Tuesday, September 16, 2008

Look at me, I'm doing research!

Somebody should start a pool on how long it will take me to start hating research. I'd say the over-under is about 6 days. But I'm loving it for now. I mean, I try and waste time on the interwebs like usual, but all the lies about "we had 23,000 people at our rally," "thanks, but no thanks," and "Saddam Hussein and his family had a personal relationship with al Qaeda and he's about to miniaturize nuclear weapons" (start about 2:30 in) make me too angry. Also, I discovered that I didn't have to learn Perl, re-invent the wheel, or pay an undergrad EECS major to re-invent the wheel for me--every baseball stat you could ever want is available in a free download at Sean Lehman's Baseball Archive.

So, for starters, using all the MLB MVP's from 1911-2006 (excluding AL pitchers), players batted an average of .285 before winning their first MVP award and only .271 after, and 74% of the MVP winners had a lower career average after winning than before. That, of course, proves nothing. But now I'm going to use propensity score matching based on MVP votes received (and whatever else I determine makes you win the MVP award) to compare the actual winner to the almost-winners, and then maybe I'll be able to say whether winning the MVP award makes you famous, gets you a date with Jessica Simpson, and then makes you ignorant/less good at baseball by association.

1 comment:

  1. i don't know if using stats all the way back to 1911 is fair--i mean, no only was the MVP not even given consistently until the mid-twenties, but the statistics that were valued back then as most valuable were obviously different than they are now. I mean, how else can you explain babe ruth's numbers in 1924 and see that he didn't even figure in the top 16 of MVP voting? He would be a unanimous decision these days with those numbers. so maybe you could split the data and figure out what seemed to be the statistics that were most determinant for MVP voting and how that's changed too--and then use those determinants to see how they changed pre- and post-MVP. just a thought.

    ReplyDelete