Slashdot Mirror

You know, I just did that search and the very first link was "The plural of anecdote is data, after all".

The author admits that he's been using the quote wrong all this time but that he, like some, will continue to use it wrongly because he simply doesn't want to change his mind based on his emotional attachment to the word anecdote. I found that ironic.

A huge problem at Microsoft seems to me to be that people there, or maybe just the leaders, seem socially unsophisticated. In fact, neither of the articles quoted below explains the underlying reason that Microsoft is buying Revolution Analytics. That needs to be explained. (All quotes retrieved Sunday, January 25, 2015, around 07:00 PST.)

In The Official Microsoft Blog there is a lot of corporate-speak, of the kind used by people with no actual interest in a subject who nevertheless want to be considered knowledgeable:
"find ... value"
"data-driven decisions"
"reduce the ... skills gap"
"enterprise-class platform"
"analytic solutions"
"advanced analytics within ... platforms on-premises"
"we are at the threshold"

From another article linked from that article, Revolution Analytics joins Microsoft, by "David Smith, Chief Community Officer":

"Microsoft might seem like a strange bedfellow for an open-source company..."

It was not a good idea to use the word "bedfellow". That word is more appropriate for a novel. The primary meaning of "bedfellow" is "a person who shares a bed with another".

'CEO Satya Nadella proclaimed "Microsoft loves Linux" '

On the surface, that makes no sense. Below the surface, is Microsoft trying to say, "We want Microsoft to be popular"?

"We're excited the work..."

That should have been "We're excited [that] the work...".

I'm not the only person who feels uncomfortable with those statements. One of the comments to that story is this one:

"What a joke. You're really working hard to try and convince readers that this is a good match, going on and on about how supportive Microsoft is of open-source. You were probably sweating while trying to come up with excuses as to why this is good, knowing that you were typing bullshit. I would suggest growing a pair of balls and just being honest, but I'm sure you've never had to do that in your career. -- Posted by: Anonymous | January 23, 2015 at 11:22"

David Smith replied to that comment: "Anonymous, I've never been anything but frank on this blog and this is no exception. I'm truly excited for the future, and I'm sure I speak for the rest of the team as well. -- Posted by: David Smith | January 23, 2015 at 11:25"

Sometimes the lack of social ability at Microsoft is shocking. The cover of the January 16, 2013 issue of BusinessWeek magazine has a large photo of Microsoft CEO Steve Ballmer with the headline calling him "Monkey Boy". See the BusinessWeek cover in this article: Steve Ballmer Is No Longer A Monkey Boy, Says Bloomberg BusinessWeek. The BusinessWeek cover says "No More" and "Mr.", but that doesn't take much away from the fact that the magazine called him Monkey Boy -- on its cover.

In many years of following such things I have never seen such disrespect of a CEO. Of course, whoever wrote the cover headline was merely repeating a common phrase applied to Steve Ballmer by people in the computer industry.

Worst CEO: Quote from an article in Forbes Magazine about Steve Ballmer: "Without a doubt, Mr. Ballmer is the worst CEO of a large publicly traded American company today."

Another quote: "The reach of his bad leadership has extended far beyond

The plural of anecdote is data: http://blog.revolutionanalytic...

No, it's anecdotes

The plural of anecdote is data: http://blog.revolutionanalytic...

The plural of anecdote is data, after all

Data _is_ the plural of anecdote.

Now, if you mean "improve sales"...

No, I mean efficiency. If your main business is sales, then it means improving sales without raising expenses, but Big Data analysis is useful elsewhere, too. If you'd actually read my comment without the gross prejudice, you'd find several examples of improving efficiency in medicine - more accurately detecting trends, showing comorbidity, and finding doctors that ordered excessive tests. Inefficiencies like that are hard to see individually, but in aggregate, outliers are much easier to see.

Some examples from other industries that I've encountered in research:

• Arranging small shipments on large transports.
• Determining which screenwriters wrote the best parts of a collaborative work.
• Turning off HVAC in unused offices.
• Arranging water towers in a city.
• Monitoring the growth and migration of endangered species

And taking that same person at the right time and showing them that generic X is equivalent to brand name except for the label and the much cheaper prices increases efficiency. Funny that. Not very "Big Data" efficient, though.

Well, yes, but that's rather unrelated to the advertising-industry obvious example I was talking about. Interestingly, that was one of the capabilities of our medical-data system. It differentiated generic and name-brand prescriptions, so researchers could compare efficacies.

Congratulations. You've proven correlation == causation. Oh, wait, no you didn't.

To quote XKCD: Correlation doesn't imply causation, but it does waggle its eyebrows suggestively and gesture furtively while mouthing 'look over there'.

Ten million people might have a cough because it's flu season. But, Joe may cough because he has lung cancer.

It seems you're suggesting that every unknown cough should be treated like it's cancer. That's exactly one inefficiency that raises healthcare costs so much. Out of fear (mostly of malpractice suits), doctors must order extra tests that don't even make sense, just in case the patient has some rare disorder. Maybe Joe also has a funny feeling in his chest that the doctor should know about. A Big Data system can highlight the correlation between cough, that odd feeling, and cancer, and lead the doctor to ask about the feeling. Then there's an indication for Joe to get the extra test, but not for the ten million with the flu. That's ten million fewer x-rays, and a few hundred million dollars in savings for the hospital, insurance providers, and patients.

So true. Too bad the plural of anecdote isn't data. So Big "Data" isn't inherently good for statistical analysis.

Well, the plural actually is data, but I digress. More accurately, data is a set of anecdotes with their subjective interpretations and biases stripped away. That makes them far more useful for statistical analysis, because they're now reduced to a mere set of facts.

Oh, and did I mention that real statistical analysis is really, really hard? I mean it-requires-a-human hard.

Just like playing chess, recognizing speech, driving a car, optimizing circuits, and all those other "hard" things that we now use computers for daily. The actual computations aren't really hard, but what is difficult is knowing exactly what computations are appropriate for the questions being asked. That's not a problem most Big Data systems even try to solve. Rather, they assemble data and prepare predefined statistics to make it easier for humans to get answers to their questions. For example, we can take a set of millions of medical procedures, and ask "which doctors order more tests than others, over one standard deviation from the average for their diagno

Many of us geeks have gone on to sire progeny

You must be new here.

But seriously, you're right that many Slashdot users have younger relatives in their mid-teens, such as my younger cousin. So we have your anecdote and my anecdote, and the plural of "anecdote" is "data".

As Raymond Wolfinger pointed out: "the plural of anecdote is data."

You misunderstand, you need a hell of a lot more than 1/2 a dozen anecdotes to prove anything.

One is that Mortal Kombat (2011) isn't ported to PCs, nor are most other games in that genre. In fact, the only fighting game series I'm aware of that sees regular PC ports is Street Fighter.

Nothing to do with people connecting PCs to HDTVs.

See also comments by kamapuaa, FunkSoulBrother, CronoCloud (again), and several more dismissing the likelihood of the public adopting home theater PCs as a viable alternative to traditional set-top consoles.

So you're quite happy to take the baseless opinion of a couple of randoms on the internet as 'evidence', wow, i pity you.

neither major video game publishers nor the general public have caught on that a PC can be connected to any flat HDTV.

Have you got any data to support that assertion?

As Raymond Wolfinger pointed out: "the plural of anecdote is data." One is that Mortal Kombat (2011) isn't ported to PCs, nor are most other games in that genre. In fact, the only fighting game series I'm aware of that sees regular PC ports is Street Fighter. See also comments by kamapuaa, FunkSoulBrother, CronoCloud (again), and several more dismissing the likelihood of the public adopting home theater PCs as a viable alternative to traditional set-top consoles.

Sorry to burst *your* bubble but your argument fails on several fronts:

1. There already exists a succesful support company based around R: http://www.revolutionanalytics.com/
2. The model of making money by providing BETTER support and releasing Free Software is proven by Red Hat, MySQL AB (pre-acquisition), etc
3. The OP doesn't sound like they need anything besides out of the box functionality, which is incredibly full-featured in R (especially compared to Excel!)

He said he wants something that is easy to implement, and only reason he is going with open source is because then he doesn't have to ask for purchase approval. Which IMO is a really stupid reason and will hurt in the long run - it's insane to take worse software just because you don't want to ask your boss if it's okay to buy this one.

Horse shit. I've seen projects die because they couldn't get software through the approval process. Better to try 10 apps that are free and run in userspace (so no need to get IT involved for an Administrator install) than to wait for management approvals, budget cycles, and IT support, and never get the project done. If I'd done that on the job, I'd have been fired for taking too long to do my work.

I also resent the implication the "free" means "worse."

Sorry to burst your bubble, but if you want good support and easy implementation, you have to look for normal paid-for solutions. Besides, open source is not synonym for free. This is especially true with specialized software or something you want good support for. Open source just means you get the code aswell, so you can implement your own additions (without use of plugins) or change it.

I'm guessing you haven't used R. Not only is there a thorough user manual, but there are books from most major statistical and instructional groups on how to use R, AND the R-help mailing list answers every R question I've ever had about it, AND there are local R user groups where you can get support similar to how LUG's work.

But unless you get an product from a company that is spending money to develop it, you never get good software and good support. No one can make both because everything in this world costs money, and developers have to live too. Open source and free software model works well for the likes of Google and Firefox because the developments get paid by money made with advertising. Statistical analysis software, and other specialized software is a different matter.

Please shut up. If your assumption were true, R would not exist. R exists, so you're just an asshat.

My advice to the original poster: Use R if you have any familiarity with programming. Any higher level math/stat course OR experience with basic programming will let you get started in R. If you've been doing this all in Excel already, you're probably ready to hop into R. If you're still uncomfortable, I'm sure one of the people who value your academic library could help out.

Why do you think R is not easy to implement? My company has been using SAS for a long time and we are finally making the change to R. As far as OP's requirements are concerned, I think R is way superior to SAS or SPSS because of its free, modular nature. It is clean, simple and suitable for a wide range of users. The commercial packages are filled with way too much business lingo garbage for me.

I personally think commercial support is overrated. I can install software on my own. I know how to browse through manuals and other information to find what I need. For a package like R, I almost always get any questions answered in at most few hours on online forums. So what exactly do I get from commercial support for my money? But, if OP needs commercial support, there is an enterprise edition of R by Revolution Analytics located here: http://www.revolutionanalytics.com/products/revolution-enterprise.php. Might be worth looking into.

Bottom line: R all the way.