from Amazon
Monday, September 27, 2010
Secret Sauce? No more!
Colin is doing what’s right:
I took a look at how often the team with the best sauce won each series, from ’06 (the first season the Sauce was used) through ’09, and the result was only 54% success – not significantly better, statistically speaking, than flipping a coin.... It’s possible that the Secret Sauce ran up against a few years where its performance was flukily low. But what we have is a model based on historical data that has thus far been ineffective at predicting results out-of-sample, which doesn’t give us a lot of reason to be confident in it going forward. So for now, we’re retiring the Secret Sauce.
I think Colin should have been more forceful and acknowledged that regardless of what the Secret Sauce said, it should have been retired. As MGL noted, how does completely ignoring a team’s offense supposed to be a good thing? The best you could have hoped for with indicators like this, is that the win% would go from .500 (all other things equal) to .510, maybe .520. With such little out-of-sample data to work with, you could never achieve anything close to a reasonable uncertainty level.
Anyway, I love that an outsider like Colin is fixing right what is wrong.
Recent comments
Older comments
Page 1 of 344 pages 1 2 3 > Last »Complete Archive – By Category
Complete Archive – By Date