from Amazon
Friday, November 21, 2008
Run-based similarity scores
Great work… to which I disagree. Pizza Cutter did similar work based on rate stats, to which I have lots of comments on his thread. My key point is this:
If you are interested in looking for similar players to Vince Coleman, you may insist that the speed components (3b per 2b+3b and sb per sbOpp) be weighted much more than you otherwise would, because you are really interested in the speed players mostly.
So, in a run-based system, the speed components simply won’t have much differentiation. However, since we know the speed is strongly tied to SB, and speed is such a huge component of a player’s skillset, I would heavily overweight that in terms of trying to find similar-style players. Same deal for HR. Perhaps this is best exemplified with the K, which is very close in run value to the typical out, but clearly, there’s a huge difference in a hitter with 180 K and 40K. Basically, the more the component tells you about the player (rather than how much runs it’s worth), the more you should weight it.