Saturday, July 5, 2008

Lies, damned lies, and statistics

"There are three kinds of lies: lies, damned lies, and statistics." The semi-ironic statement refers to the persuasive power of numbers, and succinctly describes how even accurate statistics can be used to bolster inaccurate arguments. (from Wikipedia).

I love statistics. Statistics can tell you a lot about teams and individuals in sports, but especially in baseball. Baseball is driven by statistics.

If you follow college baseball closely you've probably heard of Boyd Nation. Boyd is cited frequently around NCAA tournament time as he has developed a model that very closely resembles the RPI used to select teams for the NCAA tournament. His Web site also has various other rankings and ratings that are supposedly statistically sound.

I'll be the first to admit that I am not smart enough to understand all of the complex factors and equations that go into Boyd's determining his various rankings, ratings, and odds. I've argued more than once that some of these didn't make sense, only to be told I was too stupid to comprehend his complex models.

Perhaps my stupidity is the reason why I found it interesting that as the 2008 NCAA Division I playoffs began Boyd graced us with his probabilities of each team winning their regional, super regional, and the CWS, writing that Fresno State had a 6% chance of winning their regional, a 1% chance of winning their super regional and a 0% chance of winning the CWS.

Fair enough, except for the 0% chance of winning the CWS - they were playing - they had a chance no matter how small it was. But Fresno won their regional. And there was Boyd writing that despite the fact that Fresno was now one of 16 teams left in the tourney they had a 0% chance of winning the CWS.

Finally, when Fresno reached the CWS (after winning a super regional they were given literally a ZERO percent chance of winning) Boyd graced the Dogs with a 3% chance of winning it all.

We all know what happened. Fresno defied the odds and won it all in Omaha.

Were the odds very long? Absolutely. Were the odds 0% in the regional and super regional? That's ridiculous. Hocus pocus.

Sometimes statistics lie.

No comments: