given as how usually the dominant xDSL provider in any areas is usually the same as the company whose wire telephone service runs on (because of telephone open access rules, they may not be the person you pay your phone bill too, but that's immaterial), I'm not sure that I would agree with the comment about their inability to predict anything about your phone line.

Re:So just what am I paying for?

[SanityInAnarchy]

latency, peering, uptime, etc.

All of which it'd be really nice to know. But where can I get that information?

if you aren't used to that in advertising you are being willfully blind.

Being used to it doesn't mean I should tolerate it.

But you see, there are other industries where the actual specs are available, so I can actually make a comparison. I could compare model numbers of CPUs, which would be fairly stupid -- but I could also compare clock speed, cache size, wattage, etc, all of which are trivial to find, generally right there on the page that's trying to sell it to me.

Re:So just what am I paying for?

[SanityInAnarchy]

And what do you mean by "research"?

If I just ask people, I'm going to get "Yeah, my Internet is fast" or "My internet is slow", and what they really mean is "I downloaded BonziBuddy and now my whole computer is slow, but I still blame Comcast."

The only real alternative is likely going to be actually subscribing to each in turn and running some tests myself, which would be obnoxious in one area. The quality of ISP is a serious consideration when I look for a place to live, so how, exactly, should I do that across the country?

Re:So just what am I paying for?

[SanityInAnarchy]

For xDSL technologies, it depends on your phone line (which the ISP can't predict obviously).

As DragonWriter points out, it may not be so "obvious" that the ISP can't predict it. Even if that's true, however, is the issue with your lines, or with them overselling and under-delivering?

For technologies where you aren't spectrum sharing (e.g. FTTN/FTTH fibre deployments), they generally won't be advertised as 'up to'. You should attain the maximum speed on fibre (to a local, fast host at least - obviously if things are slow beyond your ISP's network then no amount of fibre is going to help that).

And, more importantly, do they have enough upstream bandwidth to handle their fiber users wanting that full speed to a remote, fast host? It is cool having essentially LAN speeds within my small town in Iowa, but that's not really the point of getting fiber.

Re:So just what am I paying for?

[Totenglocke]

You know, there ARE places (like speedtest) where it logs ISP and bandwidth in an area......

Re:So just what am I paying for?

[mindwhip]

Take a look at Virgin Media's customer support forums some time, you'll find an abundance of people that fail to understand and complain when they hit the clearly indicated peak hour usage caps, or don't understand the difference between meagbit (Mb) and kilobyte (kB) or are not in the 20Mbs cable service areas and are on the 'national' ADSL service and don't get that there is a difference.

Fixed that for you.

Re:So just what am I paying for?

[mindwhip]

Oops forgot... ...or have bot/trojan infested computers using the bandwidth or just plain inadequate computers or wonder why the illegal porn they are accessing from an underpowered server in someones basement in some distant, poorly connected part of the globe can't max out their connection or why the cheap wireless router they are using is suffering collision/interference problems or just can't cope with 2000 torrent connections (because more is faster right?!?) or they are saturating their upstream bandwidth with torrent seeding causing control packets to be dropped and causing multiple unneeded data retransmits or have crappy anitvirus software that slows everything down...

I could go on...

Median 3Mbit, Mean 4Mbit

[Rockoon]

FCC analysis shows that the median actual speed consumers experienced in the first half of 2009 was roughly 3 Mbps, while the average (mean) actual speed was approximately 4 Mbps

The real story is that over 50% of the users get less than 50% of the average bandwidth. I'm not sure how to explain it, but the difference between median and mean looks quite significant to me.

Re:Median 3Mbit, Mean 4Mbit

[hwk_br]

The real story is that over 50% of the users get less than 50% of the average bandwidth.

And 50% of the users like their ISPs 50% more than they deserve... Altho my ISP allocates 12797Kbps downstream for a 10Mbit plan. "Viva GVT"

Re:Median 3Mbit, Mean 4Mbit

[Your.Master]

I expect it's because during peak usage, you get less bandwidth. But also more users are using it during peak hours (by definition). So the bulk of people are using it primarily at peak hours and getting <= 3 Mbps, and the remainder get >> 4Mbps on average (possibly pretty close to 6.7 on off-peak hours).

Re:Median 3Mbit, Mean 4Mbit

[mysidia]

Reminiscent of the US tax system.... 10% of the people pay 90% of the taxes.

10% of broadband users get a transfer rate that is 90% (or more) of the rate advertised.

And ISPs like Verizon complain about 2% of the users transferring more than 2GB of data a month.

Now we know why

Only a small portion of the subscribers actually get the really high speeds that allow them to do everything they want transparently, which, of course uses more bandwidth, due to the ISP's failure to deliver true quality high speed links.

98% of users are sick of fighting their slow internet connections

Re:Median 3Mbit, Mean 4Mbit

[natehoy]

And 50% of the users like their ISPs 50% more than they deserve...

"I don't know half of you half as well as I should like; and I like less than half of you half as well as you deserve."

Re:Median 3Mbit, Mean 4Mbit

[clone53421]

Reminiscent of the US tax system.... 10% of the people pay 90% of the taxes.

Yep. Or, to phrase it similarly to GP’s statement: Over 50% of the people pay less than 50% of the taxes.

Nothing to see here. Moving on...

RCN in Chicago

[cspankne]

I have RCN (Cable and Internet) in Chicago. I have spoken candidly with technicians who come out to do installations and I have verified through several phone calls with customer representatives that they "aim" for 60% of advertised speeds. I perform speedtests, using their preferred site and have found that I am almost ALWAYS at 60% of advertised speeds. In order to get over 10 mbit/sec down, I have to pay for the "20mbit/sec" rate, and am typically around 12 mbit/sec down. If I was a normal customer, I'd easily compare the 20mbit/sec advertise rate against competition and opt for RCN's as it is the cheapest price for that advertised speed. Complete garbage and misleading to consumers. How is this legal?

Re:RCN in Chicago

[bjackson1]

It probably is due to the fact that we are on different nodes of the network, but I have RCN 20 megabit as well here in Chicago, and I get pretty constant 18mbps... Consistant enough that I think they are throttling me to 18mbps, since I've rarely seen under and never seen over.

Re:RCN in Chicago

[phantomfive]

If it's truly deceptive, then it is illegal. I don't claim to have a deep understanding of it, but as an example, a lawsuit was recently brought against the makers of Vitamin Water for being deceptive, because even the name implies it is healthy, when in fact it's mainly just sugar and water. Here's a great page describing better than I can, and it explains what cases the FTC chooses to pursue. Note that other people besides the FTC can also file lawsuits if they so choose.

It may be no one ever realized the broadband speeds were overestimated by so much, assuming the article here is true.

ISP Welcome Kits

[Korbeau]

Up to 7Mbps! Improved flavor, now without trans fat!

Re:ISP Welcome Kits

[DeadPixels]

Now asbestos-free!

Up to...

[TKBui]

I have fibre to my house (10Mbps symetrical) and I get close to that (per speakeasy.net/speedtest) 9.75Mbps/9.12Mbps. Should I sue? :-) Oh yeah... I pay $70/mth for 1.5/384 ADSL. TKBui

The story is..

[prakslash]

They set up false expectations. Unfortunately everyone does it. I have bitten by this more than a few times.

Weight loss ad told me I could lose UP TO 50 lbs. I still need to request a seat-belt extender on airplanes
My employer said I could make UP TO a million dollars a year if the company does well. I am still driving a beat up Kia
And, worse of all, that nice email ad said I could increase my length UP TO 9 inches. My wife still has trouble finding it

Meh..

Re:The story is..

[techno-vampire]

I still need to request a seat-belt extender on airplanes.

I don't know how much weight you actually lost, but just losing weight isn't always enough; you need to get some exercise to firm up your abs. Try sit-ups and crunches; you'd be surprised how much of a difference they can make, even if your weight doesn't change. And as an added benefit, if you get your stomach out of the way, your wife might just be able to see what you've got down below.

Re:The story is..

[gmhowell]

And, worse of all, that nice email ad said I could increase my length UP TO 9 inches. My wife still has trouble finding it

Meh..

Funny, she had no trouble finding mine last night.

Re:The story is..

[gandhi_2]

It seems you are talking about spot reduction, which has been repeatedly demonstrated and documented to be false.

Ab workouts don't burn more fat from the abdominal region than any other region or exercise.

Re:The story is..

[endymion.nz]

They tone the abdominal muscles, so that when you do eventually get under 7% body fat from the exercise that you do every day, you don't look like an emaciated swimsuit model.

Re:The story is..

[techno-vampire]

No. I'm talking about firming up the muscles so that you can hold your gut in better, as well as changing some of that flab into muscle. That lowers your waist size even if your weight stays constant.

Re:The story is..

[Nesman64]

Funny and insightful are sometimes divided by a fine line. Maybe internet service providers should start advertising with the disclaimer that weight loss companies use when they show a success story: Results not typical.

Re:The story is..

[clone53421]

Yeah. Non-story, if you ask me, but that’s just me. Maybe I’m jaded, but any time I see an advertisement that promises “UP TO!”, I assume that a typical case is going to be quite a bit less impressive than the number they quoted.

“Up to 50% off!”: The 2-for-a-dollar packs of store-brand batteries are now on sale 3-for-a-dollar. Everything else in the store is marked down somewhere between 0% and 5%.

“Up to 8 Mbps!”: At 2 AM, if nobody else in your region was using their service at the same time as you.

“Up to 15% energy savings!”: Or it could be 0%. (Worse yet, these ads often “GUARANTEE” it!)

Re:The story is..

[Jaime2]

They didn't set up false expectations, the reader inferred them. The beauty is that as long as people keep inferring meaning, marketers never actually have to say anything they can be held to. We decide to be lazy and infer meaning, so we get no actual commitments.

The solution is to become a hardcore skeptic. Assume everything not written in bulletproof legalese to be a blatant lie. Ask for everything in writing. Never trust a source with an incentive to be untruthful (this includes every sales person on the planet). You'll be right more than you are wrong.

Weight loss ad told me I could lose UP TO 50 lbs. I still need to request a seat-belt extender on airplanes
My employer said I could make UP TO a million dollars a year if the company does well. I am still driving a beat up Kia
And, worse of all, that nice email ad said I could increase my length UP TO 9 inches. My wife still has trouble finding it

BTW, the weight loss product had one person out of a million lose 50 pounds, and they lost it by getting into a car accident and having a leg amputated. Your employer is flat out lying to you, but it is mathematically possible. The email is lying so blatantly that if the FTC could trace the money, someone would go to jail. Only a gullible moron thinks otherwise.

What did they measure?

[White Flame]

I've got a piddly 2Mbit/sec cheap connection here in urban USA, and top out at around 200KB/sec download speed. However, some sites can't push data to fill even that little pipe. If they are measuring sustained speed of a single download, your 20Mbit/sec connection can theoretically go 2MB/sec but are the server connections you're downloading from capable of sustaining those upload speeds for common uses? What about traffic congestion further behind the point of speed throttle you're paying for?

Re:Loophole

[Spewns]

The loop hole here is "up to"... "up to" != "is"

No shit.

It's hard to believe so many people have nothing more to say than that. This is clearly dishonest, inaccurate advertisement, yet it could say "Up to 10000000000000000000 Mbps" and dozens of dorkos will still chyme in with "but it says up to! Hur hur hur hur!"

Re:Loophole

[phillymjs]

Yeah, I don't get the point of this article. The ISPs have the weasel words right in front of you, they're not hiding anything.

Now with that "unlimited" connection promise, on the other hand...

~Philly

Why, send them a nice letter...

[zogger]

...tell them you will pay them "up to" the agreed on price for their service, but you will determine what the real sums involved will be.

Re:Why, send them a nice letter...

[MojoStan]

Someone has to post the oblig. link: http://www.penny-arcade.com/comic/2009/5/01/

Re:Why, send them a nice letter...

[momfreeek]

http://www.penny-arcade.com/comic/2009/5/1/

Re:Why, send them a nice letter...

[sincewhen]

You're being funny, but I have been waiting for someone to seriously suggest this.

That is, why don't the ISPs test the line for maximum speed and pro-rata their charge appropriately?
So if you could get up to 10Mbps, but can actually only achieve 5Mbps at most, the should only charge you half.

Seems fair to me...

Re:Why, send them a nice letter...

[DarthVain]

Nice. If I could mod you I would!

"Say, I got this bill for 75.65$. However my average speed was only 54% of your maximum 10mb/s. Enclosed is a check for 40.85$. Have a nice day! :)"

Re:Why, send them a nice letter...

[DarthVain]

Now combine that with unused CAP space and you can get real savings! "I only used 10GB of my 60GB CAP further reducing my bill to 6.81$...."

Advertising..

[nephridium]

That's what advertising is! You don't compare products by their advertising, but by unbiased reviews, or by trying it out yourself (if there are short term subscriptions). A certain brand of beer won't get you automatically surrounded by hot chicks just as a certain brand of cigarettes won't turn you in to a cool cowboy sitting by a camp fire.

Now if there was a standardised benchmark to test broadband speed.. - But for that you'd probably need government involvement, and who wants that, right?

Re:Advertising..

[mindwhip]

http://speedtest.net/global.php#0 is useful... but you need to drill down to your local area and ignore datacentre ISPs...

Re:Advertising..

[SanityInAnarchy]

You don't compare products by their advertising, but by unbiased reviews,

With better-behaved industries, like (surprisingly) computer hardware, I can also compare them by specs. That's my point -- if "up to" is meaningless, is there a spec anywhere that I can compare?

Without regulated standards, who cares?

[Dahamma]

They claim "up to X Mbps". As long as some customer out there gets X Mbps, they are NOT lying. They may be completely gaming the system, but most companies in their situation would do the same thing given the cable vs DSL competition right now....

This is why there needs to be FCC regulated standards for stated services levels/Internet bandwidth based on real statistical measurements. Most cable and DSL modems out there are capable of bandwidth testing. Sample enough of them, take the median, mean, standard deviation, whatever, and allow them to state certain claims based on the results, as long as they are clear. This level of data is already required for food, cars, (some) utilities, etc, why not Internet access?

Re:Without regulated standards, who cares?

[mysidia]

No.... they are in this business to make money.

They would probably force customers who get low speed to cancel service, so they could have higher numbers. By making their service break until they switched to a competitor or informed their line can't actually support the service, and they need to cancel.

Assuming there's some intrinsic reason certain customers got lower speeds, and it would be costly or impossible to address.

Re:Without regulated standards, who cares?

[Dahamma]

They would probably force customers who get low speed to cancel service, so they could have higher numberss. By making their service break until they switched to a competitor or informed their line can't actually support the service, and they need to cancel.

That's why I said it needs to be treated like a utility by the FCC. The phone and cable companies are required by regulation or franchise contract to offer service for everyone for certain services (POTS, TV), and keep up a service level agreement or be fined, etc.

Assuming there's some intrinsic reason certain customers got lower speeds, and it would be costly or impossible to address.

This is NOT a new concept. Phone and cable companies have already been required for MANY years to fix your phone line or cable connection (assuming it is their fault and not your in-home wiring) regardless of the cost - and they do. Not always in a timely or efficient way, but they do.

Re:Without regulated standards, who cares?

[mysidia]

Phone and cable companies have already been required for MANY years to fix your phone line or cable connection

It may be that nothing is broken. Your distance from the CO is simply so large that you intrinsically cannot achieve even half the advertised speed.

It would be in DSL providers' interests to make the "distant cutoff" shorter, to increase their average speeds closer to the max, if they had to be judged based on that.

A minority of subscribers are a long distance from the CO. Beyond a certain distance, the DSL provider won't even try.

If they have an incentive to make that cutoff shorter, they will.

Mean, median, mode, lame

[Gothmolly]

The mean is not the maximum. Remember grade school math?

its a story because

[Rivalz]

Not everyone realizes that other people are getting substantially better internet for the same amount from the same company based on the same agreement.
I think its high time corporate America fully embraces the "Up to" mentality.
Here are some suggestions
1) Restaurants / Groceries (Up to meals) Only give half the people half the portions of food.
2) Gas Stations (Up to 1 gallon for $2.80) Some days we dont have to give any gas but if you go 24hrs without getting any gas we will give you a minor refund of what you paid.
3) Cell phone minutes (up to 2100 family minutes during peak hours) But really only give 50% of the minutes to half the clients and charge them more for the rest.
4) Warranty (We warranty all our services up to 2 years ( meaning we can deny your service before or after 2 years, but after 2 we will always deny it.)
5) Intrest rates ( up to 2% fixed interest rate for the life of the CD ) Up to meaning we dont have to pay anything but at most we will pay 2%.

Can anyone think of any others?

Re:its a story because

[Nethead]

Up to 15 mod points but I got only 5.

Re:its a story because

[spencebot]

Power supplies for computers kind of work like this. Many of them advertise their peak power (a few advertise the rms), but obviously peak power is not nearly as meaningful. Wiki has some other good examples on their "power handling" page.

Re:its a story because

[mysidia]

Yeah

5)Interest rates -- Up to 90 days, same as cash (use your imagination)

4)Insurance Coverage -- Up to $100,000 coverage in case of an accident, Variable-rate deductible as low as $1000.

3)Cell phone data -- Up to 2GB of data per month included (some months, you only get 100K and have to pay 0.1 cents per Kilobyte (= $0.10/Kilobyte ) for more

2)Gas Stations -- fuel additive concentrations -- Up to 90% unleaded fuel (some days, the gas will be 100% additives, but you'll never find out anyways)

1)State lottery jackpots -- "Jackpot: up to $1,000,000" -- (Some jackpot winners who got the numbers right may found they have only one $1.)

Re:its a story because

[Quiet_Desperation]

Can anyone think of any others?

I will do Slashdot posts of up to ten

Big Bad ISP Good

[Anonymous Coward]

Well, I guess I'm the lucky one. I pay extra for bandwidth to my home, and the plan calls for 15 Mbps.

I've measured that puppy several times right at 30 Mbps. Most times, it hovers around 20.

I'm lovin' this.

Most surprising of all, it's a well known major ISP.

They still couldn't run a DNS server to save their lives (Thank you, OpenDNS), but on the bandwidth front I have no complaints. (Other than cost, of course :) ). That, and my upload appears to be 640 Kbps instead of 768... Oh well.

Only 98% lies.

[AliasMarlowe]

If your "up to" only applies to 5% of your customers, you're scamming them.

Try 2% getting close to the advertised speed, and 98% getting a lot less. According to OFCOM - the UK regulator - only a small fraction get anything close to the advertised maximum (unless they're on fiber). For example, regarding ADSL2 which was 'up to 20Mbps':
"65% were getting less than 8Mbps, 32% between 8 and 14Mbps, with just 2% getting 14-20Mbps."
http://www.bbc.co.uk/blogs/thereporters/rorycellanjones/2010/07/ofcom_broadbands_broken_promis.html

Re:Only 98% lies.

[green1]

I'm curious... do these speed issues pertain only to certain countries? or certain technologies?
The reason I ask, is that my ADSL connection is advertised as "up to 15 Mbps"... the speed I get? 15 Mbps. now that doesn't mean 15 Mbps from any site, (you loose a fair amount just by virtue of the fact that you aren't always the only one going to a specific site, and there are often busy routers in between) but it DOES mean that I have 15 Mbps available, this is usually easy to see by downloading several things from different places and seeing a total of 15 Mbps.

Further though, I'm not a rarity here, I am an ADSL tech, I go to people's houses when they complain that they aren't getting the full speed they were expecting. What I've found is that there are only a few reasons why they wouldn't get the maximum:
- They don't understand that they can't download at 15Mbps from a server hosted on a private ADSL line somewhere else in the world that only has a 600kbps upstream rate!
- Their computer is so laden down with garbage that IE takes 25 mins to even load, let alone transfer any data
- Their P2P program in the background is using 14.999999999999Mbps of their 15Mbps total leaving them with nothing.
- And lastly, and extremely rare, is someone stuck in an area that just can't do more than 1Mbps (long lines back to the DSLAM), This is quite rare at this point, but more than that, it means that they won't be paying for a 15Mbps connection because we won't sell them one if they don't qualify. they'll be paying for a much lower speed (and if they somehow DO get on to one of the higher plans, they get changed to a slower plan as soon as we realize we can't provide that speed)

Now that said, our largest competitor, a cable company, used to quote speeds that weren't even physically possible with the equipment they were offering... but even they are quoting much lower speeds than they used to (though I don't know if people actually get those speeds on the shared cable topology)

Re:Only 98% lies.

[HereIAmJH]

I'm curious... do these speed issues pertain only to certain countries? or certain technologies?

...

Now that said, our largest competitor, a cable company, used to quote speeds that weren't even physically possible with the equipment they were offering...

I think you answered your own question. Advertised speeds seem to be accurate up to the point that the provider starts oversubscribing their infrastructure. Certain countries might have regulations concerning how much a service might be oversubscribed. But all bandwidth technologies have choke points, and the ones that are closest to the customer are the ones least likely to meet their advertised rates. Cable and wireless (3g) are using a shared medium all the way to the customer. DSL isn't shared until the customer gets to the DSLAM, and even there you have the experience/culture of engineers who are used to working in a much more regulated environment.

Another example of cultural disconnect. Time Warner used to continually try to get me to add home phone service to my cable/internet service. Yet it never occurred to them to install backup power capabilities to their network or encourage customers to install UPSs for their cable modems. During a storm related power outage my systems would stay online due to my precautions, but the cable network would be out, thus knocking out essential phone service if I used theirs. The Telco on the other hand has been regulated to require the local loop to be available unless physically damaged.

So I don't think it's unusual that you see ADSL performing at the stated rate. Your infrastructure has been designed conservatively. And Telcos currently have an abundance of bandwidth in the local loop due to the build-up during the 90's for fax machines and modems. Now people are shutting off those lines and going to alternatives. Cell phones for kids and alarm systems, e-mail instead of faxes, DSL and cable internet for computer connections.

As bandwidth demand increases (on-demand TV, streaming media, etc) and regulations are relaxed on Telcos so they can 'compete', you'll see them oversubscribing their links between the DSLAM and the CO simply because it's horrendously expensive to run new lines. And if other providers are allowed to continue they are at a financial disadvantage if they don't follow the industry.

Re:Only 98% lies.

[Jedi Alec]

I'm curious... do these speed issues pertain only to certain countries? or certain technologies?

Here in the Netherlands to my knowledge no one advertises using "up to" doublespeak the way it's done in the US and UK. Then again we have well over a dozen ISP's to pick from on the phoneline front and there's more breaking into the cable market as we speak, so any corp trying to pull a stunt like that would start losing customers rather quickly.

Speaking for myself, my ISP says I get 50/5 and so far I have yet to see it be any less than that, assuming whatever is on the other side can actually spit out the bits at that speed.

Re:Only 98% lies.

[Seth024]

I get up to 675KB/s download speed on my 6Mb/s connection (that's the maximum; when I'm downloading big files from a good server)

675 * 1.1 (ATM overhead) = 742KB/s = 5.94 Mb/s

Unfortunately when I upgraded to a 12Mb/s connection I remained capped at the previous speed because "I'm too far away from the central".

Re:Only 98% lies.

[delinear]

How about a sliding scale of payment - if I'm using half the theoretical bandwidth of someone living next to the exchange and paying the same amount that's patently not a fair system. Alternatively if I only get 10mbps on broadband, when I decide to pay for a 10mbps connection instead why does my speed then drop to 5mbps - my distance to the exchange hasn't changed so that must be a technical issue on their side (even if it's oversubscribing each tier of service). It's not like people are even getting slightly less than they're paying for, we're talking about more than half of all customers getting less than half of the advertised speed - that's a massive disparity.

Re:Only 98% lies.

[jcrousedotcom]

Another example of cultural disconnect. Time Warner used to continually try to get me to add home phone service to my cable/internet service. Yet it never occurred to them to install backup power capabilities to their network or encourage customers to install UPSs for their cable modems. During a storm related power outage my systems would stay online due to my precautions, but the cable network would be out, thus knocking out essential phone service if I used theirs. The Telco on the other hand has been regulated to require the local loop to be available unless physically damaged.

While this is not an edorsement of cable VOIP over POTS I did want to point out that my cable providors EMTA actually has a battery backup built into it and lasts for quite some time when the power goes off. I assume (probably incorrectly?) that they have backup power at their 'Central Office' end of things. Rarely does the power go off here for more than an hour or so and if it does it is likely the result of a hurricane or some other natural disaster and physical damage to the infrastructure is likely. To tell the truth, I only picked up the land line service because it the package saved me money by bundling. FWIW its nice to have a land line but I rarely use it except for my alarm system. My cell phone for personal and my Packet8 VOIP phone for my company gets most of my talk time.

Re:Only 98% lies.

[green1]

I've been out to troubleshoot "it's fine until 5pm" many times... it has NEVER been a problem with the DSLAM being oversubscribed.
I've had it be a few things:
- failing copper lines (heating and cooling cycles thorughout the day can cause it to fail, or have errors, at specific times of day) usually easilly fixed. (once they decide to call in instead of just cancelling!)
- Most commonly, it's fine until 5pm really means it's fine until their own kid gets home from school and turns on their favourite P2P program. Remember, you aren't sharing a line with your neighbours, but you ARE sharing one with your kids!

Re:Only 98% lies.

[HereIAmJH]

While this is not an edorsement of cable VOIP over POTS I did want to point out that my cable providors EMTA actually has a battery backup built into it and lasts for quite some time when the power goes off. I assume (probably incorrectly?) that they have backup power at their 'Central Office' end of things.

I don't know where Time Warner's system fails. I just know that every time my power has gone out my equipment has stayed up but the cable network is down. I have a 1000va ups that powers just their modem, my router, and my switch. It will run for hours without power.

OTOH, my mom lost her power for a week recently due to a storm. Landline worked fine the whole time.

The incentives are all wrong

[Ichijo]

With unlimited plans, the ISP's incentive is to prevent you from using up all your bandwidth, because infrastructure costs money, so if you used up all your neighborhood's bandwidth, they'd have to upgrade their network.

With a per-megabyte plan, the company's incentive is to provide you with more bandwidth than you could ever possibly need so that nothing will prevent you from downloading as much as possible.

If we want fast pipes, we should be asking for pay as you go data plans.

Re:The incentives are all wrong

[T Murphy]

If they switch to usage-based billing, they'll just target the numbers so they get the same or more money if customers don't change their habits, and they'll have less incentive to upgrade their system as customers won't be using that extra bandwidth once they learn how much it costs them.

I think the best model for fair use of the neighborhood pipe is to have a moving window (say 72 hours), and your speed is a base (say 128 kpbs) plus a fraction of the available bandwidth determined by your usage compared to your neighbors who are currently using the pipe. That way you only get penalized for being a "bandwidth hog" at times when others want to use that bandwidth, but you can still check the weather, email, etc no matter what without a problem.

Sure, the above system does nothing to encourage ISPs to upgrade their systems, but I'm convinced the only thing that will is true competition or strong government regulation, both of which are doubtful to happen soon. Therefore I'll settle for a fair system to manage what bandwidth we can get.

Re:The incentives are all wrong

[mysidia]

Pay as you go data plans are unfair. The more data you transfer, the more you pay on average for the same network resources.

We should be asking for 95th-percentile network utilization-based billing

With price breaks for off-peak hours and a premium for prime time utilization.

The guy who downloads a few GB of Linux ISOs from 2am to 6am, or over several days at a trickle... (i.e. rate-limited on their end)

Should not pay as much as the guy who maxes out his connection at prime time when the network is at peak utilization.

The ISP should also allow you to apply QoS markers to your packets and honor them to reduce priority of traffic, and low-priority packets should be less expensive, since you were essentially marking your traffic as specifically drop-eligible in case of any congestion.

Re:The incentives are all wrong

[phantomfive]

Wow, that is the most insightful thing I've seen in a conversation about ISPs in a long time. Excellent point.

Re:The incentives are all wrong

[bill_mcgonigle]

If we want fast pipes, we should be asking for pay as you go data plans

Yes, but we need a competitive market for that to play out. So long as municipalities are granting local monopolies (and not owning them) both the consumer and producer incentives are all wrong.

The current détente is the telcos charging about as much as they figure an average consumer is willing to pay, and then providing just enough service at that price to keep the regulators off their backs (which is some function of the consumers being on the regulators backs and the backroom deals).

We can do better.

Re:The incentives are all wrong

[the_one(2)]

http://en.wikipedia.org/wiki/Token_bucket

Re:The incentives are all wrong

[DocSavage64109]

I really like that idea, but who would enforce such a pricing structure, as all the ISPs care about is charging as much as possible. Maybe if the government regulated rates much as they do with water, gas and electricity? I wouldn't mind a monthly fee that was itemized and made up of connection fee, maintenance fee, and actual bits transferred (possibly billed on a QoS/primetime scale) as long as all of those fees were fair and accurate.

Re:The incentives are all wrong

[TheSync]

If we want fast pipes, we should be asking for pay as you go data plans

I'm really not sure this is the issue. US broadband speeds are most limited by our longer local loops compared with other developed countries.

The networking bandwidth costs are a very minor issue compared to the capital costs of rebuilding shorter copper loops (AT&T uVerse FTTN) or putting in new fiber ones (Verizon FiOS FTTH).

So indeed, the "speed tier" plan makes sense, but only if the money you pay to get a faster tier actually goes to building you a better local loop.

Re:The incentives are all wrong

[evilviper]

With a per-megabyte plan, the company's incentive is to provide you with more bandwidth than you could ever possibly need so that nothing will prevent you from downloading as much as possible.

A) Metering all your users is expensive.
B) Complex billing for all your customers is expensive.
C) Maintaining customers who download very little and pay you next to nothing, is expensive.
D) Charging per-MB will increase prices and likely reduce demand, so while they might be encouraged to upgrade the last mile (or just take off the speed-limits they currently impose), they will probably actually reduce their upstream bandwidth.

Re:The incentives are all wrong

[mysidia]

The ISPs would have opportunities to charge more, and not have to buy as much capacity.

Think about it... if as an ISP, your users have to pay more for high utilization, but not more for large amounts of data transfer spread over time, then it actually discourages behaviors that could cause congestion on the network, and encourages behaviors that relieve congestion.

It gives customers a financial incentive to limit their transfer rates.

Which increases the amount of oversubscription you could achieve, without service being negatively effected.

Re:median and the mean

[gmhowell]

I'm more interested in the mode.

Up to is what I get

[the eric conspiracy]

I wonder how many people really care? For email and web browsing does it really matter? Would the 50% difference even be noticeable? How many people even measure their speeds?

Now for a heavier user like me and many other slashdotters who do a variety of things on the net the difference is significant, and I hold my ISP accountable when I get more than 10% less than the up to number. Since I have Optimum Online Boost that means 30mbs down and 5 up. Which I get unless there is some serious interruption due to act of God. And if I didn't get it without good reason I'd be on to another ISP, which in this case would be FIOS, which has a good reputation for actually delivering the up to number.

The current state of affairs is simply due to the fact that 50% of advertised is probably good enough for most people so they don't raise a ruckus.

Re:Up to is what I get

[IRoll11!s]

I don't use Hulu, but I don't think it would matter much on Youtube. I've got a 10/2 from Verizon FIOS, and as noted by a few previous comments they are one of the few that actually do a good job of providing the maximum throughput at any given time. I swear a few times I've actually gone over 10Mbps.. Anyway Youtube seems to have very aggressive... hmm no, "optimistic" streaming settings on the server side, and will grudgingly dole out just enough chunks to keep it a few seconds ahead of the playback whenever possible. It is rare I use even a 10th of throughput on any given video, and I've taken to pausing anything I want to watch for 30 seconds or so, go do something else, then come back and hit play to avoid any potential interruptions.

Re:Up to is what I get

[clone53421]

So... if I say that “I don’t even watch YouTube”, have I become the 2010 version of “that guy”?

Tell me what the CIR is...

[rotide]

I don't care what your maximum is as I'm merely going to consider that the maximum burst speed under optimal occasions. What I want to know is, what is the CIR? Ya, I know, it doesn't exist in residential service, well, that sucks.

Re:Tell me what the CIR is...

[rdebath]

Actually, in some places, they do. Except they call it "contention ratio". If you have an "up to" 20Mb at a contention ratio of 50:1 that's 400kb CIR. Of course, they will often bury it too.

What they can't tell you with ADSL is your PIR. For example, the advertised is 20Mb but the reality is whatever your modem syncs to. (14Mb for me)

For BT wholesale connections (in the UK) the rations are 50:1 for a residential service and 20:1 for a 'business' version. But that only covers the BT part; from your phone socket to the ISP's lan not their interconnects.

Re:Loophole

[mark-t]

It's not dishonest because their "up to" speed is still *actually* achievable with their infrastructure under absolutely ideal conditions.... the fact that such ideal conditions generally don't ever happen is irrelevant.

Misunderstanding the meaning of the word "lie"

[Anonymous Coward]

It seems people like to stretch the meaning of the word "lie" to mean "inconvenient to me." You see, as long as someone can achieve the advertised maximum speed, it is not a lie. If it really was a lie, then that would illegal, and opening the companies for not only civil lawsuit, but criminal charges.

Just because most users don't get the maximum advertised speed, and the average/median/mean speed is not the maximum speed doesn't mean the company is falsely advertising their "up to" speeds. There are various conditions that can affect the speed of the connection and this is usually outlined in the advertisement and/or the agreement.

The only thing the ISPs may be lying about is possibly the implication of capacity to support those connections. However, because they don't advertise "we have the capacity for everyone to use 10mbps" you can't really catch them in a lie.

Re:Loophole

[Sancho]

No, "up to" means that the connection, including provisioning, is capable of transmitting data at that rate. If it's impossible to transmit data up to that rate, then that would be false advertising.

Re:median and the mean

[Haeleth]

Come on, did you fail maths or something? If the median is 3 and the mean is 4, then the average is 3.5.

This is News?

[Ironhandx]

In other news, water is wet.

Seriously, is there anyone on slashdot that wasn't well aware of this? I think its even safe to say 99% of people GLOABALLY don't get speeds as advertised. In fact I had a connection that I know for a fact could never, ever possibly hit the advertised speed as advertised was 5mbps and the modem was throttled to 4 mbps max in the firmware. Not that I even had to worry about that as 3mbps at 2 am on a good day was like greased lightning compared to normal rates. It was the "Premium" service however. My packets would get prioritized over others, basically guaranteeing me a 10-20 ms ping drop from being prioritized locally and I was playing a lot of MMO's at the time, so a 1mbps connection with a lower ping was worthwhile.

Re:No problem with Cox

[MichaelSmith]

Yeah cable can be great for burst speeds. When I used Optus cable I once downloaded a fairly big file (say 30MG) with wget. The command actually completed before I noticed it had finished. I thought it had failed and looked for an empty file, but it was all there. Must have come down the line in one chunk.

Did also believe the governments MPG ratings?

[Anonymous Coward]

Did also believe the governments MPG ratings?

Re:Qwest

[Phroggy]

is that 160 kilobits per second, or 160 kilobytes per second?

You might need to multiply or divide by 8. Units are important.

I wish

[silentace]

Too bad I can't pay "Up to" 100% of my cable/internet bill.

ok

[nomadic]

My first modem was 2400 bps. It was slow enough that I could read text as it came in. My next modem was 9600 bps. Door games ran a little faster, which was cool. My third modem was 14.4k. I was able to download Doom. It took me 6 hours. You all need to calm the hell down and get some perspective. Bunch of spoiled babies.

Why Some ISPs should be sued...

[Bazar]

Here in NZ most of the providers will advertise plans as "downloads as fast as your connection can handle", then an upload cap of 256Kb/s.

The result is that even if you connect to the exchange at 18Mb/s, you'll find using more then 3Mb/s impossible because of the tiny upload.
Each packet you receive by tcp requires confirmation packet to be sent. Each sent packet uses your upload bandwidth.

So by all practical measures, your download speed is limited by its upload speed. As fast as you can download is false advertising, as its as fast as your upload allows.

I'd love to see companies sued over false advertising, as it is misleading the public into thinking they are downloading as fast as possible, when they are only getting a fraction of their true potential.

Re:Loophole

[queazocotal]

In many cases it's completely impossible to accurately determine the speed that you will get before the equipment is actually installed.

So it's news to Ars Technica...

[John Hasler]

...that the mean and the median are both less than the maximum? Not surprising.

Re:So it's news to Ars Technica...

[Dunbal]

...that the mean and the median are both less than the maximum? Not surprising.

Not necessarily: What is the mean and medium of the following set of numbers?

5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5

ahem

[thatskinnyguy]

"Ars Technica has an article detailing the difference between ISP advertised 'up to x Mbps' speeds and the actual speeds, in addition to some possible solutions. They find that on average, the advertised speeds were 'up to 6.7 Mbps' while the real median was 3 Mbps and the mean was 4 Mbps. This implies that ISPs were falsely advertising by at least 50%."

...In other news scientists have discovered that water is, in deed, wet.

Re:median and the mean

[Anaerin]

Perhaps my sarcasm detector is malfunctioning here, but WTF?

The way I was always taught it, the Mean is the Average, the Median is the middle-most number of the sorted set. So if you had:

• 10
• 3
• 2
• 1
• 1

Then the Median would be 2, and the Mean would be 3.4

My new ISP

[munky99999]

I formed my own new ISP. My service was UP TO 500 mbit; it was simple you just got the 56k modem and dialed in. Naturally I charged as if you got the full 500 mbit. Ofcoarse there are unnameable customers who got the full 500mbit. It was just distance sensitive as in you could only get 500mbit if you were connected at the central office.

Upto on its way out in UK

[sane?]

The UK regulator has been talking to the advertising standards people, and its likely the voluntary 'code of practice' will get toughened up to prevent ISPs using the 'upto' get out in advertising. Of course there is the risk that Cameron will scotch things, but the requirement for 'typical' speeds is likely in the medium term future.

Re:Running at 4% in Hong Kong

[nhtshot]

Come across the border into the mainland...

I'd kill for 1.5Mbps!

MOD PARENT UP

[nhtshot]

In the business telecom world, we've had CIR as long as I can remember. We don't have the ambiguities that plague residential communications. Of course, we pay substantially more for the privilege.

I think the label that they suggest is s GREAT idea, just put a CIR on it and we're good to go.

"Up to" means "less than"

[Clovert Agent]

It's standard marketing bullshit. Every time you see "up to" in an ad, replace it with "less than". "Up to 10mbps", "up to 80% shinier hair", "up to whatever". If one out of the entire sample/customer base experienced an anomalous outlier result, they will claim "up to" that. You're statistically unlikely to be the anomalous outlier, therefore you will experience less than what they're claiming.

"Less than" is more accurate anyway. What you experience may be anything in a wide range of values below that, but you KNOW you won't experience more. So do the mental substitution, and I promise your perception of advertising will change as a result.

Re:"Up to" means "less than"

[Quiet_Desperation]

I feel your pain, Mr. Agent, but, still, "up to" and "less than" seem mathematically equivalent to me in the real world.

To my local ISP's (Charter) credit, I have their "up to" 5 Mbps service and I quite often get that. Any downloads that come under that could just as easily be a problem elsewhere than the link between Charter and me.

They haven't added any tiers in a while, though, so that Verizon FIOS 25 Mbps (up to) might need trying.

Re:"Up to" means "less than"

[weicco]

In Finland one ISP markets their mobile internet to work "full speed all the time". Full speed is cleverly defined as "the maximum speed of the network at the time". So even if their service is transferring 1 byte per sec it's still "full speed" because it's "the current maximum speed of the network".

But I must say the same provider's ADSL works like a charm. 10 mbps as promised if only the other end can upload that fast.

Re:"Up to" means "less than"

[sorak]

But if I see a price advertised as "less than", or "under", then I mentally substitute "almost exactly". I remember a long time ago, one of the Atari systems was advertised as "under 50 bucks", because the price was $49.99. This was the first time I realized that principal.

I wonder if there is a marketing bullshit translator somewhere.

Re:"Up to" means "less than"

[TheSync]

It's standard marketing bullshit. Every time you see "up to" in an ad, replace it with "less than"

Indeed, your local loop distance (and to a smaller extent other issues like loop noise and loop quality) will affect your broadband speed.

Oh stop

[Sycraft-fu]

Seriously the amount of whining geeks do about Internet speeds is amazing. The up to thing is perfectly fine, they are not ripping you off no matter how much you try and cast it as that. What they are doing is telling you what the rate cap of your line is. The actual throughput will vary based on load of the segment and so on.

Don't like it? Buy a better plan. You find with business plans bandwidth is shared a lot less so you get more. Often they'll have commitments to certain minimums. Still not good enough? Get a high end connection with a CIR. This is a committed, "never to go below" minimum that you will get no matter what.

Yes it costs more money. Deal with it. Higher quality of service costs more money, and I don't want to hear bitching about that. If you can't understand that you are being petulant.

For that matter, your financial example works against you. The more of a commitment you want on a rate, the lower the rate. In the stock market you can get gains "Up to, maybe exceeding, 10,000%" You pick the right small cap stocks you can make a killing. However there are no guarantees. You can lose your principal easy. You want more stability maybe you look at corporate bonds. Lot harder to lose your money there, but you discover interest is much lower.

Likewise you can put money in a high interest savings account, and they'll offer you an interest rate, mine is 1% right now. However that can change. They don't guarantee that. I get whatever they can offer right now. If I want I can buy a CD, lock in a rate. However that comes with more costs, in that I can't get the money whenever I want and the rate won't rise.

What it comes down to is the more commitment you want the more you pay. With net connections guaranteed bandwidth is expensive. If you demand it, great, but pay the price.

Me I go half way. I want better than consumer connections so I get a business class line. However, no CIR. Sometimes my speeds drop and I live with it. If they are too low or for too long I can call the ISP.

Re:Oh stop

[Anonymous Coward]

Seriously the amount of whining geeks do about Internet speeds is amazing. The up to thing is perfectly fine, they are not ripping you off no matter how much you try and cast it as that.

I am on dial-up. The connection is snappy and the latency is low. Adblock and noscript ensure I make few simultaneous connections. I also block content until and unless I want to hear a sound or see a picture (videos are saved for work with the T1s). What makes my experience speedy, is having a computer that works and knowing how to operate it. I doubt many intelligent people whine about speed. This is about advertised speed which ISPs use as a dick-measuring contest. A real-world comparison telling you what bandwidth you can expect (in addition to how much you can download, latency, and other tidbits) ought to be welcome by all.

What they are doing is telling you what the rate cap of your line is. The actual throughput will vary based on load of the segment and so on.

Not when they design your final mile to be less than the rated speed. The top speed should be achievable at low bandwidth usage times. It is not, you have been sold a bill of goods. NOW, that may not be worthy of action on your part if ther service is otherwise worthwhile. However, as honest, thoughtful people we should strive for a world in which technogobblygook isn't used to cheat the less knowledgable. When I upgrade to DSL - or whatever - I will get the lowest speed because it is more than I need.

Re:Oh stop

[cynyr]

so if i have 3 "up to" tiers, 3Mbps, 6Mbps, and 9Mbps, and the costs are x, 2x, and 3x. You are saying that it is fine that the 9Mbps tier run at 2Mbps 99% of the time, and cost 3x that of the 3Mbps tier that gets the same speed?

Heres the thing, when you see three tiers like that you assume that the 6 is 2 times as fast as the 3, and the 9 is 3 times as fast as the 3... Now if there is one person close to the office that has a dedicated line(only house down that path, mandated to have the line by local government), and they pay for the 9Mbps line and get 9Mbps most(>80%) of the time, then they can claim up to 9Mbps.

The normal consumer expects each tier to be faster than the last most of the time. We understand sometimes things break, or get congested, or a tree falls on a line, but "we oversubscribed the line, so only 1am-4am gets full speed" isn't a good excuse. If the only thing that cause max speed to be lower than the 9Mbps, was TCP/IP, frame, and FTP/etc overhead, then fine. None of the data shows that they advertise raw bit rate, and you get less based on your transfer protocol. The real issue with this, is how is the consumer to make an informed decision on whether DSL A is faster than Cable B. Both say up to 20Mbps.

Re:Qwest

[Anaerin]

160 KByte/sec = 1280 KBit/sec = 1.25 MBit/sec, so you are getting 83.3% of what you're paying for (And chances are the other 16.6% of your bandwidth is going into TCP packet headers, packet loss/retransmission, and plain and simple "distance from the provider" losses).

Those durned "boosts"

[gaelfx]

I'm guessing this is just a rehash of old complaints against pretty much all major ISPs in the U.S. regarding their traffic shaping practices, in which they give you that max speed only for a few seconds at a time when you are initially downloading the contents of a webpage, but then they cut it off after maybe 10 seconds to prevent you from actually enjoying any kind of video or pictures that you might be enjoying.

They seem to believe that we webbies ought to adhere to their standards for what "normal" internet use consists of, rather than using the web the way we would like. It's a very rationalized argument that probably works very well on judges, most of whom don't actually understand much, if anything, about what most people would consider "normal" web traffic, and their ignorance is probably largely due to the fact that they have to spend so much time researching all the useless information out there about the frivolous lawsuit of the week.

Ok, maybe a little out there, but does the point get across? It's like everyone in the U.S. has been using Catch-22 as a political/business model.

Means

[drumcat]

Just like other items, regulators should force ISPs to disclose mean/median if they want to use their "up to" shit.

won't let me post without a subject.

[euphemistic]

This won't be popular with the libertarians out there, but how about it's just mandated that you can only advertise the theoretical maximum when you clearly outline what the median speed is for your existing customers and that the stated median has to be re-evaluated before any new piece of advertising can go out. It can go in the small print at the bottom of the screen/poster/whatever but it has to be visible on there somewhere.

I for one know that this information would be the first thing I'd look for when choosing a new connection/ISP and save a fair bit of digging in their websites.

Re:won't let me post without a subject.

[SecurityGuy]

As one of those libertarians, I propose this test:

Go read all the laws that apply to you. If you can't due to sheer volume, there are too many, they are too complex, or both. Proposing law for trivial problems is really not a good thing.

The sky, it's, it's...

[Cesa]

Hey guys look! The sky is blue! It's BLUE!

Speed to where?

[gmuslera]

Having a country/ISP with badly oversold bandwidth, im very used to the difference between the speed of your connection to your provider, versus the speed you could eventually get, depending on hour, day of week, or things like that to "internet". The physical connection to my provider could deliver that bandwidth, and that is the speed that is sold, but usually cant get that speed to the sites i visit.

Regarding lies or not, "up to" means "less or equal". They would be lying if you get more than that speed, but getting even 1bps would met their claims.

Re:Qwest

[mindwhip]

Or maybe you (or your ISP) are getting your bits and Bytes mixed up...

A speed of 1.5Mb/s (Mega bits per sec) will give a download speed of around 160kB/s (kilo bytes per second) since there is 8 bits in a byte and normally some overhead...

So if you have a 1.5Mb/s service and get 160kB/s you are in fact getting around 100% of the advertised speed...

Or maybe you already knew this... but the difference between Mbs and kBs is something that a lot of people (customers and ISP support) just don't get...

Awful Headline/Summary, Great Story

[leppi]

This headline and summary are horribly written. I'm not an English major by any stretch of the imagination, but even I can pick out a loser when I see it. It's misleading, factually inaccurate, and leads to a number of useless comments like "UHHH, DUH? Who Thought that up to 10Mbit meant 10Mbit? Are they stupid? Not a story!". The article clearly states in the first freaking paragraph that this isn't news to hardcore nerds, but *is* news to the unwashed masses. Take a look at the difference a headline makes:

    Ars: Your fears confirmed: "up to" broadband speeds are bogus

    Slahdot: ISPs Lie About Broadband "Up To" Speeds

I'll leave analysis of the fugly summary as an exercise to the reader. Slashdot editors really need to clean things up. It happens multiple times per day, is frustrating, and really dumbs down the comments. It's not youtube yet, but it's close.

...Officially added to the noise

Depends on who you have...

[XenophileJKO]

I have to say it depends on who you have... I have charter cable and they advertise upto 25mbit down and 3mbit up. It obviously depends on where I am trying to download, but I get consistent speeds of between 23-30mbit down and a pretty solid 3mbit up. I'm not a huge fan of charter cable, but they have always been right at or above what they advertise for me.

Translation

[darth dickinson]

When you hear "up to" just mentally translate it to "less than". Solves many problems...

It's not a lie

[Weaselmancer]

"Up To" means "Less Than or Equal To".

What are all the numbers you can name from zero "up to" 6.7? Would you expect to encounter 3 and 4 on your way up to 6.7?

It's misleading maybe, but it's not a lie. They are publishing their maximum possible speed. YMMV.

Re:It's not a lie

[DimmO]

Would you expect to encounter 3 and 4 on your way up to 6.7?

Four shalt thou not count, nor either count thou two, excepting that thou then proceed to six point seven.

Re:It's not a lie

[Surt]

And if you are conveniently located near a junction, you can sometimes get more than the advertised up-to speed. I had comcast 5mbps service and routinely got 30mbps.

Re:It's not a lie

[yukk]

Ahh, but is it lying to say "This truck can do up to 100" when in fact it's been speed limited at 60 ?

Also, all the natural numbers up to 6.7 should include 5 and 6 !

Re:It's not a lie

[beh]

Indeed, if it's speed limited, then it would be a lie.

With your broadband modem, if it's configured for 6.7MBit/s, then that is your speed limitation. Whether the network behind it can serve it is another matter.

The article itself, on the other hand, is doubly bogus - for one thing, they don't seem to get the wording 'up to', the other thing is that the compare the MEDIAN speed to the 'up to' speed.

Picture this: Your sports car can go UP TO 300km/h. There is no speed limit on the motorway (in Germany, at least) - yet traffic moves at a median speed of around 130km/h.

Does this mean the 'UP TO 300km/h' on your sports car brochure is wrong?
No... The median speed has nothing to do with what the car would be capable of.

Same thing here - if they were to say 'up to 6.7MBit/s' in the brochure, but their observed TOP speed over half a year would never go past 4MBit/s, they might have a case. Saying the median is lower than the top speed - only one word springs to mind: Duh!

Re:It's not a lie

[delinear]

It's more like saying "this truck can do up to 100 MPH" and then omitting to mention that the actual maximum road speed is 60, the 100 is only when it's going over a cliff. It's not "technically" lying, but it's going to be effectively untrue to the vast majority of people who buy the truck and the truck company is fully aware of that fact when they pay for the advertising campaign. It's nice to see that finally, here in the UK, ISPs are now starting to offer a pre-installation speed check so you at least have some indication of what speed you will get. Of course if they do the check off peak it's still going to be misleading, and they're only doing it because they're scared of the growing public backlash forcing legislation on them but even so it's better than nothing. Now if only we could get them to stop lying about their "unlimited" download plans.

Re:It's not a lie

[Panoptes]

Here in Jakarta we have bit of a puzzler. The advertised speed of First Media's 'FastNet' service is "up to 6mbps", so I monitored the connection for a day. For 30 minutes you get just under 6mbps, as advertised. Then, for the next 37 minutes, you get just under 4mbps. And so on, throughout the day. I'm annoyed because there's no mention of this in their account agreement - even in the small print. So although I'm getting 6mbps for some of the time, the maximum that I can get over a period of time is significantly less. Do other ISPs in other countries pull this one, or is it an Indonesian one-off?

Re:It's not a lie

[hedwards]

It's the largest number they could get away with. They could've used up to 4 gbps as well. This isn't math, if you're claiming up to that amount, then you damn well better be able to provide it. The assumption being that at some point during a given month that you've got that much bandwidth. What's going on is simply a matter of fraudulently selling people less bandwidth than they're advertising to make more money.

Re:It's not a lie

[hedwards]

I haven't been car shopping in Germany, but around here we don't expect to get to drive our cars that fast. I can't recall the last time Ford or Toyota advertised that their cars could go X mph. Consequently nobody expects to be able to drive that much faster than the speed limit of the local freeways. There is a well understood constraint that the highway patrol will pull you over for speeding and that during rush hour you going to be constrained by traffic. However this is a completely different matter. ISPs are advertising and charging for bandwidth and then claiming they don't have to provide it because they only promised up to that number and since most ISPs don't provide the tools to actually measure it most people don't know whether or not they're getting what they were promised.

The reason for the up to figure is that things get congested at times and things like latency conspire to drop it somewhat from the peak, not we don't feel like investing in the infrastructure necessary to provide what we're promising. Worse still they'll lie about the actual speed they're providing you if you bother to ask. If in a given month you aren't spending significant periods at or near the number they quoted you then they're doing something wrong. If they can't provide you the service they're promising at the price they're charging that's a business model that needs to be fixed. Other wise this really ought to be called what it is fraud and handled by the FCC or local AG's office.

Re:It's not a lie

[cgenman]

If I sold you a sports car that could do "Up To 300 km/h," but failed to mention that is only when being dropped out of an airplane, would that be lying?

It comes down to what the problem is. If it is congestion on other people's networks, then the ISP is fine. But let's be frank: that's not the problem. If the problem is with the ISP, then there is a problem. To cover some basic grounds, the main reason why a line is slower than the advertised speeds are:

1. Oversubscription. ISP's sell far more upstream capacity than they actually have. In a well balanced system some of this is fine, necessary even, but the degree of oversubscription is silly.

2. Lines that can't possibly live up to advertised speeds. DSL springs to mind. By nature of DSL technology, you're guaranteed to never hit the maximum line speed, which is frequently the number the advertisers plaster up there.

3. Outdated switching tech at the ISP. This should never be a problem, but happens more often than it should.

4. Shared lines. Comcast seems to be getting better about this. But if other local users effect whether or not you can reach your advertised speeds, you're almost guaranteed to be on the same loop as Mr. McFilesharer.

I don't know. I'm hard pressed to think of another area where "up to" can be used as the only number a consumer encounters, rather than an average. Can people sell "up to" 12 ounces of orange juice, and deliver 6? Can you be billed for "up to" an hour with a personal trainer, even though you only got 1/2 hour?

Re:It's not a lie

[tehcyder]

It's an age-old advertising trick, as in: "up to 100% dandruff free hair guaranteed!"

Re:It's not a lie

[clone53421]

Indeed. And since when was it equivalent to the median speed?

Looking at the recorded weather statistics from last week, the high temperature was an average of 99 degrees Fahrenheit... but the median temperature was 88! The actual recorded weather data is falsely advertising by 12.5%?

And why the hell are we using median anyway? It makes more sense to use the mean...

Re:It's not a lie

[tophermeyer]

With your broadband modem, if it's configured for 6.7MBit/s, then that is your speed limitation. Whether the network behind it can serve it is another matter.

On that point I would have to disagree. When I contract service from the provider I am not interested the differences between their hardware and network capabilities. If they sell me a service package with a stated speed, I expect to be able to get that speed to my house.

I understand your point about the technical limitations of the available network (and as a geek I would be interested in knowing more about the network test results to my home). But if the service provider is not able to provide 6.7MBit/s to my home, they shouldn't be advertising and selling that capability. More importantly they shouldn't be charging me for that capability. Geeks know enough to understand how they are being shafted, but the average consumer probably does not.

Re:It's not a lie

[realityimpaired]

The fine print usually specifies the minimum... I work for a DSL provider, and for the 7mb plan, it specifies a synch rate of 3008/800 => 7040/800. If your synch rate is less than 3008/800 then there's a problem that needs a tech. visit, which we'll book for free, and if it can't be fixed we'll switch you to the 2mb plan at no cost (which specifies 512/512 as the minimum). Usually, we'll even refund you the difference for having sold you a service you couldn't get.

We don't advertise it, but usually for customers on the 2mb plan we'll put them on 2496/800, and for customers on the 7mb plan we'll put them on 8512/1080 or 8512/800 if it's possible. As a courtesy to customers who are on the old 5mbit plan (which isn't sold any more), we'll usually put them on the 8.5mbit profile too. That's because of OS overhead and packet routing overhead, if you're actually synched at 7mbit, you will probably only see 6-6.5mbit depending on what OS you're using. By putting you on an 8.5mbit profile, you'll actually see 7mbit download. We do similar things for customers on the 10mbit, 16mbit, and 25mbit plans, though usually the 16mbit and 25mbit plans are using VDSL modems, not ADSL, and those tend to have less overhead and better throughput.

Of course, if you have a defective filter, a badly installed jack, a short in the wiring, or poor quality wiring in your home, you're still going to have issues. It doesn't matter if you're synched at 8512/800 if you're getting about 50% packet loss or experiencing 20-30 resynchs/day because you went to the dollar store and bought a 100' extension cable to put your DSL modem (which does have a wireless router built in) in the same room as your laptop computer... which does happen a *lot*.

The ISP's are in it for the money. But they're also going to know whether you're going to see the advertised speed, and we're pretty good at telling you when you won't see it, and at not selling you the service in the first place if you're not going to reach the minimum.

Re:It's not a lie

[parlancex]

Additionally, nobody seems to remember that protocols have a non-trivial overhead and assume their bandwidth consumed is equal to the throughput of only their payload bytes. Just as a quick example of how big of a difference this can make, enabling jumbo frames on a high throughput iSCSI link can often increase the bandwidth of that link by up to 50% by reducing the amount of packet headers and increasing the number of payload available bytes per packet.

Re:It's not a lie

[networkBoy]

It's certainly creative...
I was on AT&T (SBC) and was on a 6Mbps plan. I could only get about 3.4, so I called and downgraded my plan. Then instead of the expected solid 3Mbps, I started seeing 1.2-2.6Mbps.

SHENANIGANS!

Re:It's not a lie

[clone53421]

I’m using a Pentium, you insensitive clod!

Re:It's not a lie

[pruss]

"Up to" does not mean "less than or equal to". Then they could truthfully advertise that they go up to a googolplex bytes per second. To my ear, "up to x" means something like "less than or equal to x, and sometimes reaching x (or at least arbitrarily close to x)". That's still compatible with the average speed being about half of x, as long as on rare occasions one hits x. And it's not even all that misleading, I think. Maybe if the average speed was about a tenth of x it would be misleading.

Re:It's not a lie

[nametaken]

7 is right out.

This they will make sure of.

Re:It's not a lie

[Quirkz]

Hm. I would have thought the appropriate car analogy was the advertised gas mileage, where they do indeed advertise "up to" X MPG on cars. They do this because environmental factors and personal driving tendencies do affect the mileage you get, and are not directly under the control of the original manufacturer.

For reference, my 97' Honda Civic was advertised as getting something like "up to 35 MPG highway/28 MPG city." In practice, I don't usually quite hit those numbers. 30/25 might be more realistic as the average. In heavy traffic in Chicago I probably had really bad weeks were I was closer to 20 MPG. On the highway, in bad conditions or when driving too fast, I've seen 30 MPG. On the other hand, on a good day I've actually been over 40 MPG when driving in certain conditions (cruising slowly on mountain highways, for instance).

It is of course absurd (and illegal) to advertise a rate that it's not possible for the customer to get, but it would also be absurd (and probably shooting yourself in the foot) to only advertise your mean or median value when in good conditions you can get a lot more. Whether it's criminal, a lie, good business, or accurate enough -- I would imagine there are standards for determining that, but I don't know what they are.

Now, I personally *would* argue that the numbers I cited for my car are close enough to the advertised numbers that I think Honda gave me good information, and I would argue that the numbers cited for broadband rates are NOT close enough to the advertised rates to be truthful advertising. Still, I'd expect broadband speeds to have to have a wider standard deviation than mileage -- there are a lot more outside influences to slow things down. To counteract that, though, I'd expect broadband to be advertised with a rate under the real theoretical max, so that your good days could start to make up for the bad days.

Re:It's not a lie

[Weaselmancer]

Exactly! Thank you for getting my point. When someone says "up to", what they are saying is "you won't get any better than this." It's not a promise of a minimum, it's the promise of a maximum. "You will not exceed this." That's all it really means.

Re:It's not a lie

[Weaselmancer]

"Up to" does not mean "less than or equal to".

Yes it does.

To make it easy we'll stick to integers and start at zero. Now - name all the numbers starting at zero "up to" ten.

0 1 2 3 4 5 6 7 8 9 10. Done.

The mathematical operation you've just seen is "less than or equal to."

Re:It's not a lie

[SomePoorSchmuck]

Exactly! Thank you for getting my point. When someone says "up to", what they are saying is "you won't get any better than this." It's not a promise of a minimum, it's the promise of a maximum. "You will not exceed this." That's all it really means.

And that's okay with you?
That doesn't make it any less deceptive. It just makes it deceptive AND semantically meaningless.

By your logic, it would be perfectly acceptable for them to sell you a contract promising "Up to 1.21Terabytes/second", and then deliver no more than 1Mbps in practice. According to you, all it really means is "You will not exceed 1.21Tbps".

That is absurd logic which requires a Cirque de Soleil contortionism of the English language.

Re:It's not a lie

[SomePoorSchmuck]

Or even a Cirque du Soleil contortion... ugh.
I fail at the Canandianese language. ;-D

Re:It's not a lie

[Weaselmancer]

No, "up to" means exactly that. Up. To.

It means you have to be able to demonstrate that the service you are selling is capable of getting up to that rate. It does not mean that it is a minimum. It's simple English.

We agree on the deceptive part though. Keep your eyes open for "up to" in advertising. You'd be surprised how often it shows up. Once you know it's a ceiling and not any other promise you'll be shocked.

Re:It's not a lie

[Golddess]

If I sold you a sports car that could do "Up To 300 km/h," but failed to mention that is only when being dropped out of an airplane, would that be lying?

Yes, because without additional acceleration, it'd only get up to 35.28 km/h. :P

Re:It's not a lie

[SnarfQuest]

If you are paying for an "up to", you'd hope that you'd reach near that value every now and then, instead of never reaching half that value.

Re:It's not a lie

[SomePoorSchmuck]

But the reality is, you can't get those speeds in practice. By "you" I mean the set of all individuals paying for the service. Is it possible, given the current structure of their network, for all persons on the network to get that maximum speed? The issue isn't the speed of the technology (the routers and switches and modems) but rather the amount/speed of your transmissions across their network, which is what you're paying for. Just because the technology is capable of going that fast doesn't mean the specific service you're buying actually can go that fast. Because if it could go that fast for everyone using the system, it would. And if it doesn't go that fast for everyone, then there must be some reason why it doesn't. And that reason is why even though the theoretical limit is higher, you in fact can't get those speeds.

Stop the "oversubscription" canard.

[Estanislao Mart�nez]

If by very little control, you mean oversubscribing the line by a factor of N, then okay, whatever.

The whole point of shared computer networks is to oversubscribe the shared lines so that intermittent users can get higher maximum bandwidth than they would otherwise for the same cost. I don't think that you have some revolutionary plan that would allow every node in the Internet to get a dedicated line connecting it to every other node and thus get a guaranteed 24/7 full bandwidth to any node you wanted at any time--and if you had that plan, well, for the same amount of investment we could build a network that delivered higher average bandwidth to all the users by sharing the lines when they were not using them.

Re:Stop the "oversubscription" canard.

[Tanktalus]

Better than that - recently my ISP, Shaw (shaw.ca) increased all of its plans by 50% without changing rates. So I was getting 10Mbps, now it's 15Mbps. That should, in theory, get me up to about 1.9MB/s. However, they also apply a SpeedBoost "technology" (yeah, they just allow extra bursts at higher rates) such that I've seen 2-3MB/s speeds from some mirrors. And wherein the boost is supposed to be for brief durations, though they never really say how brief, I've had sustained 2-3MB/s speeds for 10-30 seconds before settling back down.

Their customer service sucks. But at least I almost never need to talk to them. And when I do, I generally ignore their questions about operating system. Works better that way. Especially since the only times I call are when the cable modem itself no longer has the upstream light on.

Re:Stop the "oversubscription" canard.

[Shakrai]

But at least I almost never need to talk to them. And when I do, I generally ignore their questions about operating system. Works better that way. Especially since the only times I call are when the cable modem itself no longer has the upstream light on.

Time Warner likes to blame problems on "routers". Their techs are smart enough to figure out that your provisioned MAC address belongs to a Linksys WRT54G but not smart enough to realize that you can't blame said Linksys for a network outage when the cable modem itself loses sync. More's the pity. I've gotten to the point now that I clone PC MAC addresses for all of the routers that I install for people. It's just easier that way if they ever need to call TWC support.

Re:Stop the "oversubscription" canard.

[Capt.DrumkenBum]

Shaw sucks. I once waited all day for their tech to show up, to replace my modem. (I didn't want to change it out, they did.) He/she never did show. When I called they wouldn't even give me an expedited appointment. I had to wait 2 weeks for their next weekend appointment.

The only good that I can say is they are not as bad a Telus.

Up to means...

[woboyle]

What "Up to" really means depends upon a number of factors, and the type of connection you have (DSL, Cable, FIOS, Satellite) is a big part of that. With DSL, it depends upon your distance from the nearest hub/switch. My business DSL from AT&T advertises "up to 6mbps". However, I get a consistent (barring cable problems) 5.1 to 5.2mbps. Because of the technology used, having a bunch of connections in the neighborhood should not be an issue - all of them go to a switch that is connected to a fiber optic cable that goes back to the central office. Cable is another issue. You will get advertised speeds (or close to it) ONLY if no other connections in the neighborhood are active as they share the cable bandwidth and other connection links. FIOS pretty much gives you what you pay for since it is fiber to your system and the aggregated fiber connections multiplex into very high bandwidth WDM trunks. Satellite is somewhat iffy since a number of people are probably going to share a transponder on the orbital bird. So, DSL should be somewhere close to what you pay for, cable depends upon traffic, and FIOS will probably be the most true to promised bandwidth. I've had everything but WiMax and satellite over the past 10 years and FIOS was the best (not available where I am right now). DSL has been the next best, though its dependence upon copper wires (and in my neighborhood, those are 50+ year old lead cables) is its biggest drawback.

No, really?

[rickb928]

One of my more annoying clients finally got rid of the DDS2 lines and got local Internet connections everywhere. While making PiX VPNs actually work 24x7 and maintain out-of-band service connections was overwhelming for a while, I got a kick out of his expectations. He wanted to deliver software everywhere, and while it worked, it was never fast enough.

His big complaint was for a specific location in the middle of Maine. The local cable company promised him a 12MB down/3MB up connection. In 2004.

The cable modem had a 10Base-T connector. Yeah, and the PiX had 10Base-T also.

He never quite understood why the cable company would advertise the speed, and kept pressing me to get the modem in the 'right mode' to go over 10MB. Kevin, this one's for you.

I'm figuring my current Cox modem has a 10/100 port. This is good. But I'm not getting anywhere near what they claim.

When I lived in Portland, Maine, there was a period when Time-Warner delivered stupendous (for the time) speeds, and no one really seemed to care. It took a while until they sold enough users and started to tax the system, then it got slower and slower. Back in 1998 or 1999 or so I dumped 9GB from Novell's ftp site overnight. It was much faster than I expected, but I had ditched the ISDN BRI line I had for work, and it was night-and-day faster of course. Plus my BRI terminated to our T-1, and shared that with a dial pool and stuff. Ah, those were the days.

Surprised?

[QuietLagoon]

This implies that ISPs were falsely advertising by at least 50%."

.
.. and this is a surprise ... why?

Does anyone here really expect the likes of Comcast to be truthful in their advertising?

Deliberate obfuscation?

[Narcogen]

They find that on average, the advertised speeds were 'up to 6.7 Mbps' while the real median was 3 Mbps and the mean was 4 Mbps. This implies that ISPs were falsely advertising by at least 50%."

This implies that the author of the post does not understand the distinction between "peak" (meaning by "up to") and average or median.

C'mon, people. Yes, ISPs are evil and do bad things. That doesn't give us license to deliberately confuse things in order to make them look worse. It's unnecessary.

Re:Deliberate obfuscation?

[Arker]

This implies that the author of the post does not understand the distinction between "peak" (meaning by "up to") and average or median.

You might get that impression from the blurb here but if you read TFA it does not, in fact, miss that distinction. What you are missing is the assertion in the article that, although they do advertise 'up to' they make that phrase inconspicuous with tiny letters next to very large numbers, and focus on those numbers. So while the advertisement may be technically correct, it's also deceptive nonetheless, and many consumers who are not knowledgeable in the area are deceived by this.

It is true

[tburt11]

Yes. I can confirm it it true. I have a Cable TV Internet Service in Los Angeles. I ordered their second tier offering, Something like speeds up to 15 mb/s download. After the tech left, I ran a test, and I was getting like 5 mb/s. Alot less than what they promised. I called to ask, and I was told that performance vaires by location. OK, was my reply. If I get only 5 mb/s, then why should I not pay for the 10mb/s plan instead of the 15? I still only get 5, right? The tech asked me to wait a moment, and returned to say that my account had been setup at the standard 10mb by mistake. She fixed it and now I would get 15. I ran another test, and now I get 8 mb/s. So what changed? If I get 5 on the 10mb plan, but it only takes a quick config change to get me 8, then why can I not get 8 on the 10 mb plan?

Oh blow it out your ass

[GrumblyStuff]

So on average, they're delivering 50% of their 'up to' speed and that's alright? What if it was 40%? 30%

How low would they have to go before you would say "Hey, this is a fucking rip!"?

Re:Oh blow it out your ass

[Sycraft-fu]

It would depend on what speed I was getting vs what I was paying. If I paid $50/month for a connection that was "up to a gig" and only got 5%, 50mbps, I'd be just delighted. If I paid $200/month for a connection "up to 1 mbps" and got 100% of that I'd be looking for a new provider.

I also do not at all mind the idea of a burstable connection. That's what I have at work. If I tried to use 100% of the connection 100% of the time I'd be hearing from higher up real soon. However it can go really fast when I need something.

I'm a pragmatist about it. I look for a connection that is as fast as I decide I want for a price I'll pay.

However seems most geeks have a real entitlement complex about it. Well here's the solution: Don't buy it. go with something else. Like I said, if you want CIR you can have it. However it'll cost you. You want high speeds for cheap? Well then it is probably shared bandwidth which means that you will have lower rates during peak times.

And yes, this is true even in other nations that people love to hold up as having perfect Internet, like Japan. In fact it is often more true there, because of the way it is set up. Guy there told me how great his 100mbps connection was because he could download a CD worth of data in a little under 10 minutes. Had to tell him that's not 100mbps speeds, that's 10mbps. Still nice and fast, but not 100mbps.

This is why I love FiOS

[Golbez81]

I'm paying for 35/35 and I download "up to" 50Mbps =]

Re:This is why I love FiOS

[BorgDrone]

50Mbit on fiber? Am I the only one who thinks that is ridiculous ?

I'm getting hooked up to fiber in 2 weeks and I'm quite annoyed that they offer nothing better than 200/200. I would have expected at least gbit on fiber, I don't see a reason they need to limit it at all and gbit seems to be a good point since most consumer grade computers come with gbit NIC's nowadays.

After a certain point, it doesn't really matter how fast your internet connection is, you're not going to use any more bandwidth anyway. I have about 500-600GB of traffic a month on 120Mbit cable, this will not significantly change if I go to 200/200 or gbit. All the limits do is ensure I have to wait a little longer for a download to finish, they are waisting my time for no good reason.

Re:This is why I love FiOS

[Golbez81]

You have to realize how most of the fiber bandwidth is devoted to the HD cable. (This could be solved by IPTV, but FiOS only uses that for On Demand) The bandwidth that is left over depends on if you're on a BPON or GPON (Gigabit BPON). MOST people on FiOS are on BPONs. * 1310 nm wavelength for upstream data at 155 Mbit/s (1.2 Gbit/s with GPON) * 1490 nm wavelength for downstream data at 622 Mbit/s (2.4 Gbit/s with GPON) * 1550 nm wavelength for QAM cable television with 870 MHz of bandwidth This bandwidth is shared by 32 people usually. 155/32 = 4.84Mbps per person upload 622/32 = 19.44Mbps per person download So they are technically over subscribing their bandwidth. I'll gladly admit that FiOS is setup retarded and I've complained about it for years, but its the best that most people have... Hopefully U-Verse fiber will solve that issue when it comes out and offer higher bandwidth, but AT&T has a long way to go on upgrading their network to Fiber. =[

Most people dont understand statistics, or fallacy

[JohnnyComeLately]

Mean, Median, Mode all are useful, but not when used in most advertising. If the lottery advertised, "Most of you will get, at best, a $1," would you buy it? No, they say, "You CAN win a bajillion dollars." Will you? Not likely. You are statistically MORE likely to hit the "up to" advertised speed of the broadband speeds on the internet, just possibly not at YOUR (mom's) house (basement). I get the "Up to" speed, but it's during the day. Friday evenings I drop from an AVERAGE (not peak) of 16 Mbps to around 8. I can stream two netflix movies and still surf.

It's funny, I read all sorts of comments but no one realizes it's everywhere. Your network card on your computer is likely "Gigabit," and you wouldn't have bought the 100BasteT card next to it for $5 less, however for various reasons your Gig card rarely goes over 10 Mbps. So maybe Cisco, Dlink, LinkSys, and the like ARE LYING! I mean, everyone knows that wiring termination can be substandard (cross talk reduces bandwidth), people run Windows which is stuffed like a pig at a festival, and various other bottlenecks.

That said, it's painfully onvious because if the original author said, "The average is 5 Mbps, and the standard deviation is 5Mpbs," a large number (if not all) would fail to realize this covers over 10Mbps, the "stated max."

Comparing this to physical goods is foolish

[AsmordeanX]

I see posts here about people saying that if a snack product was sold as "up to X amount" that wouldn't fly for consumers.

Of course it wouldn't. If I buy a bag of M&Ms (they happen to be sitting in front of me) it's very easy for Mars Inc. to work out that the bag has 200g in side of it (give or take a gram). If it is 194g, well throw another candy coated nut in it. Ah, 200.5g. Close enough.

Selling access to something that is mostly beyond your control is foolish to try and actually guarantee a speed.. It's like building a wind turbine and being annoyed that it isn't producing the advertised 6.5MW of power on a calm day.

Now what I would like to see would be a standardized speed test that they would be forced to post as an average and make that the big wording. "7.5Mbps AVERAGE* *Average speed may vary, maximum speed of 15Mbps."

It really means ...

[Katchu]

It really means that the speed will never exceed the "up to" speed.

Re:Learn some GD math and shut up.

[White Flame]

I'm trying to think of how to sensibly respond to such a way-out-in-left-field response, but the only thing I can think of is:

*whooosh*

My "up to" is getting slower...

[S-100]

Two years ago, on a good day I could nearly saturate the 10 mbps link from my cable modem to my router. Now, it never even gets half as fast, with the same connection topography, same hardware, same cabling. Funny, the ISP now sells a "turbo" upgrade, which I guess means I have the right to pay them more money to get back what I got from them years ago.

Burst and Sustained

[anethema]

This is already a well commented story and I usually don't enjoy commenting on those since my comment will likely just get lost but here it is.

At least with cable modems the DOCSIS configuration file for your modem is given two speeds. Sustained and Burst.

If they spec you at 10Mbps sustained and 2.5 Burst, you will see 'up to' 10Mbps. For normal browsing, it should be like being on a 10 megabit connection, but for downloading, you will quickly be relagated to sustained speed.

The file also has provisions for setting these speeds based on ports. They may really give you ten megabit on port 80 but less on others, its up to your cable ISP.

Call

[S-100]

Good thing the cell phone companies don't sell Doritos. In that case, you'd have to pre-buy the maximum number of bags of Doritos you might want to eat every month, whether you actually eat them or not. And if you eat more than you pre-paid for, they will cost $10/bag.

The other problem

[Sycraft-fu]

Is that bandwidth is expensive to guarantee, nobody will pay for it. For example I work on a university campus. We have gig over most of the campus internally and a good bit of bandwidth out to the Internet. Net effect is really fast transfers. Downloads normally fly and web surfing is limited by your browser not the network. However that works because it is a massive sharing model, an "up to" model. I get "up to" somewhere in the real of 800mbps because that is about the total bandwidth I could use. Realistically I get maybe 100mbps on most transfers and I'm fine with that.

So, what if we switched the model, rate limited every device and gave everyone a committed rate? Well then people would get about 200kbps. That is about what you get when you divide the bandwidth the campus has be the total number of systems. Ouch. That would suck. Also the result would be that most of the bandwidth sits unused most of the time, just idling there so that you can commit to having that 200kbps should someone need it.

Much better not to run things that way, I think.

Nobody likes pay as you go

[Sycraft-fu]

That's why we are where we are. Back in the day ISPs constantly tried to offer this. Particularly when the Internet was young and bandwidth was scarce, this was the best way to do it for high end connections. You buy a DS-1 or DS-3 and get the full transfer rate, but pay for what you use (or usually pay a flat fee for some and usage after that). It allowed for the ability to offer higher rates to more people for less money. They'd show businesses how it'd cost less. Didn't matter, people didn't like it because they could get hit with extra charges. They wanted unlimited.

You just can't have it all ways. You can't have cheap and fast and dedicated and so on.

Personally I think there needs to be less bitching, particularly if the complaint is with low end broadband service, which is (at least in my area) what 3mbps is. That is Cox's "value" tier. You pay very little for it. That's fine, but it is for people who really don't do much. It doesn't surprise me that it is slow. Pony up more cash if speed is important.

Re:Nobody likes pay as you go

[Tim C]

which is (at least in my area) what 3mbps is. That is Cox's "value" tier. You pay very little for it. That's fine, but it is for people who really don't do much. It doesn't surprise me that it is slow. Pony up more cash if speed is important.

The fact remains though that if you are paying for 3Mbps, you should be receiving 3Mbps. If in fact you are receiving "up to 3Mbps if conditions are favourable, but less most of the time" then it should be labelled clearly as such.

Re:Nobody likes pay as you go

[Chowderbags]

To be fair, I think there's two main problems with pay as you go: ISPs gouge on data costs and there's a lot of traffic to your computer that you didn't ask for and definitely don't want. The latter one is impossible to avoid - even if you keep your computer locked down, you'll still get traffic from a shit ton of people that have no business contacting your computer. Seriously, I run Peerblock and see everything from big corporations, to small firms, to government entities on the other side of the work. Either I'm on every watch list ever made, or the average person is getting so bombarded with traffic that overage charges would happen to everyone every month (to the ISP's delight, no doubt).

It's a non story, stop modding this tripe up.

[mjwx]

Not everyone realizes that other people are getting substantially better internet for the same amount from the same company based on the same agreement.

Not everyone realises that speeds with DSL and cable technologies are entirely dependent on the environment.

Lets hear that again, DLS and cable speeds are entirely dependent on the environment.

Now, DSL attenuates over distance, ADSL2+ close to the exchange is around 24 Mbit, at 3 KM from the exchange, its about 8 Mbit. The laws of fucking physics wont let us change that one and the physics police are a bitch to deal with if you even try.

Cable does not attenuate as fast as DSL but it's shared so speed is dependent on the number of users your particular network segment.

I though Slashdot users had an understanding of technology, now we are just making bad analogies that, when you do have a modicum of understanding make absolutely no sense.

Can anyone think of any others?

No,

and you should feel quite retarded for thinking of those. It's like a saying a packet of crisps will automagically deteriorate the further away from the shop you get thus it's "up to 200 grams", because that is as feasible as an ADSL connection that maintains full sync speed at 5 KM from the DSLAM.

Re:It's a non story, stop modding this tripe up.

[Rivalz]

You sir are a idiot. Reading your argument was like listening to a 5 year old with a thesaurus.

They are not entirely dependent on the environment.

Most ISP's prioritize traffic, throttle bandwidth, and do a lot of stuff you might not understand.

What I was trying to convey is the art of overselling in my description and analogy.
Which seemed to completely elude you.

I'm not going to try and educate you because I could care less what you think.

Between X/2 and X

[kthreadd]

A few years ago the local ISPs over here got tired of this discussion and started advertising their services as "between X/2 and X Mbps", where X was what they had previously said "up to". Turned out better for everyone.

Well, there is a solution of sorts:

[Penguinisto]

...what if I, in return, promise to pay the ISP "up to" $45/mo for their service?

Oh, that's right - they'd cut me off. :/

/P

Re:Well, there is a solution of sorts:

[natehoy]

If I had mod points, I'd have given you you "up to" +4,000,000,000 Funny.

Re:Well, there is a solution of sorts:

[cynyr]

did your contract with the ISP come with a SLA? no? right, so you agreed to pay $45 a month for what ever they give you... hmm oddly i'm not sure a contract like that would hold up in court though.

Re:Well, there is a solution of sorts:

[tacokill]

You do realize you have that same option, don't you? It may not be palatable to you but it's there nonetheless.

Re:Well, there is a solution of sorts:

[Shakrai]

If everybody switched away from ISPs that pulled this crap they would stop doing it in short order. Just switch to another local provider and this will all go away. It's not like your local government cut a deal with them giving them monopoly status in exchange for bribes^Wfranchise fees or anything.......

Re:Well, there is a solution of sorts:

[commodore64_love]

>>>what if I, in return, promise to pay the ISP "up to" $45/mo for their service?

That would be okay if that's what your contract read. It does not. It specifically says you will pay $45 flat, plus applicable taxes and in exchange they will provide upto XXX Mbit/s. So they are not in violation if you get less than XXX, but you would be in violation if you only gave them $40.

Not sure if it's true, but I've heard that DSL Internet is more reliable than Cable Internet. Why? DSL gives a dedicated phone line for every home, where cable has a shared line (broadcast to everyone). So if the kid next door starts downloading a Bluray, that will slow your cable internet too.

I know I've always had 7Mbit/s DSL just as advertised, without any slowdown. My cable friends can't make the same claim.

Re:Well, there is a solution of sorts:

[TemporalBeing]

Not sure if it's true, but I've heard that DSL Internet is more reliable than Cable Internet. Why? DSL gives a dedicated phone line for every home, where cable has a shared line (broadcast to everyone). So if the kid next door starts downloading a Bluray, that will slow your cable internet too.

I know I've always had 7Mbit/s DSL just as advertised, without any slowdown. My cable friends can't make the same claim.

Yes, Cable is shared - it's a basic Ring Network. However, it's typically more reliable than DSL exactly for that reason as they allocate more bandwidth (i) on the local loop and (ii) a bit more at the hubs (e.g. Central Office) for the loop than their DSL counterparts. So while at peak you might get a lot less on the local loop, at non-peak you will be able to get a lot more than you XXX Mbit/s.

Re:Well, there is a solution of sorts:

[SanityInAnarchy]

DSL gives a dedicated phone line for every home, where cable has a shared line (broadcast to everyone). So if the kid next door starts downloading a Bluray, that will slow your cable internet too.

However, both ultimately share one or two lines upstream to eventually connect with a backbone -- so it might take more, but if enough of your neighbors are torrenting, that certainly can affect your DSL.

I know I've always had 7Mbit/s DSL just as advertised, without any slowdown.

It's been pretty reliable for me -- but then, so has fiber. Both situations are fairly unusual -- it seems like in most of the country, they oversell their upstream bandwidth, also.

Re:Well, there is a solution of sorts:

[Self Bias Resistor]

Exactly. Just ask Penny Arcade:

http://www.penny-arcade.com/comic/2009/5/1/

Re:Well, there is a solution of sorts:

[non0score]

Really? Time Warner vs. AT&T? Oh, great choices there.... Or do you mean how Verizon customer service responded to me, "Sorry, your address is within the AT&T area, and we aren't allowed to put FiOS there."

Then again, I'm not sure if that was meant to be tongue-in-cheek....

Re:Well, there is a solution of sorts:

[Vash21]

If everybody switched away from ISPs that pulled this crap they would stop doing it in short order. Just switch to another local provider and this will all go away. It's not like your local government cut a deal with them giving them monopoly status in exchange for bribes^Wfranchise fees or anything.......

Although that is the solution you would expect to work for a capitalist society, it unfortunately isn't feasible for most consumers. The problem is most of these companies have a monopoly simply because of a real lack of competition, all because there are other ISP's, doesn't mean that they can offer much better than even the lowest quality service of the larger ISP's.

Where I go to college for example has three possible ISP's, Charter (owned by comcast), which I have, AT&T, which doesn't offer service at my apartment, and the service it does its terrible at best from what i have heard, and some local company which i believe to be owned by a regional ISP, but also doesn't would not provide service at my residence. I was looking at other ISP's where i go to college because my internet would become all but unusable at the times 12pm - 10pm Sun - Sat. Charter's service was becoming so terrible that I even looked into satellite internet only give up and wait for Charter to fix the problem. Apparently one of my neighbors had something wrong with their modem (or had become part of a very active bot net) and was completely clogging up my connection. They fixed the problem about 2 months after I first started calling them multiple times a week, replaced several modems, and several visits from Techs and to the local office.

My case is just an example of a temporary issue where the idea of Capitalism would have been great had it been available to me, sure i could have switched to satellite internet, (didn't do much research on it before i ruled it out to be honest), but i highly doubt it would have hurt Charter's bottom line enough for them to actually fix anything wrong with the system.

In my opinion, the biggest flaw in Capitalism is the fact that it relies on the consumers to be proactive in order for things to be improved. In my experience most people care about few things enough to actually do anything about it and when they do, they have few to no options with which to do anything, even if it is because of deals cut with local governments. It's a pretty sad state that I hope things like net neutrality (not the Google-Verizon proposal...) and nationwide broadband access help address.

Sensassionalist reporting from ars

[guacamole]

Since when "up to" either mean or median? As others had pointed out, "up to" more likely means the theoretical maximum, and I don't see what's wrong with that. Network bandwidth is not like a bag of chips. If internet was like a bunch of PCs are on the same LAN then indeed, 100Mbps would mean.. exactly that much. But as soon you connect to a WAN or any other network, then it's pretty much impossible to guarantee anything. During the peak hours, you sure bandwidth with others, and there may be indeed network bottlenecks somewhere where you could get the "up to" speed during the off hours, but during the peak hours network speeds likely will drop. This is the reality of the networks. So, how do they propose that ISPs advertise the network service capacity? Should they quote mean or the median? How is it going to be measured? Seems like a big can of worms. Just leave it alone. If you don't like your service, regardless of the advertized "up to bandwidth" you can always opt for a better or more expensive option or if you're lucky find an alternative provider. Pretty much everywhere I have lived, I had at least two options, cable and DSL, (sometimes FIOS), with various levels of performance.

Re:Sensassionalist reporting from ars

[Dunbal]

and I don't see what's wrong with that.

      Well next time you apply for a job, what happens if HR tells you you can earn "up to $120,000" but actually only pays you $5000? Still nothing wrong with that? What if the numbers on a fuel pump at the gas station showed the gallons that it theoretically could be pumping into your car for that price? What if the electric company delivered "up to" 120V but actually only delivered 5V with spikes up to 120V instead of an RMS value? Still don't see a problem with being deceptive? Then you get the service you deserve.

Re:Sensassionalist reporting from ars

[Red_Chaos1]

Your argument is flawed. The electric companies *can't* let the AC vary, because things only work at the specified voltage. The same cannot be said for broadband connections. Besides that, none of your examples has *ever* used the "up to" business. However, I bet your cars speedometer displays that you can go "up to" a certain speed, but I doubt it actually could get that fast without help or a long stretch of road (provided it doesn't have a limiter in the ECU).

That ISPs use the "up to" line is nothing new, they've been doing it for years. The maximum may be what you are provisioned, but if you pay attention to the paperwork you sign, you will find they only guarantee up to a percentage of that provision. For example, Frontier, who provides DSL under the Frontier name and a few others, only guarantees 70% of provision. So it you had a 10Mbit line provision, you're only guaranteed to get 7Mbps. If it goes below that, then there is an issue and they will send out a tech to check things out. But as long as your connection doesn't test under 7Mbit and isn't showing frequent disconnections or packet loss, then you're fine.

Re:Sensassionalist reporting from ars

[Dunbal]

The electric companies *can't* let the AC vary, because things only work at the specified voltage.

Not true. You can easily power a lightbulb or small electronic devices at different voltages because these devices do not draw any significant amount of current. Most small electronics convert AC to DC and then only use 5 volts - unless the power supply design is crappy they should also be able to deal with somewhat different voltages. You usually have 125 Amps max coming into your house, and a lightbulb draws a tiny fraction of the 15 amps your plug is usually wired for. You can vary the voltage between 90 and 150 volts and you won't notice any difference in the lightbulb.

The problem happens when you try to run an electric motor (like your fridge or washing machine) - these draw vast amounts of current especially on start up. If you are under-volted you are likely to burn out the motor because it won't get the current it needs - (remember V = IR? It's all related). The electric company does not vary the voltage because they would be liable for buying everyone a new fridge every few months.

The internet provider, however, doesn't care if it takes me 5 hours to download a file instead of 5 minutes. What they are getting away with that, unlike the electric company being liable for a burned out motor, the ISP doesn't deem itself liable for the extra cost to me for running my computer for 5 hours instead of 5 minutes.

However, I bet your cars speedometer displays that you can go "up to" a certain speed, but I doubt it actually could get that fast

The salesman never assured me that the car could go 220kph when selling me the car. I also expect the speedometer to indicate the correct speed when I'm driving, because the police will certainly assume that it is. You try the "my speedometer was wrong" defense in traffic court and see how far you get.

That ISPs use the "up to" line is nothing new, they've been doing it for years

That doesn't make it right.

Re:Sensassionalist reporting from ars

[Red_Chaos1]

"That doesn't make it right"

I didn't say it did. But ISPs aren't the only ones playing that game. It's a typical sales thing, and it isn't new. So why is everyone suddenly up in arms about it now?

Re:Sensassionalist reporting from ars

[guacamole]

Like I said, you can't compare the maximum throughput to a bag of chips (sorry, salary is the same). What do you guys come up with these stupid analogies? You can't read? You can't understand? Networks and other shared resources are different. Point. It's more like a highway. The government wouldn't let me drive faster than 70mph on the interstate I commute on, but there is plenty of times when my car is crawling at 20mph, because there is plenty of other people trying to use the road at the time time? One hour later the road clears up again. Understood?

Not everyone

[Tridus]

One of the reasons why I love my ISP is that this doesn't happen. I pay for a 20mbps down/5mbps up Fibre line, and I get that. Not just in speed tests either, downloading from fast servers (like Impulse, and Steam if you pick the right region) gets me that full out in real world usage.

Actually they bumped it up to 25 just recently due to competition, and I only noticed because everything I was downloading sped up. It's awesome. Hell, these guys don't even do traffic shaping.

(This is Aliant FibreOP, in Canada. Only available in New Brunswick right now.)

Only in USA?

[paziek]

Is it only in USA? I didn't R whole TFA but I think it mentions just them.
For example here in Poland, if someone would do that he would be most likely getting trouble from our own kind of FCC, just that ours has huge steel balls thats used to slapstick ISP no matter they size or whatever. Everyone kinda fears them. It could be fine if for some time speed drops by say 30-50%, but if it does that regularly (like in peek time) it might mean trouble for ISP. It will definitely be trouble if some1 with "up to 2Mbit" is getting average 1Mbit all the time.
Its considered more like "today there is transmission of some super important event and a lot of people will stream it and saturate our lines, so speed might drop for like 1-2 hours, but we explicitly said we can't guarantee speed all the time" safety.

I have had 2Mbit connection for 4 years and it NEVER dropped below 2Mbit for reason other than my destination. Its a standard consumer one with usual "up to" clause. I also work for a local ISP and I know they try their best to provide max speed for every customer, while business just gets higher priority in fixing stuff.

That's funny...

[EmagGeek]

I have Verizon Fios. I pay for 25/25 and that's what I get, if not more. There have been times when I've sustained 35Mbps for an entire Ubuntu torrent download...

I would imagine that with the ludicrous speeds being made available, most "slowness" is likely to be because of poorly configured customer equipment rather than wide-scale "lying" by the ISP.

Re:That's funny...

[TheSync]

most "slowness" is likely to be because of poorly configured customer equipment rather than wide-scale "lying" by the ISP.

The sad truth is that MANY cheap home routers can't keep streams of more than a few Mbps going. I needed a bunch of cheap routers on a commercial 10 Mbps connection, and I found I really had to shop around to find a "gamer" quality router to keep the TCP streams flowing near that speed.

In an age where you can pick up GigE Ethernet switches for under $10/port, it amazes me that some routers can't push TCP down the line at more than 1-2 Mbps.

Re:That's funny...

[EmagGeek]

That's a very good point. Most cheap home routers use slow processors that cannot handle the throughput being offered in modern Internet services. Fortunately, Verizon provides a decent router that can.

It's unreasonable to expect to get top-tier performance out of the bottom-tier hardware. Yet, that is precisely what most consumers expect. When they are disappointed, it turns into "fraud" or "lies" by the ISP, when it really is just consumer ignorance.

I'm reminded of lotteries

[ThatsNotPudding]

"Powerball is up to $64,000,000!!" No giant asterisk on the giant sign that you will only see about 45% of that money once all the taxes are taken out (double taxation as the voluntary tax has already been paid). When the idea of legalizing state-run lotteries in the US were being debated, David Brinkley remarked: "Organized crime already has a lottery; it is called the Numbers game. The difference being the odds of winning are better, and you don't pay any taxes on the winnings".

Oh go

[djdevon3]

That's because we're passionate and you're obviously NOT. Where would we be if Christopher Columbus said "Yeah I'm fine with the world being flat", or if JFK said "Before this decade is out we won't put a man on the moon because it's too far and the task too hard" or if John Carmack was content with 8 bit graphics.

We don't just want more speed. We want progress and at a fair price. Because these huge companies don't have adequate competition they're happy to stick us with higher bills and stagnant technology. It's obvious you don't understand technology so please put something against your head whose mechanics involve a firing pin and "make it so number 1".

No; we get 'up to'

[ThatsNotPudding]

the advertised speed - for the first few seconds (Speedboost, anyone?). Akin to incredibly loud local TV commericials that merely have to be at an averaged sound level of the preceding program, thus allowing massive peaks.

Huh?

[MBGMorden]

I hate the ISP's as much as anyone else, but how did this make the front page? In this particular case 6.7 was clearly advertised as the MAXIMUM speed and these guys are complaining because their AVERAGE speed is lower?

News flash: maximum and average are two different concepts.

Look at Sweden

[rundgong]

Haven't you learned yet that when it comes to broadband you should always look to see what we socialist bastards in Sweden have done :-)

Here it is mandatory to advertise speed intervals that show what you can reasonably expect. One of the biggest DSL providers (Telia) currently sell three packages: 1.5-2 Mbit, 6-8 Mbit and 12-24 Mbit.

That being said, it is of course a problem to correctly advertise speeds that greatly depend on factors that are out of the ISPs control. If we are talking DSL the quality of the copper cable and also the length of the cable (the location of your house) are huge factors in determining the maximum speed you can get.
I also think a general increase in technology awareness has made most people aware that just because they advertise 12-24 Mbit, it does not mean you can actually buy that subscription. If your house is in a remote area maybe you can only get the 6-8 Mbit package.

Penny Arcade Reference

[rfelsburg]

Mandatory: http://www.penny-arcade.com/comic/2009/5/01/

The whole thing is frustrating

[sarbonn]

I live in one of those areas where we've been stuck with low levels of Internet ability. My housing complex subscribes to Suite Solutions, which has been one of the worst Internet services I've ever had, at prices that are comparable to really good Internet provider services. It rarely works, and on weekends, it just goes out from Friday until some time on Sunday more times than not.

My alternative that the housing complex recommended was AT&T. Yeah, I know, but all commentary aside, it was what I was left with as an option. They signed me up for the medium tier of service, which was supposed to be 1.5 MB for speed but has never been faster than 500 kb ever. Turns out, they weren't able to set up the medium service, but just never bothered to let me know. So I've been stuck at 500 kb for the last year.

A few days ago, Clear moved into our area, and now I'm able to get 5.5 MB as a constant, and it's advertised at 6, so it's not that bad of a claim versus result.

But each one of the crappy services claimed to offer so much more than they delivered, and they have zero remorse for what they do. And they never will.

In Canada...

[da_guy2]

I know bell advertises "up to" 7 mbps but the technology only goes up to 6 mbps.

eh

[buddyglass]

What really ticks me off are the misleading claims about being "faster than the competition". Time Warner is always claiming that its cable modem service is "four times faster than AT&T Yahoo DSL". Listen to the disclaimers, though, and you realize they're comparing the base cable modem service (6 Mbps) against AT&T's low-end service (1.5 MBps) without mentioning that the 1.5 Mbps service is significantly cheaper. AT&T also offers a high-end DSL service that is (drumroll) 6 Mbps and approximately the same cost as Time Warner's base package.

Bah! Meh...

[genw3st]

It's quite obvious that most people replying to this post didn't read the actual article.

Yes, we all know that "up to 1,000 Mb!" internet service is a joke, and most people don't see performance anywhere near what is advertised. As some people have already pointed out... "DUH". It's biased advertising, what do you expect - that's how advertising works. Secondly, the article explained several reasons that people will not see the entire 6.7 Mb in terms that even the least tech-savvy person could understand.

I agree with the FCC's assessment. Essentially, they summed up the fact that it's pretty shady advertising, but ISPs are not doing anything illegal. The proposed solution is a new type of speed rating, similar to a nutrition facts label, which I think is genius.



On another note - I think it's pretty simplistic and idiotic to compare the facts and issues related to this topic to, say, restaurants. The article mentions how ISPs have little power over what happens to their service farther down the land lines - which is a fact. They have no control over interference, poor quality routers, etc. You can't make a comparison between that and a cheeseburger. The restaurant has complete control over their ingredients until it reaches the customer's hands physically, so no, you cannot make that comparison.

Interestingly, you might draw the analogy that McDonald's offers a quarter-pound patty on their burgers. This is typically the pre-cooked weight, meaning that there's a lot of liquid included in that measurement. In addition, you might order a double quarter pounder with one patty being less than a quarter pound, and the other being more than a quarter pound. Who flippin cares, as long as you get similar sized patties?

The moral of that story is... don't eat at McDonalds. Fix your own damn burger, and stop complaining!

Missing option...

[denzacar]

You DO get 6.7 MBit/s - connecting to your ISPs servers and network.

But from your ISP to wherever... well... they can't really vouch for that. So they don't.
And they put a clause saying exactly that in your contract.
Then again, most people don't read contracts.
Just think of all those EULAs you've OKayed over the years.

Re:Missing option...

[adolf]

*shrug*

I get 12Mb/1.5Mb from my ISP* over VDSL. It works fine, and it's what I was sold.

Some sites are slower, some are faster. A fast site has no trouble saturating the pipe with a single TCP stream, though, in either direction.

Torrents are fast, too. There's the whole "using so much bandwidth in one direction or the other that an ACK can't get through," but with reasonable settings in uTorrent, things are reasonably fast.

I call that good enough: By all appearances, my entire ISP's network is geared to support my high-speed whims, at least to their border links. What happens beyond that? Who cares -- it's the Internet. (Every time I run traceroute to diagnose a slow transfer, I find that it's not my ISP's fault, but someone far downstream.)

*: AT&T Uverse in NW Ohio, for whatever that's worth. It's completely likely that YMMV, and that things will be very different with this provider in different areas, in which case you're quite welcome to bitch all you want.

The purpose of Up To

[b4upoo]

The whole idea of using "up to" in a description is to lie and misinform. This is the very type of gimmick that makes businesses all look evil.

Re: The purpose of Up To

[EmagGeek]

No, the term "up to" is to account for the fact that most people are litigious assholes that will sue the moment they get 1bps less than the "advertised" speed without it.

The average consumer (and about half of Slashdot, it seems) is ignorant of the technical limitations involved with providing network access.

Complaining about "up to" speeds is akin to suing Netgear because you don't get exactly 100Mbps between your laptop and your file server when copying your files. There are several other factors than simply the line rate that need to be considered, such as whether the consumer is using a router that can handle the throughput (and I imagine most don't because they don't want to pay for it).

I like that/ need a math nerd

[zogger]

Someone with ties to Washington (legislative process), or a mass letter/comment writing effort to the FCC and see if a regulation would be in order to address this issue. They sure as heck always take the full price for the service. Now they want caps plus speed limits, etc, fine, there should be a mathematical formula they are required to use when the charges go out every month. Taking into account speed plus transfer. (seeking math nerd input there for this formula)

    And if they ALL had to follow that formula, every ISP, maybe they would be more interested in honest advertising and actually improving infrastructure. This would also address the issue the ISPs have with bandwith "hogs", their term. The more you get, and the faster you get it, you pay more. Less, pay less. About the same as any other utility (sort of).

Re:Loophole

[clone53421]

This is clearly dishonest, inaccurate advertisement, yet it could say "Up to 10000000000000000000 Mbps" and dozens of dorkos will still chyme in with "but it says up to! Hur hur hur hur!"

The problem isn’t the “dorkos” who know that their “up to” claim means utterly nothing. The problem is the morons who thought it actually had some significance in the first place!

Up to 100% of people will disagree with this post...

Re:median and the mean

[clone53421]

Then the Median would be 2, and the Mean would be 3.4

So the average of them would be... 2.7?

(whoosh)

Obligatory Penny Arcade

[Maarx]

The Fivefold Mother

Mine doesn't

[TheTick21]

My cable company advertises up to 50Mbps and in every test I've done I get around 50Mbps. There is a cap, but it is very reasonable and you can buy more for a reasonable amount as well. Very happy so far.

Zero is a number

[karlandtanya]

"Up To" means absolutely nothing.
If it's less than that, they told the truth.
If it's more than that, you're getting something for free. Where's the damages?

Note that the price you pay is never "up to". That's always "starting at"...

1200? Bah. Whippersnapper

[Petersko]

"I launched my first BBS at 1200 bps and at the time it was a pretty good speed"

Hah. My first BBS was on 300 baud. I could read faster than the text could scroll.

(Queue the acoustic coupler fellows)

Re:1200? Bah. Whippersnapper

[greed]

Since 300 bps was also 300 baud, and the coding was very, very simple (four tones, one and zero for each of originate and answer), you could push it to 500 or 550 baud on a really clean line. Somewhere between 450 and 500 I started to need the pager to slow it down a bit... by 1200 bps, 'more' was mandatory.

Those were the days when, if you had an interactive 'chat' with someone, it was the sysop. 'Cause with only one phone line, the only other keyboard was the one on the PET running the 'board itself....

I started with direct-wire modems, but did use an acoustic-coupler device later, in 1990: it was the ordering terminal for a hardware store. You went around the store with this hand-held gizmo, typing in SKUs and quantities or scanning barcodes off the shelf stickers. Then, when you were ready for the weekly order, you strapped it to the telephone receiver and pressed the Big Order Button.

It was great; head office got electronic orders, and the Mom & Pop franchise stores didn't need to have a full-fledged computer terminal, modem and all the other stuff to go with it. The order pad would probably have been needed anyway, and so they just put everything in that.

Been around for ages

[Jorl17]

In Portugal, you basically know that 16MB means 8 MB after 16^2 hours on the phone with four people!

as advertised here

[astar]

I am rural and still happen to have fiber to the house. The claim is 20/5. I could get higher. I only checked once. I did the check with some prominent test your speed site. I suppose it might be accurate. Got 21 up. Hah, got a static ipv4, but I run ipv6. You should have it as good. Realize this is a coop telephone system and not exactly driven by "share-holders value". Ah well. I am hardly committed to some fox tv brand of capitalism or obama brand of bail out the speculators. I figure if people are sane, about any system of economic superstructure will work. But capitalism is pretty good when you want small and nimble, pushing on the edges of the possible.. But I am not sure that a coop is really a capitalism thing.

Silly help desk people figure this must be some big name branded service. I guess everything has to be a brand name to work, It is true there is sort of a regional cooperation going on here, but the local guys have been burying fiber for a decade.

Perhaps all the slashdotters in a state should get together and start doing isp coops.

in other news...

[smash]

... arstechnica misunderstands the term "up to", and confuses it with "mean".

Anecdote

[ThePhilips]

Reminded of the joke:

- Is the Intenet available in heaven?
- No.
- Why not?
- Because all ISPs burn in hell.

FCC is in it for political gain. Biased!

[Shompol]

FCC analysis shows that the median actual speed consumers experienced in the first half of 2009 was roughly 3 Mbps, while the average (mean) actual speed was approximately 4 Mbps

FCC has not been mandated with controlling the internet yet, they have no business measuring the speeds. They want to show that there is a problem to get this piece of pie: something important to regulate.

ISPs aren't lying, they are giving the wrong stat

[davidwr]

"Up to" means peak instantaneous speed, not peak sustained speed or typical sustained speed or even typical peak speed.

If for any single second this month I get the "up to" advertised speed from my ISP, my ISP is not lying. It's being very deceptive, but it's not lying. There is a difference, even if it's a hypertechnical, er, I mean, hyper-greedy one.

MTBF

[C10H14N2]

I just arbitrarily pulled up the specs on the Seagate Barracuda ST3500630AS 500GB SATA Hard Drive. Its Mean Time Before Failure is 700,000 hours. That's EIGHTY YEARS. They have an annual failure rate of 0.34 %. Are Seagate lying to you when they say your hard drive probably won't fail until sometime in 2090, yet your data center just had three drives fail within six months?

Now, that's a fairly isolated piece of hardware where you could rightly expect that the reported stats given compliance with environmental limits should be fairly independent. Yet, not only do people "get" the fact that, no, their hard drive probably won't be happily spinning into next century, they /plan/ on it knowing that at one end of the curve, it is /expected/ that some portion will perform far, far less than the mean.

Your network connection failing to deliver even an arbitrary /average/ speed as you connect to Ouagadougou /obviously/ is subject to many, many external considerations. I would expect all spittle and histrionics from your average schmo on the street, but to have a bunch of dweebs pretending they don't get it is pretty silly. Yeah, if your connection is consistently falling far short of spec, fine, investigate, complain, switch, whatever. But this OMGCon$piracy b.s. is just stupid.

And how do you know that?

[SanityInAnarchy]

The fiber only covers the last mile -- they could just as easily oversell the same upstream connection that they oversell with Cable, DSL, and everything else.

Mine's like this too...

[sean.peters]

I have Verizon FiOS service, and it's advertised to be 15 Mbps. I've checked it numerous times with Speedtest, and every time I check, I get... 15 Mbps. So they're not ALL "lying".

Storewide Sale!

[phorm]

Storewide closing sale! Up to 60% off!

Which almost always means: 60% off the crappy product that nobody ever buys (or a pair of socks, whatever) and 5-10% off the overinflated prices of anything else in the store...

So yeah, it does happen fairly regular in other industries.

Puffery vs. Fraud

[Steve Hamlin]

There is a legal difference between puffery and deceptive advertising.

One shouldn't trust that "fantastic" means the same thing to both the buyer and seller, but one also shouldn't be allowed to deceptively mislead in commerce - that's called fraud.

They don't lie, they just don't know

[MaQleod]

I work in support for an ISP. The problem is that many ISPs deal over copper and copper is unstable. Two circuits on the same DSLAM card, provisioned for the same speed, say 6.0 DSL, may give one customer 5.7 mb/s (some loss allowed for packet overhead) and the other 2.7 mb/s. This doesn't mean the ISP is cheating the customer who gets 2.7, it just means that the copper pair that that customer is on is incapable of supporting faster speed, usually because the margins (signal to noise ratio) are too low because there is too much resistance on the pair (either due to distance or degradation of the copper). They say up to so and so amount because they don't know up front what the state of the copper running to each and every structure is.