Tuesday, June 15, 2004
To see what is in front of one's nose needs a constant struggle
There is an article today on ESPN.com by Rob Neyer titled "Differential Equation -- The Yankees may have the best record in the 'real' standings, but they're nothing to get geeked about"
Thankfully Neyer's been moved to ESPN's paid content so I'm not even tempted to read his Bill James re-treads any more.
In this case I am assuming that Neyer is using some version of the baseball pythagorian theorum which uses a teams runs scored v. runs against to "scientifically" determine what its win-loss record should be.
I am also assuming that according to whatever sabermetric he's using, the Yankees have more wins then they "should" have.
Here's my problem:
You have a theory. The theory says X should be happening, but X is not happening. Y is happening.
Do you:
A) Consider that your theory may be wrong in some areas
or
B) Deem the data "bad" and stick to your theory ???
Yes, sometimes bad data happens -- I fail to see how 2 1/2 months of baseball would be bad data, but I don't know that much about stats so I'll have to reserve judgment there.
Is Neyer saying that the 50 plus games we've played is too small a sample? I don't know. I do know that he's an arrogant little prick and I don't like him.
Here's my theory as to why the Yankees' record may not be what the forumla predicts:
MARIANO RIVERA.
You're going to win a lot of close games with him in the pen (Flash Gordon too), which will probably skew your predictions a bit.
The Yankees also have a lot of come from behind wins. I know sabermetrics people don't believe in clutch hitting, but fuck them too.
|
Thankfully Neyer's been moved to ESPN's paid content so I'm not even tempted to read his Bill James re-treads any more.
In this case I am assuming that Neyer is using some version of the baseball pythagorian theorum which uses a teams runs scored v. runs against to "scientifically" determine what its win-loss record should be.
I am also assuming that according to whatever sabermetric he's using, the Yankees have more wins then they "should" have.
Here's my problem:
You have a theory. The theory says X should be happening, but X is not happening. Y is happening.
Do you:
A) Consider that your theory may be wrong in some areas
or
B) Deem the data "bad" and stick to your theory ???
Yes, sometimes bad data happens -- I fail to see how 2 1/2 months of baseball would be bad data, but I don't know that much about stats so I'll have to reserve judgment there.
Is Neyer saying that the 50 plus games we've played is too small a sample? I don't know. I do know that he's an arrogant little prick and I don't like him.
Here's my theory as to why the Yankees' record may not be what the forumla predicts:
MARIANO RIVERA.
You're going to win a lot of close games with him in the pen (Flash Gordon too), which will probably skew your predictions a bit.
The Yankees also have a lot of come from behind wins. I know sabermetrics people don't believe in clutch hitting, but fuck them too.
|
