Slashdot Mirror
MSN Search Engine Favors IIS
Scud writes "It appears that if you want to rise up in the rankings over at the MSN search engine you would do well to host your page on IIS. Ivor Hewitt has done a study and it appears that by using IIS, you are likely to increase your odds of a higher listing by several percent."
...such anticompetitive behavior from Microsoft?!? Wow. Who'da thunk it?
Paleotechnologist and connoisseur of pretty shiny things.
who would have thought?
Argh, just when they removed it from Visual Studio requirements, it's skewing up search results. MSN Search should be banned for being dishonest.
It's clearly biased towards Internet Explorer too, the results I get back in Firefox are mostly irrelevant blogs and pages full of adverts.
"Those who cast the votes decide nothing; those who count the votes decide everything." (attrib. Joseph Stalin)
So what's going on? I have no idea, I doubt it's all a big conspiracy... but some possible explanations spring to mind: Perhaps the MSN search has simply been coded by developers used to talking to IIS machines and so it just does that job better? Perhaps the MSN spider is taking advantage of some specific IIS features to provide enhanced indexing?
In other words, there are some explanations out there other than "MS is biased and there's a conspiracy and they are trying to take over the world"...
One man's Funny is another man's Offtopic.
... to think ms wouldn't use all it has. Obviously it hasn't yet learned from google, that being evil is bad. And bad guys get punished.
Is there any truth to the rumor that having a picture of Bill Gates on your site makes you #1 in your category?
I'm a big tall mofo.
And who is the silly person who would expect it to be otherwise? Have you actually been listening to the news at all over the past decade? Have you learned nothing? The real story would be if the ranking did not rise if it were housed on an IIS server. Otherwise it's a nothing, I would have assumed that.
2. Why would i want to install IIS, when i have a better alternative Mac OS X?
3. Why would i bow down to Satan, when he's dumb, stupid, slow and...
"Doing what i can, with what i have." ~ Burt Gummer
... and they still think they can beat Google to the game. When are they going to realize that what made Google so successfull was the fact that is has been so unbiased in all ways imaginable, including not accepting payments to get higher rankings.
Google makes money by prioritising quality. Microsoft makes money by prioritising money.
Go figure.
...you gotta do something to pump up your buggy, non-mainstream, insecure webserver.
Yeah, right.
My page is titled "San Andreas Radio" and if you Google it, comes out #1 or #2 every time.
MSN it and it comes out about #7. Either they're being paid to reduce its rank (it's a bit subversive), or they don't like the fact I'm hosted on Linux, or they simply don't have a very good search engine.
If I put the exact unique title of a page into an engine, I expect that page to be #1.
so that MSN _thinks_ you are using IIS; look at the agent and send a different server name string
Mirror of Ivor Hewitt's site
I think I have never used MSN search in my life. I suppose other people do, but how many? Anybody know MSN search share percentages?
San Francisco Photographers
The control over what webserver you will use is typically limited by your hosting provider. While many provide the choice between Unix-based servers and Windows-based servers, many do not.
For those who use hosts that do not provide these services, I don't think it appropriate to think that they are simple SOL. Rather, the better quality your website provides, the more relevant it is to the topic you discuss, the better it will fare in any search engine. The type of webserver you are using becomes nothing more than the tiniest fraction of your search ranking.
It just happens that people use "is" a lot in their search queries. Perhaps Apache should rename itself to "The Apache."
Whose the most evil this week? Up-and-comer Google, or Microsoft, the old standby?
Though seriously, scummy as this is, it doesn't surprise me.
"The Sage treasures Unity and measures all things by it" - Lao Tzu
I have stopped using other search engines since i have found google.
if it favours iis machines, it makes it that much easier for virus writers / script kiddies to play about with them if it displays them in preference to other web servers.
To be conclusive, it needs to be a controlled experiment with the same text and same outgoing/incoming links.
Just the webserver alone changing. This can happen by taking a popular site and then changing what it reports to the MSN search robots.
But until such an experiment is done, the data is open to too many interpretations.
The link to MSN search on the main story links to beta.search.msn.com. It should be noted that MSN Search is out of beta for a while now - the correct links should be http://search.msn.com. It's not like it's Google or something - trying to keep everything in beta for years to escape criticism.
It's really unlike MS to skew things towards their products... I'm sure it's a mistake or a "Linux Zealots" distorting the facts...
My site is first or nearly first in google using relevant search terms. But in MSN it never shows (even if listed). Maybe also the use of PHP is harmful for MSN ranking? M.
For years now, the company where I work has had all it's Apache systems reporting that they are IIS 5.0 systems. Just a quick change in a single file before compiling and there you go!
think back 1995 Bill tells us MSN an not the Internet is th future. I will not use it and my webservers will not run IIS. my 0.2 cent
Microsoft has its own search engine? When did this happen? This is the first I heard of it. I have never heard one of my friends say, "Hey just MSN Search it!"
or its just not been indexed by the msn bot yet.
Now I have a reason to run IIS instead of Apache.
- Crow T. Trollbot
and this is "news", why?
Well, MSN never promised they wouldn't be evil.
So it seems fair to me.
Just change the server response line if the GET or POST comes from Redmond, WA to say you are some version of IIS. I can already see the recommendations coming from the SEO folks.
Nobody uses MSN, anyway. Or, at least nobody I know...
If you disagree with me on social issues, then it's pretty clear that you are a narrow-minded bigot.
Surprise, Surprise! Who could ever imagine this possible?!
"...Suggesting crappy Microsoft based products to Idiots of Tomorrow, Today!"
Apparently it's a very slow news day. In the interests of being remotely on topic.. (yes my karma will suffer dearly for this)
Why would this be any real surprise to anyone? MSN being MS is obviously going to give preferential treatment to their own products. This may be by design or strictly because IIS servers respond to some proprietary (yes I said it) requests that other servers won't.
I don't necessarily see it as an evil thing, but it's not entirely philanthropic either.
The world according to SComps
It looks like your post isn't running on IIS.
What you are saying is Google is still the best search engine out there?
Thanks, MS, for clearing up that confusion.
Seriously, it's like these guys are out to do as much evil as they can get away with.
Mod me down with all of your hatred and your journey towards the dark side will be complete!
...to use the Google search engine. I mean, sure it's expensive to use MS's engine, but by paying that extra bit to use Google, I get less evil results. And no ads, placed by people that are paying Google, either!
Give me a damn break. It's MS's own freakin' web site, and they can do what they want (assuming this study is even really exposing causation, rather than correlation). No one is paying MS to be neutral in any way, any more than they are paying Google or Yahoo, each of whom tilt results according to all sorts of standards, ideals, and sometimes total capriciousness. Now, lower ranking for sites that can't figure out the difference betwen "its" and "it's" would be worth paying for.
Don't disappoint your bird dog. Go to the range.
One of our clients has a web app hosted on an IIS box and their main website hosted on apache. The web app ranks higher than the main website when doing a search for them.
Excuse me, kind sirs, but MSN Search is in beta. I'm sure if you calmly file a bug report, they will address the matter before MSN Search is formally released to the public.
Honestly, the thought that anyone would rely on a beta service...
For more information, click here.
Is this really any different than being able to pay Google or Overture money to put your page near the top of the page? Why not get a preference to people who use your product?
I have a technical websites with lots of visitors. I don't need MSN. I'm going to block MSN spider from my website to make MSN much more inferiour. Everyone should do this.
Did anyone anywhere claim that their results WOULDN'T do this? That they would be the premiere unbiased source of accurate search results, playing no favorites? Nobody forces anyone to use it, so don't.
"I'm a leaf on the wind. Watch how I soar."
-Hoban Washburn
compared here
I knew something was up! When I search on "warriorpostman" in Yahoo and Google, my personal website (running on Apache) shows up, but the only thing that shows up on MSN's website are my Slashdot comments!?
I suppose it's not proof of a conspiracy, but still...
IIS stands for "Invasive Internet Search"
Please note:
Before you configure IIS to run 32-bit applications on 64-bit Windows, note the following:
IIS only supports 32-bit worker processes in Worker Process Isolation mode on 64-bit Windows.
On 64-bit Windows, the World Wide Web Publishing service can run 32-bit and 64-bit worker processes. Other IIS services like the IIS Admin service, the SMTP service, the NNTP service, and the FTP service run 64-bit processes only.
On 64-bit Windows, the World Wide Web Publishing service does not support running 32-bit and 64-bit worker processes concurrently on the same server.
After configuring IIS to run 32-bit Web applications on 64-bit Windows After you configure IIS 6.0 to run 32-bit Web applications, IIS stores 32-bit DLLs and ISAPIs in the %windir%\syswow64\inetsrv directory. All other IIS files, including the MetaBase.xml file, are stored in the %windir%\system32\inetsrv directory. File access to the System32 and sub directories are transparently redirected based on the bitness of the process making that file access (64-bit processes have full access, while 32-bit processes have access to System32 redirected to Syswow64). If your legacy applications have specific 32-bit file access needs and you notice application failures, see if the application needs to reference the new %windir%\syswow64\inetsrv to resolve the problem.
That's all it takes to resolve the problem!
He who knows best knows how little he knows. - Thomas Jefferson
My site is first or nearly first in google using relevant search terms. :P
You mean if you type your sites url into google
... what an awful fucking post. I hope your horribly transparent attempts to curry favour with the "slashdot crowd" get modded into oblivion.
Don't they realize that Google became the top search engine because the results were good? If they mess with results then people won't use MSN Search, and then the bias won't have the desired effect on web hosts. Its really a win-win for the anti-MS crowd that they do this.
Is change your webserver's outgoing HTTP Header for Server to
HTTP/1.1 200 OK
Server: Microsoft-IIS/5.0
and let everyone think you're using IIS.
There are 2 kinds of people in this world. Those that can keep their train of thought,
Unrepeatable
Ill designed
Speculative
Not exactly scientific, is it?
Did you read the article? Oh wait, this is slashdot and you're an anonymous coward, of course you didn't.
The methodology and the word lists are available, and the author has stated that he plans on making the perl scripts used available, so the experiment is repeatable.
It also seems reasonably well designed and not particularly speculative.
At some point, somewhere, the entire internet will be found to be illegal.
The problem is that, if you the client (the searcher) search on a website that gives emphasis to hits that have paid you in some fashion, that's dubious. One of the early Google tricks was to get the sponsored links out of your face and denoted as such. In this case, however, it's actually worse*, because the sponsored links are being unfairly weighted for something utterly unrelated to content.
* Assuming it's true. This claim seems rather strong. Microsoft has no reason to do this without publicizing it ("use our products and be higher ranked"), and publicizing it would destroy their credibility. I personally believe that there is a better explanation than the one we are left to infer.
Microsoft is in a weak position in search, but not so weak that they would need this kind of duplicity. Why, when they can make themselves the default search engine of so many, would they go through the effort of making a wacky backdoor that will only bring them bad news? I'm not buying it.
Who Cares? at least it still shows sites that don't use IIS. Now if it only showed sites running IIS, that would be news.
Most of those useless keyword, domain parking/hijacking, and spam sites out there run on Linux+Apache because the owner can host thousands of those domains fairly inexpensively, and that's the key to all spam: minimization of operating expenses so you only need 1 out of 100,000 users to click/buy to turn a profit.
These sites don't have any real content, they just point to other sites and/or exist to spam you with advertisements. Some of them have googlebombed their way higher into the rankings.
My guess is that MSN does a slightly better job of filtering those useless sites out of the index at the present time, OR the "googlebombing" techniques they use aren't as effective with MSN's indexing. Since they almost exclusively use Apache that would have the false appearance of favoring IIS.
This is just a guess, but it seems plausable.
Natural != (nontoxic || beneficial)
That Google might *gasp* be biased towards Apache!
...Googled for anything using MSN!
If you disagree with me on social issues, then it's pretty clear that you are a narrow-minded bigot.
None of my pages have any problem ranking well on MSN Search and I've never served a web page off of IIS. What defines "several percent" anyway? If someone forges their web server header information only to gain a 5% chance of increasing their ranking, I hardly think that's worth it. In my experience the MSN search engine seems to value keywords, page titles and the number of pages that link to the target page much more heavily than the type of web server. I'm not a fan of MSN, nor would I advocate using it on a regular basis.. but this type of story seems to be a bit of anti-MS FUD if you ask me.
shop.envescent.com - Computer hardware and more.
that I would chalk it up to the MSN beta search engine just plain sucking.
;)
Or at least the criteria it uses to determine matches is not as strong as it should be.
What interests me is that apache is still the top web server
"Fighting the underpants gnomes since 1998!" "Bruce Schneier knows the state of schroedinger's cat"
Correct Netcraft is, that is a large amount of webservers that are going to get little or no ranking on MSN because of their choice of Webserver.
/. article stating Ballmer giving a 90% discount on goods/services if the Germans didnt go to Debian.
Apache 41819229
Microsoft 12420068
Sun 1836275
Zeus 610819
Regretably this seems to be the case with MS, giving preferential treatment to those who will use their products. Id have to cite an earlier
"God of Rock, thank you for this chance to kick ass. "
The same was mentioned like 3 months ago... I don't even bother to look for previous /. link.
Dear Microsoft Employees: It has come to my attention that some of you are under the impression that there is another web server other than IIS. Please know this is not true. It is about as silly as this rumor that there is another browser besides Internet Exploder and another email program besides Outlook and Outlook Express. The EVIL Linux / Open Source movement has it out for us. However, upon review of the data collect from our super-worm, it appears that Linux and Open Source may too be a big hoax. We can find no evidence that either one exists. So, please continue on making your software as insecure and unstable as possible. Our marketing records indicate that such "features" such as the "Blue Screen of Death" actually cause people up upgrade their older versions of Windows to our newer, more intrusive versions. Thank you for your hard work. Until we rule the world, Bill Gates
I'm not a troll, but I play one on Slashdot.
One thing I've noticed is it skips results from time to time... makes it really annoying when you're trying to check your positioning.
For example, I do a search, and on the first page it shows 1-9. On the second, it shows 11-20, on the third it might show 21-27. It is TOTALLY annoying.
As far as I can tell, this ISN'T due to sponsored links.
On the other hand, we use apache, and we're doing well, but that might be an exception.
Cheezit! We're boned! - famous 31st Century bending unit
My site that I run from home is Apache on Debian, and Microsoft's MSNBot hammers the shit out of it daily. It accounts for over 15% of the hits I receive. Go figure...
It may be ill designed, but it's up yours to redesign it
It's not speculative but observative (gather facts, trace graphs)
And it's damn not unrepeatable, the whole polling process is explained and the scripts will supposedly be made avaible (even though any half decent coder should be able to pull out that kind of scripts easily in the language of his choice)
Does sound scientific to me boy
"The way we can tell it's C# instead of Haskell is because it's nine lines instead of two." -- wadler
Tha article actually links to an older smaller version of the analysis. There's a more comprehensive wordlist at: http://www.ivor.it/goog
There is likely to be some bias due to self-selection - site owners who run on Linux/Apache are less likely to go and add themselves to MSN search.
Let's try this one. Google uses Linux is biased against Microsoft-based IIS servers!
:)
Oh wait..I forgot, we love Google and hate Microsoft.
Amazing how two sides of the same coin can be overlooked, isn't it?
I quick search for "linux" indeed confirms the ordering assertion made for this example for the search engines.
But what is more interesting is that Google found about 299 million hits, where as MSN found over 336 million hits. I say, Whaaaa....??
It must have been something you assimilated. . . .
conspiracies require more than one party
So you're saying that "government conspiracy" should actually be "government policy"? Interesting.
I'm not a Microsoft fan by any stretch of the imagination. So I was pleasantly suprised to find that my sites get good rankings for some searches where Google puts them on page 999, way below lots of totally irrelevant stuff.
BTW, did anyone consider the possibility that Google is prefering Apache sites, rather than MSN being the baddies?
is when all the extra traffic from higher rankings crashes the IIS servers that much faster!
Reject Fear - Embrace Hope
I would like to be able to recognize it if it spiders my server.
Banned from life is my vote.
If it can do what it likes then so can we. Everybody who is put off by M$FT's behavior should open up a hundred or a thousand hot mail accounts, and subscribe to every spam list there is. Subject these accounts to every bad practice regarding smap avoidance. Fill up all of M$FT's hotmail drive space. I bet it won't take more that a week to bring hotmail down this way.
While running the MS Antispyware scan, it picked up that my host file may have have been tampered with. That's great stuff! Only problem is that I set up my hosts file to send most ad producing sites to 127.0.0.0. Only redirection they complained about was that ads.msn.com was being sent to 127.0.0.0 instead of the holy land of banner ad goodness. And to give fair credit, I get my hosts file from http://everythingisnt.com/hosts.html
I traded all my mod points for these magic beans.
Remember, the guys working on the MSN search engine certainly use IIS to host their intranet sites, and whatever internal webservers they use to test against are probably IIS as well, at least in the most cases. They are likely to consider bogus results for their own sites (both internal and external) more critical... that's not malice, that's just human nature. Even if they consciously work against that, they're more likely to notice problems there first.
And search engine tweaking is more an art than a science. It's an evolutionary process, with feedback loops and strange attractors. So if there's any difference in the behaviour or design of Apache or IIS that would be visible to a search engine, it's likely to lead to a slight bias in favor of the server software that the servers they pay more attention to run.
Ok, maybe it is just by chance this is happening.... At the moment, no-ones really bothered, as Google has the market...
.. But, what if MSN search had the market? Let's face it, The way IE is setup, new mom and dad users will end up using MSN to search, because it's what IE takes them to should the page they typed in not be found. There is a good chance that, like it or not, they could become the search engine to be seen on in a few years, regardless to what Google do. Of course, I guess it would be not too difficult for Microsoft at this point to turn up the bias algorithm, forcing the world to adopt IIS..... Yes, there would be lawsuits, but, from past experience, this wouldn't do much.
The question is, what can we do about it? It seams to me, that Google needs to do the same NOW, ie, any sites on IIS, to be page ranked down, (pushing the market to dump IIS), and do this until Microsoft decide that there bug (or hidden agenda) is bad for profits, and repent. They may not have the ability to do this in a few years.
I did a search for /. and quickly got slashdot, which does have pics of BG's true form.
I prefer the "u" in honour as it seems to be missing these days.
But you do choose your hosting service. There's a lot of competition out there. You can choose what server your website is running by choosing the hosting service that runs what you want.
Ask some one who knows me ... I AM
Am I really that bad
-Gowan
MSBPodcast.com The opinions expressed here are my own. If you don't like 'em... Think up your own stuff.
Leads me to think: is it significant? That is, can we exclude (to a reasonable certainty, that is, p>0.95) the possibility that the effect seen cannot be attributed to chance or some other criterion MSN uses?
Ivor says at some point The initial set of words indeed showed a significant difference between the results from Google and the results from the Beta MSN search..
But what does he mean? I would be interested in what kind of significance test was applied, what the exact results were. Just looking at the ratio of percentages doesn't tell me enough... One should go back at the original data (seems provided, good) and check if the effect is actually trustworthy or just, in Ivor's words, "Odd. Pure coincidence perhaps."
Before seeing some analysis of significance, I don't believe anything...
The only difference in the HTTP response is just that IIS adds headers and that IIS has that stupid HTTP Continue on handling SOAP via ASPNET.
Just telnet to almost any Apache web server and type GET / and then to an IIS server and do the same thing. Look at the top. Almost all non-IIS web servers return no default headers.
Microsoft.com:
redhat.com
Moving it to apache is only natural if you're dealing with rational beings.
Driving on public roads and spending much time on the net both suggest there's a severe shortage of those.
The Fine Article states that while Google's results are comparable to Netcraft's server survey results (that is, their share of Apache and IIS represents the respective market share), MSN seems to favor IIS. So no, Google does not favor Apache.
No matter because GOOGLE will CRUSH you!!!! ARRRRRRGH!
the preceding comment is my own and in no way reflects the opinion of the Joint Chiefs of Staff
Who even uses MSN search? Even my most senior computer using relatives have turned to either Google or Yahoo without my prompting.
Luck favors the prepared, darling.
Have they gone ahead and implemented that thing about assigning you a hosting provider at birth then? What a shame. Back in my day, we used to be able to pick our hosting provider based on what they provided and what they charged for it.
Ah, the good ol' 1900's.
--MarkusQ
Sadly, this fails to surprise me. This news story doesn't sound too different from the one posted yesterday about Kerry-contributors being banned from some engineer gathering.... Why should a search engine give a flying fuck about what http server a box is running?
since they run a few of their own sites (or at least they used to) through akamai's proxy servers (running Linux) I wonder if they give these sites an exemption? Because let's face it, if detecting apache is -1 point, detecting Linux is probably -10...
TODO: 753) write sig.
Was your immediate reaction to fire up google and msn and compare them?
Mine too! Then I realized "Dang It! What a clever marketing scheme!"
What can I do (by negligence or by design) to get my business /.d too???
--- Attorneys Assisting Citizen-Soldiers & Families -
of Google -vs- MSN here. Just another datapoint in the continuing saga. Got your popcorn ready?
Clearly, people using IIS just make better websites.
I have a site that is about 10th on Google. Hosted on Apache and Fedora. But on MSN Search We're number 1! YEAH BABY!
Oh wait the site sells stuff for FrontPage.... well that blows that theory.
Maybe if you have a lot of MS positive stuff you get higher too...
I can't remember the story but I seem to recall that search results are not factual merely opinion and in the US protected by the 1st amendment.
Maybe google favors non IIS websites? I can't believe this guy didn't use yahoo in test as well.
Have you ever been to a turkish prison?
This just in, The Yankee Group uncovers the truth behind the PHP name, in a report commissioned by Microsoft (MSFT). Laura Didio is the analyst who made the startling break through:
"Over the course of a study into the roots of PHP, we discovered that several Psychology sites use the software. That got me thinking," She commented. "See, Psycology has a 'ssss' sound at the beginning, and yet it starts with a 'P'."
Soon after her statement, she put on a tinfoil hat and ran out of the room. Screaming "They're out to get me! God, save me from Groklaw! Pleeeaaase!"
So, what is the mysterious epiphany she had? "Pspawn Hof Psatan". In a related story, Microsoft updates it's Get the Facts (tm) campaign...
Ivor also provides us with a fascinating overview of his garden's trees
Nothing costs nothing
Nobody uses MSN search anyway. Most computer users I know use Google, and they range from geeks, to casual users, all the way down to those who think the Internet Explorer icon is "the Internet". MSN search is irrelevant and this is just another example of why it will continue to be well into the future.
You can have a "conspiracy of one" if that person acts in multiple roles.
As an example, let's say that one person is a company's bookkeeper and CFO. (This isn't uncommon in small companies.)
As a bookkeeper she cooks the books to cover her embezzlement.
As CFO she prepares false financial documents for her company and its investors.
One person, criminal acts in two roles, so in many states she can be charged with conspiracy in addition to embezzlement.
BTW, this isn't a "conspiracy" in the legal sense since it's not a crime to give preferential service on the basis of web server. It's sleazy unless it's fully disclosed, but it's not a crime unless they actually sell the search engine as an unbiased tool.
For every complex problem there is an answer that is clear, simple, and wrong. -- H L Mencken
We're talking about a convicted illegal monopolist who has no qualms about leveraging one of their products to give another a boost. There is no indication that Microsoft has changed its behavior since they got their little slap on the wrist. Assuming malicious intent behind all of their actions is simply the way to go.
In other words, there are some explanations out there other than "MS is biased and there's a conspiracy and they are trying to take over the world"... "
It's called plausible deniability. "Why, no, we had no idea this would happen. You say it's an interaction with an IIS feature that causes this to happen? Heavens to Betsy, we never thought of that."
Microsoft people aren't stupid, and they ARE trying to take over the computer world, or haven't you been paying attention to what they say and what they have done? The engineers that built MSN Search would certainly be aware of any interaction that fits with IIS features to provide enchanced indexing. They would have been all over it from the beginning. And a side-effect means that IIS sites come out higher? Great! It's a feature that benefits us, they would think.
Of course MS is biased. Of course they would have noticed this. Of course they like it.
I think that might work, but I'm not sure. I mean, it may be possible, but there may also be some roadblocks. I don't have experience in this area, but you might find something on Google. MSN might give you different results. I hope that helped, my apologies if it didn't.
Where's the paired t test?
It's just coincidence that there happens to be a bias that makes IIS-hosted sites measure higher by this metric. ;-)
As copyright owner of this comment, I authorize everyone to defeat any technological measure which limits access to it.
Apache freely issues advisories and patches. It will issue an advisory if even one user faces a minor risk.
Microsoft (and nearly all other proprietary software companies) tries to hide problems to protect their perception in the marketplace. You usually only see advisories for major problems that will become public knowledge anyway, and numerous other fixes are piggybacked on the big ones.
But beyond that advisories don't really address the quality of a product. They're one metric, but nothing more.
For every complex problem there is an answer that is clear, simple, and wrong. -- H L Mencken
Will some of the lone gunmen out there explain to me how, assuming this is some sort of ploy by MS, how it benefits them?
You can't advertise it e.g. "Use IIS and get preferential rankings in MSN search!"
It won't get more people to use MSN Search; if anything it will get less people to use it as they get less accurate results.
So..Who Benefits?
It's been out of beta for months. (Unlike, say, google groups, or froogle, or gmail, etc.)
Best Buy can have you arrested
6% differences across the board. Normally, 6% is something to be expected simply because it's using different search algorithms. Normally. Look at the distribution of specific words. It varies a fair bit, even occasionally with MSN returning less IIS based sites than Apache...although usually more. But it ALWAYS totals out to 6%. 6%. 6%. Why are all 8 averages showing a difference of almost exactly 6%? An even marginally random experiment ought to show variations, unless MSN itself is tweaked for 6%. Now, I'm not an "M$ hater" but it seems kind of odd that MSN would be set to favor IIS sites on average 6% of the time. It seems equally odd that the reviewer results would always average out to 6%, particularly given the small size of the dataset; that all 8 graphs converge to the same central value is odd, to say the least. I'm not going to suggest any particular theories on what's going on, but this is a pretty shoddy report, if you ask me. Hardly thorough or definitive. More like a guy in his basement looking for another reason to hate Microsoft. Kinda like a (stereo)typical slashdot member, come to think of it. So, I'm thinking to myself, what the hell is going on? Why do the results look this way? I graphed the alternate's detailed results (quick excel graph with sort, nothing special, will make available on request). One graph of the difference in Apache results, one graph of the difference in IIS results. And well, the results are cubic. No question about it. R squared values of .982 and .973.
Looking at the data, it becomes apparent that there's one single search result which skewed everything, and that's the search for "plexiglas". Drop off that result and now things are starting to look just a little bit inconsistent. Flip back to the original article and notice a small handful of similar results that are also off the graph by a fair bit.
Six Percent. The nearly cubic curve, the constant 6% differences. I'm going to stop here, but please people, just think about this. Something's strange here, to say the least.
Gates favors Windows
"... likely to increase your odds ..."
to have the possibility to maybe increase the chance of an opportunity to perhaps increase the odds of raising your rank?
Thats a pretty definitive article
then maybe the other search engines should favour Apache & Linux [eg] Alta Vista, Excite, google, Yahoo etc...etc... if Microsoft wants to burn bridges & make themselfs an island then let them, Microsoft is small when compared to the rest of the world & the WWW...
Politics is Treachery, Religion is Brainwashing
Why not just change the server name in the http response header.
----- If communism is a system where the government owns business, what do you call a system where business owns govern
Then, in 2007, came Longhorn, with integrated web search using not Google, but MSN. Joe Sixpack didn't care, but MSN was so damn convinient he forgot about Google - effectively forcing Google Inc. with its costly development department out of business. Later - oh surprise - all results you got for "Linux" on MSN were advisories to ditch it for Windows. He who controlled the search result, controlled the industry. (Maybe I should put some fake Frontpage-Meta-Header to my webpages to increase Rankings on MSN ... just to be sure)
Screw the FSM - Real geeks believe in the Invisible Pink Unicorn
You are quite right. I wasn't implying malice. I was just pointing out that there seemed to be a weird minor statistical anomoly.... and I was curious what was causing it.
If you look more deeply, you will see that he ran another test using 1000 words and the results were almost identical.
Now, to fully understand if that is significant, one needs to know the variance within the results, but still -- 1000 results is pretty good number.
The real "Libtards" are the Libertarians!
One of the cool differences between my engineering stats class and my management stats class was the way statistics were explained. In engineering, they said pretty much what you just said...
But in management, they said "You have to weigh three factors: your experience, the gravity of the decision, and the statistics. The propper way to interpret a statistic is: The data, all by itself, makes me XXXXX suspicious." You replace XXXXX with a little (70%), somewhat (85%), very (90%), extremely (95%). I think this is a cool way to look at it - and it allows us to Bash MS with bad data, based on historical information...
while (sig==sig) sig=!sig;
aol mail on the web switched to IIS aspx
At the same time you are controlling your content and changing only the web server that serves it up, all of the sites you are ranked against may potentially be updating their content, switching web servers, changing hosting providers, etc. The sites that link to you and all of your competitors may be changing as well.
to search? I wonder what percentage of people uses MSN search anyway?
Believe it...
First off, I looked at the difference in means for Apache rankings in MSN and Google. 61.5% (MSN) vs. 64.3% (Google) for 970 observations Right there, you ought to be able to eyeball it and see significance. But, to make sure, here are the results of a t-test which checks the likelihood that two matched sets have different means (forgive the crappy formatting):
M G
Mean 0.615061856 0.642948454
Variance 0.01100624 0.008740111
Observations 970 970
Hypothesized Mean Difference 0
df 969
t Stat -10.51551356
P(one-tail) 7.26569E-25
t Critical one-tail 1.646427658
P(two-tail) 1.45314E-24
t Critical two-tail 1.962415113
As you can see, the P is 1.45 x 10^-24, which at least makes us think the results are not pure coincidence. I don't intend on speculating on the causality, though...
You could just change the HTTP Server header that Apache sends out. Someone should try it for a few weeks and see if it really makes any difference.
If you have mod_header installed, just add the below line to httpd.conf:
Header set Server "Microsoft-IIS/6.0"
In major countries like Germany, IIS is already down to around 3% of the server market. Even world wide, most people have the sense to run Apache. You can look at the percentages, but every time an IIS farm is rolled out, shortly thereafter, they wise up and drop it for Apache or any other product actually suited for being connected to the network.
Frankly, I'm not sure why this article even made it to Slashdot. Is slashdot or OSDN participating in this year's marketing tsunami by doing product placement ads? Please let's go a week without MS articles, there's enough shilling going on in the discussion without them.
Beta is broken and the link to classic doesn't work. Stop wasting our time or there won't be anybody left here.
It was also proven years ago that Internet Explorer displays pages served by Microsoft IIS faster, by tinkering with TCP/IP.
Does anyone use MSN search?
I know that I have not yet event tried it. Google works and has become a household name like Band-Aid. Heck it has actual become a noun meaning to search the Internet. If give biased results it will just push more people to stay with Google.
See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
I'm sure it was not intentional. I suspect that a few percent is much less than they intended.
1. MSN sucks, who uses it anyway.
2. IIS totally sucks, who uses it anyway.
All my sites have this in robots.txt:
/
User-agent: msnbot
Disallow:
This is mostly because their robot has been abusive in the past, but the idea of rendering their searches a bit less useful for folks who could benefit from the information posted (the sites are mostly informative, not sales pitches) only makes me happy.
And no, this has nothing to do with Microsoft paying Ralph Reed $20,000 a month for advice on kissing the Religious Rights' ass. It's just a way of saying they can kiss mine for free.
"with their freedom lost all virtue lose" - Milton
Who the f*** uses msn search?
While this currently does Yahoo and google, it would be neat if it includes MSn at some point.
m l
It's a neat way of seeing how the two engine's compare.
http://www.langreiter.com/exec/yahoo-vs-google.ht
"Not knowing when the dawn will come, I open every door." - Emily Dickinson
First i'd like to point out that the average of the "Diff" % column for apache is -8.30% while the average for the "Diff" % column for IIS is +4.48%. This means on average MSN returned about 8.3% fewer apache sites than Google, but only 4.5% more IIS sites. That leaves a 3.8% loss.
But thats enough on the numbers, frankly I don't think they show much at all...let's ask a few questions...
Did this guy:
Check any of these IIS websites to see if they were using robots.txt to block (or partially block) Google?
Or stop to consider and investigate whether or not IIS webmasters slightly favour MSN search for submission?
Or test against the other top 4 search engines, other than Google?
Or factor in technologies like the tendencies toward ASP, Java servlets etc on the win32 platform over PHP, Perl etc on linux and how these effect spider performance?
Or use more specific topical search strings and compare relevence by reading the pages to prove less relevent/worthy pages were being bumped up because they are using IIS? (i.e. a human comparison rather than just comparing against other possibly biased engines)
This is interesting on first sight but really doesn't show anything (yet?).
For a few more reasons.
Ok, so from now on my webserver returns MS IIS when msnbot visit's my system, instead of the normal Apache.
I think that it's a little paranoid to believe that the convention white=good and black=bad has to do with race. Unless mass PC paranoia has changed the meanings of those terms for you. (Kind of like the word niggardly has no race at all in it's background, but people still get upset over the racism of the word.)
If you were serious about wanting more/better research you'd do some yourself. The analysis isn't that hard (again, if you're serious about wanting it).
If, however, you just want to be whiney, you're doing fine.
This reminds me of that scene in Austin Powers (Pt.1) when Scott Evil reminds his father Dr. Evil that he can easily shoot and kill Austin once and for all rather than keeping him alive.
"Scott, you just don't get it, do you?" Then he's left hanging with no explanation.
Why do people still use IIS/Microsoft-centric web technologies? It's almost if people like making things harder on themselves. They should just stick to what *they* do best: OS's, office, and video games.
I'm tired of being a Scott. Right...
I know Google rocks and all, but how can we be sure that Google doesn't favor our favorite open source http server over IIS? I mean really, that's the most unscientific research garbage I've ever read. Thank you /. for killing my last 3 bain cells.I know Google rocks and all, but how can we be sure that Google doesn't favor our favorite open source http server over IIS?
/. for killing my last 3 brain cells.
I would be more inclined to think it's Google's searches that are actually skewed, because they have been perfecting their results for a lot longer and everyone knows the sites worth deserving a top 10 position, run Apache.
I mean really, that's the most unscientific research garbage I've ever read. Thank you
RTFA! Google and MSNS aren't the only search engines he tested - they are just the only ones worth talking about. So you would have to conclude that /everybody/ is discriminating against IIS.
Add something like this pseudocode to your server:
if $Browser = "MSNSearchBot" then $Server = "Microsoft-IIS/6.0"
One line blog. I hear that they're called Twitters now.
Mr. Hewitt doesn't seem to have performed any kind of statistical analysis...therefore while his "findings" are interesting (and worthy of further investigation), they don't conclude anything whatsoever.
Off the top of my head I would say some kind of ranked-sum test (Mann-Whitney, perhaps) to determine if the distribution for either search engine differed significantly from the other would be appropriate in this case. But it's been a while since I took stats.
Did he change his web site? What I had said: "This is the full list of the current search words/phrases used by the validator" and he gave some 20-30 searches. If you use 1000*K observations and you get the same results, it's a different story.
lol, you have to read what my response, was in response to. My response was supposed to be satire, not bigitrous in itself. I am sorry if you took offense, non was intended. I was attempting to point out what you stated in plain words, the connotation that these things have racist meaning s is utterly ridiculous. Anyone who tries to paint them that way is searching for a fight, and they know it.
Microsoft favors IE.
And ZiffDavis, C/Net, Didio, Foley, etc keep saying something about a kinder, gentler Microsoft. Right.
LoB
"Anyone who stands out in the middle of a road looks like roadkill to me." --Linus
I think the quotes were important, try it again. While it comes in at #1 on Google, it comes in at number 2 on MSN (oh the horror!) if you don't count the 2 advertisements above it (which I don't think that you should).
well, we use linux servers so I just checked some of our keywords. We are #1 at times, on the first page with others, and off the first page with a few more. And this is all without optimizing our pages to try and do better at msn.
So either everyone in our keywords are using linux or this is full of MS hatred gone too far.
The real "Libtards" are the Libertarians!
Before seeing some analysis of significance, I don't believe anything...
You realize that this is no place for calm, rational thinking, don't you? Maybe you're new to Slashdot. If so let me provide some assistance with your response:
1) Anything Microsoft does is bad
2) If you think they've done something good, or you can't tell whether something they've done is good or bad please review (1).
That should help you fit in around these parts a little better. There are some nuances that you'll learn over time, but I've covered the basics. If you have difficulty with these concepts, please visit room 101 for assistance.
-Adam
searching for my site name (terrbear) gives 2 pages of results on MSN. more than 10 on google.
best college pickem site ever: pickem.terrbear.org
Perhaps other factors are more significant in your case.
My site is in Italian, therefore google's result may be affected by language options. If I search "numismatica", "monete italiane", "catalogo di numismatica" using google, my site is always in the first positions:
& me ta=
http://www.google.it/search?hl=it&q=NUMISMATICA
conversely, on MSN it does not appear in the first 20 pages. But it has been indexed, in fact if you search the exact name it appears.
M.
It might not be signifiacant, but at least it may give ideas to some slashdot reader of the msn search dev team.
I have discovered a truly marvelous proof of killer sig, which this margin is too narrow to contain.
When I learned that Microsoft was trying to muscle into the search business, I added a robots.txt rule to exclude them from my web site. I figure that the less useful their results are, the harder it will be to do evil.
I could imagine (some) other Apache users like myself doing something like that, but I can't imagine any IIS user doing it...
Slow down a second.
First, anytime you see P 10^-24 for a sample of only 1000 and means differing by 3% you should be suspicious immediately.
And 969 degrees of freedom?!
Buddy, your stats are waaaaay off. There is just 1 DF in this data.
My guess is that you tried to feed categorical data to a t-test thing and it barfed (as it should), so you fiddled with it until it claimed to produce a successful result (by somehow misinterpreting each of the 970 observations as a degree of freedom) and now you've got P 10^24, which is lunacy.
thumbs down on your methodology.
Microsoft is a company, the first founded the company, then they got into different projects that gave them money. If tomorrow the production of musical shoes probes to be a better market for Micro$oft, they would switch to it.
In the case of google it's diferent. There is a project, there is an specific research, first of all, then, they built a company arround it. That's the difference beetween them. Google is a group of scientists that built a company arround an specific development. Microsoft is a company that has as only purpose to make money, selling whatever shit happens to fit the market today.
So, don't be surprised when m$ products are biased, low quality, etc, because their purpose is not to develope quality software, but to just make money. OTH Google has as a purpose to develope quality search technologys, and then, make money out of it.
WTF am I doing replying to an AC at 5 A.M on a Friday night?
While I agree that Microsoft demonstrates and has continued to demonstrate some seriously anti-competitive practices, this study simply shows that the results differ for web server types between the two search engines. It doesn't demonstrate anything beyond that. While extending the comparison to 10,000 incidences from the current 1000 will add a finer grain to the results and help to eliminate the problems of small sample size, the study's author needs to add other search engines to the study to provide some control for the comparisons. If multiple search engines show one result and one shows another then there's evidence for a problem, but right now it merely shows that the distinctly separate systems used by Google and Microsoft produce distinctly different results. This shouldn't be surprising, or even notable.
(YAWN) Oh, cool! More evidence M$ is Evil! I don't mean "evil" as a descriptor, I mean Evil itself!
Coudn't you infer that the results of this study show that Google favors apache? They are in compitition with Microsoft after all. I am sure that both search engines are using different algorithms, why assume that Google ranking is truely "correct". Why not study a greater number of search engines and see if all show a bias one way or the other.
gods forbid MS starts making cars, lest we forget...
h tml m ade_cars.html 1
http://www.ehumorcentral.com/Directory/Jokes/829.
http://www.jardmail.co.uk/misc/mscars.shtml
http://www.jokesunlimited.com/jokes/if_microsoft_
http://www.maxwebportal.com/article_read.asp?id=2
etc. etc.
(though if you look at the last one, there's an entertaining response towards the bottom...)
Though isn't this a bit more like the Phone Company only telling you about Chevron?
-- it's ridiculous how many people misspell ridiculous... (damn, damn, damn...)
somone has to. makes sense for it to be those guys...
http://milkshake.dexy.org
Where is it written I have to perform whatever statistical analysis you deem necessary before publishing results?
I simply found something curious and wrote about it. Big deal.
If someone wants to spend the time doing funky heavy statistical calculations, then they can do it.
It's even longer since I took stats, and I didn't get a very good score.
I'm still a bit skeptical of this analysis. It may be the type of thing which doesn't lend itself well to T-tests, or that search engines are complicated enough that butterfly effects cause massive statistical differences for no real reason.
For example, from yourr data I find that relative to Google, Yahoo is weighted in favor of apache by 66% to 64% and that's also highly significant at 10 ^ -13.
I haven't checked Teoma's difference of 68% to 64% but I'm sure there's a p even more astronomical there.
So if we believe your analysis, that MSN is weighted in favor of IIS is no bigger news than that Yahoo is weighted in favor of Apache (assuming Google is some kind of gold standard).
Point being, these data need to be approached with some caution. It could be that any given search engine has a huge chance of being statistically significantly different from every other engine.
If that's true, then the chance that MSN is statistically in favor of IIS is..... 50% ie a Flip of the coin.
Well we know what the next windoze iis trojan is going to use find its targets.
I can see the slashdot future - "trojan uses msn search for iis victims." Sweet.
Send Peter Clifford Francis Macrae comdoms to 23 Bedford St, St.Neots, PE19 1AX, England
Maybe you should RTFA...
From the article:
"I decided to try and analyse the results produced by Google and IIS and compare them for bias."
"A perl script was used to perform searches against Google and MSN and scrape the results."
No other search engines are mentioned in the article.
Your mind looks a little cramped. Why don't you stretch it a little?
A perl script was used to perform searches against Google and MSN and scrape the results...
Where does it say anything about other engines?
Search for Canadian Airborne Regiment on either MSN or Google and my PHP site comes up number 1, so I expect it is just because of different ranking algorithms that yours does not.
The reason Microsoft only got slaps on the wrist is ironically because judge Thomas Penfield Jackson could not hide his contempt for the defense, and therefore the case was effectively dismissed on the grounds of a biased judge.
That incredible fuckup and all its implications pain me to this day. To think, that the evidence was so clearly not in Microsoft's favor that the judge was openly (and unfortunately, vocally) hostile towards the defendant, so much so that the decision for the plaintiff was tossed out... The landscape of my industry might look a heck of a lot more interesting now if it wasn't for that.
* The penguin is the cutest mascot ever,
* and the Sky is still blue
Yep, MS is still a monopoly and acting as such! Never saw that one coming... :-)
If you think ANY search engine is 'fair' then you are nuts..
No law stating what has to show up in the results..
Sure its nice if they are honest and upfront, but they dont have to be.
---- Booth was a patriot ----
My site is a webcomic/rant site that uses some fairly... colourful language. Apparently MSN search has a porn filter, so we get all sorts of traffic from porn searches that didn't turn up any porn. In fact, MSN search is our number 2 referer, beating out every marketing campaign we've ever done.
If you search MSN for things like "anal fucker", "hardcore sites", or "why is leah remini fat now" there's a good chance UAC will be right there on the first page. And our site is PHP and Apache all the way.
Bite the hand.
I have to note that these allegations that IIS hosted sites rank higher in MSN search aren't good at all -- as far as I can tell, believing that MSN tweaks its results in favor of sister products and companies taints the search results to me.
In other words, now MSN search has, for me, become an inferior search tool.
Do I believe that MSN has fudged things in its favor? Yes. Might believing that encourage me to switch to a more neutral search engine? Yes again.
It looks to me like Microsoft took a gamble on this one, supporting IIS at the risk of alienating users if they got caught, and they got caught.
In Soviet Russia, us are belong to all your base.
One of the best way to test this would be to take a page (create a page with some stuff that would come from a search with no existing results. For example, "The Slashdot I Love Microsoft and Hate Linux Page") and put the same HTML file on an IIS server and on apache. Submit the site to both google and MSN, submitting the apache one first then followed by the IIS one) and see how the pages rank. This could confirm whether one is prefered or note.
Now, is there anyone out there with an ISS server interested in trying this out?
Check out http://www.tnl.net/blog
Try typing "online music".
On Google the top two references are iTunes and iTMS. On MSN you'll have to go through a few pages before you'll see anything about iTunes.
Yeah, I trust Microsoft to provide unbiased search results. Sure I do.
m.m.
2/3s of the sites Google takes me to are powered by Apache. ;)
In other news, Windows is found to heavily favor MS executables as opposed to others.
Here's what MS public relations had to say:
"We wanted to guide people to sites that would represent Microsoft positively, and--call us nutty--it seemed like those who want to kick Microsoft out of town would have some difficulty doing that," says Microsoft spokesman Drent Tuffy.
m-
You catch enchiladas by picking them up behind the head and holding them underwater until they don't kick anymore -VeGas
Yeah the trouble is that the person who submitted the story linked to the old "original" result set i.e. the "/orig/" in the url rather than the more complete more recent results at: http://www.ivor.it/goog
I guess the "MSN against Google" report is more attention grabbing.
Incidentally, Google has its own financial incentive to be biased: to direct traffic to sites that have Google ads.
Thanks for the link to the original. However, now I'm even more convinced it's nothing! Look at the variation between the four engines: the MSN results actually don't stand out, even though they are the lowest for Apache. For example, there is more difference between Google and Teoma than between Google and MSN. So, are we going to accuse the other search engines of manipulation, too? They exhibit the same level of variation from the apparently unquestionable Google reference.
I don't think MSN searching engines are SMART enough to recognize Bill Gates pictures on webpages...(when i did some searching with MSN search)
I only found pages(highly-ranked) that had "bill_gates" or "Microsoft_Owns" as prefixes
for their image and video files
buffering...
Why is this thing even posted? "Bad guy furthers his scummyness!"
Yeah, you have to look at it from his point of view.
"ALL MONEY IS MINE! IF YOU HAVE ANY, KNOW THAT I'LL BE COMING ROUND LATER!"
Yeah Bill, we know. The internet isn't big enough for your balls and anybody else.
MSBPodcast.com The opinions expressed here are my own. If you don't like 'em... Think up your own stuff.
This is why I still don't use anything MSN has to put up like this since I know the damned thing is biased 6 ways from Sunday.
:-)
It's another pointless tool for the Micro$oft Monopoly Machine.
Google and Yahoo Rule!!
All content in this message is copyright (c) 2008. All rights reserved. RIAA is prohibited here.
There's only one reason to favour IIS-based sites, and it's unquestionable, for any search engine wanting to make it big.
Obviously, the folks who build world-class, mission-critical websites on a webserver as prone to issues as IIS are gifted individuals, whose content is guaranteed to be superior.
Heh.. if they want to favour IIS-based sites, I say let 'em. It'll only seal their fate as a useless search engine.
Search engines can return very different results.
Here's what happens when you use the exact same search strings in a couple of different engines:
Search #1: oled driver
google: 1
ask: 1
yahoo: 1
msn: 3
altavista: 1
dogpile:3
MSN ranked my site lower than google, ask, yahoo, and altavista.
Search #1: pci express serdes
google: 3
ask: 120
yahoo: 9
msn: 3
altavista: 9
dogpile:6
MSN ranked my site lower the same as google, and ask.com ranked me at the bottom of the 12th page!
FWIW, my site is hosted on IIS.
I'm at www.nextsierra.com
I manage a few e-commerce sites and I see roughly the same types of percentages the poster above is quoting. For example, 90 day stats for one of my sites indicates 1% for msn as opposed to google which is 40% of my referrers. Now which search engine do you think I optimize SEO for?
Microsoft uses its own web server for it's search engine!!! Coming at 11:MS touts IE as worlds best browser!!!
I am Spartacus
do you know how to do this? please tell me how! :)
Just change the server response line if the GET or POST comes from Redmond, WA
What does the home town of Nintendo of America Inc have to do with anything?
Point being, these data need to be approached with some caution. It could be that any given search engine has a huge chance of being statistically significantly different from every other engine.
I agree with this...
I was merely pointing out that the results aren't due to chance-- the means are different with high statistical significance.
At an experimental level, all it proves is something that we already know, that there are two different methods producing the two sets of results, no more.
It definitely doesn't imply causality. In fact, I'd wager the difference is due to some other driving factor thats correlated with the server type than some sort of 'if server is apache, then reduce score by 5.'
But hey! ... ...
...... but if your ranking is your business, you need to rank in MSN; ... all that goes to MSN !
Recompile apache and change server version
use a proxy when msnbot comes
and you still comply with "show same content to bots and humans"
on the other hand: i do not use msn, beacuse others show better results usually
just think of grandma who never heard of google and just types stuff in IE
Try typing 'purchase music'. On MSN, iTunes is in the the first page of results, while on Google it's nowhere to be found in the first 5 pages. So you can take that however you want.
Fair point.
the filters. Keep you kids safe, use MSN.
Sorry about the writing. Robot fingers, you know? Cliff Steele in DOOM PATROL #23
- Gets robots.txt like a good spider
- Gets default.asp
- Gets / on the site
Hmm, interesting. So MSN Bot is looking servers running ASP pages.