Data should tell a story

Data isn't useful because it's fun to look at. It's only useful, when it is used to tell a story. In turn, that story affects how people make decisions. Otherwise, it's just math.

http://www.intelligententerprise.com/showArticle.jhtml?articleID=196602724

Why I Don't Care

The recent spat of the AIG bonuses serves as an excellent example for why I think tools like the Google Visualization Toolkit API are important. Here are the basic facts. AIG received something on the order of 170 billion dollars from the US government. That is A LOT of money. They then paid executives, including executives who helped cause the mess, 165 million dollars in bonuses.

In was, politically, stupid as hell. In the middle of a financial crisis, you don't reward failure. In fact, I don't understand why a company would ever reward failure, but that's not really the point.

My point is that this got BLOWN wAY oUt oF PrOpOrTiOn. Click on the convenient picture to get a graphical understanding of AIG's financial malfeasance. A mere 165 million is a laughably small proportion of 170 billion, although it doesn't look like it until you stick in all the zeros.

Bailout - 170, 000, 000, 000.00
Bonuses - 165, 000, 000.00

I'm not going to sit here and defend the logic of rewarding failure. It doesn't deserve the effort. But, the resulting hoopla has been equally outrageous. To help us all understand this, let's break this down into numbers we have all dealt with. If I give you 100 dollars because you are in financial trouble, I am not going to worry too much if you spend 10 cents on a piece of bubble gum at Wal-Mart. The bubble-gum isn't actually necessary and is probably not a good use of the money I lent you, but it is nearly the exact same ration as the bonuses paid by AIG. It is literally pennies on the dollar.

For the Precision Nazis - I rounded. It would be more accurate to say that it would be as if you spent 9.7 cents of the $100.

Data Presentation - The future is the web.

After watching a recent Hand Rosling video (this guy is incredible), I started thinking more and more about how we use data to inform our understanding of the world around us. Before reading any further, you gotta watch this first:

Hans Rosling Shows the Best Stats You've Ever Seen

There are two important things that I want to focus on in his presentation, and neither of them have to do with his actual message regarding our understanding of the "Third World".

1. The presentation is NOT boring.
2. Because it uses dynamic graphics (video).

Rosling does a good job keeping his presentation engaging and focused. He does this by being an entertaining presenter and using rich multimedia presentation methods. This is not another boring Power Point presentation. Instead, it uses complex graphics to make it easy to see the relationships and ideas that he thinks are important. Furthermore, you don't really need Rosling to tell the story. Because the technology he uses can be embedded into any website, there's no reason why his presentation could not be prepared and presented by a website, in an entertaining and engagin manner. The only thing you would lose are his fun jokes about the state of medical students and faculty as they relate to the understanding of chimpanzees.

Compare his presentation to a typical government report, written by a consulting company like the one I work for. Sure, we have graphics and data in our report; but it is not engaging like Mr. Rosling's presentation. In fact, our report can never be as captivating as Mr. Rosling's presentation because our work is always designed to be presented on paper. Paper is not a dynamic medium. We can show two dimensional graphics, but showing how relationships change over time would require the use of 3-D graphs, which tend to confuse most people. Rosling's presentation is incredible simply because you don't have to be a quant to get something out of it.

Fortunately, this software was deceloped in an open manner and is now being made available by Google for everyone to play with. The software is called gap minder. To play with a version similar to what Mr. Rosling was using follow this link. To learn more about the current underlying Google API, see this link.

Not that I need "Yet Another Project" but I really want to look at this some more. I am a real nut when it comes to data analysis. Typically I use tools like R, PSPP, etc. Although these tools are good, I have typically used them to present data in a traditional chart or graphic on a piece of paper. But, both of these tools (especially R) are much more flexible and could be used to prepare an analysis that is more dynamic. Plus, programs like R and PSPP provide a way to test hypotheses, which I presume the Google Presentation API does not. (I could be wrong about this, I haven't looked yet.)

I think technologies like this can be and SHOULD be used to engage the general public more in data-driven debates. Often these debates are dominated by a simplistic understanding of the numbers, driving the discussion in directions that are not always in our best interest. Numbers do not lie, people do. (In fact, people use creative numbers to lie.)

More thinking must be done. And, in a moment I will give you an example of how the public can be easily led astray by faulty numerical analysis.

Ctrl <-> CAPS

I just swapped the keyboard mapping for my Ctrl Key and CAPS key.


If you understand what that means, save me. If you don't, just toss me some pity.

But, I have to admit that I am happy to welcome my new emacs overlord. . . . . . .