- Posts: 1982
- Joined: Wed Dec 11, 2013 11:55 am
Main Line Expos wrote:"I'm not a statistician but I've managed to do a couple regressions in Excel (I have trouble with LINEST sometimes though, can't always get it to work). The one I can think of off the top of my head is I regressed how different parks affect the run environment. So I start with the average run environment for a given cap/DH, and then I'm able to tell it my park (and my average away park) and it adjusts the run environment up or down accordingly. Surprisingly BP singles have a slightly bigger coefficient than BP homers, which I never would've guessed. I then use that run environment to customize my run expectancy matrix and wOBA coefficients. Then I use those to get my WAR values. Identifying which players correspond to good teams would be an interesting regression experiment but I'm not sure it would be more precise than the values I get from just counting the cards using my version of Pelletier's system.
In truth there's a lot more I could be doing with the data. The main things I've used it for so far are determining run environments for different parks and caps, and determining popularity of players at each cap, i.e., what % of leagues is a player rostered in at a given cap. But the player card values are just derived from Diamond Dope."
Really interesting, I appreciate you sharing. Are the T-stat significant in your regressions? I would assume 500 is enough data for there to actually be a meaningful relationship.
I really like your concept of identifying which players at which caps have been used the most. That seems like a very simply and intuitive approach to narrowing down the player set. I suspect that would work on both the hitter and pitching side. I would think the next level would then be to maximize those players by the proper park. Again really interesting.
As I mentioned early, without delegating this to an analyst it will have to be my retirement project. Do you use Python to scrap or R or some other language. Curious if you like one better than another.
D
Darn it, I thought the secret was drafting Bobby Wallace.