Just cracked the new Ian Ayres book Super Crunchers. Interesting, if a bit oversimplified in the early chapters. In the tradition of a Freakonomics, Ayres is an economist looking at the mathematical/statistical explanations of things. In the early chapters, there seems to be a case made for the logic of numbers over (or as opposed to) the value of insights or intuition.
For example, it's implied that wine tasters are obsolete in the first 10 years of bottling, whereas a rain scarcity and high temperature formula is a great indicator. And seemingly, it is. But does that make the wine tasters obsolete in this phase? I think that's what the book is missing so far. The need to mesh insights with statistical evidence. Hopefully that's where the book is going.
We've all seen Six Sigma (aka "Six Stigma") and other systems of metrics gamed, abused, or simply overused. Is there a good system for integrating metrics with insights? Is there an organization out there that's doing anything remarkable with storytelling and econometrics?
Subscribe to:
Post Comments (Atom)
2 comments:
As soon as I saw what you were blogging about I assumed I would disagree with you, I was wrong!
Darn right there is a need to mesh insights with statistical evidence. No matter how much we know about anyone, by way of surfing habits or data, the human being is complex with 'feelings' that change on a regular basis.
I think any business person would agree that if success (financial inparticular) was as easy as analyzing stats, we would be reading your blog from our Yacht in Tahiti.
There has been so much great work done lately around complexity science, babs. My favourite resources come from Dave Snowden. Check this out, for example: http://www.cognitive-edge.com/articledetails.php?articleid=56
Snowden and others make an important distinction between complicated and complex, that you have in your lucid comment as well. Complicated is elaborate and tricky, but "knowable". A 747 airliner, with all of its bits and moving parts is very complicated. The airline staff though, is complex. Their behaviour cannot be engineered or predicted.
Any biological system or entity is complex. "Feelings" are a big part for humans. Mutations and other chemistry we don't understand also contribute.
I would challenge your last paragraph. I actually believe that Levitt, Ayres and others feel that it is as easy as analyzing stats. We're just not looking at the right stats.
Post a Comment