THE FUTURE SEASON FORECASTER

This is something I've been kicking around with Sal and Seung for awhile now-- trying to come up with our own predictive tool, like Baseball Prospectus's PECOTA, for softball. I've concluded that this is a basically useless endeavor-- incredibly difficult with little payoff besides finding out if I'm as smart as I think I am. We're not professionals; Havelock will not be considering this when deciding whether to sign Alex or Zach to a multi-million dollar deal to patrol centerfield. But the idea refuses to die.

We do have an incredible statistical database, with basic batting stats going back to 1983, and the full range of stats starting in 1985. Pitching was not always tracked diligently, until Joe Gerber came along, and fielding is impossible to track at this level. Gerber, by the way, left me a valuable statistical database consisting of-- I believe 9-- complete seasons from the 1990s and 2000s. For those seasons at least, I could (if I accept this mission) weed out the park effects from the different fields we've used. It's easier to hit in the winter games, for example. Hastings produces a lot of doubles. Dobbs produces cheap homers. When SFLOI used Dobbs as its primary field, there was a HR explosion of epic proportions, similar to what would happen if the Colorado Rockies moved to Mexico City with 275-foot fences and built eight clones of Ted Williams with DNA taken from his frozen head. When I was hitting at Dobbs in late 2008, I hit 7 of my 12 homers in a single month. Jeff Appell homered 19 times in a season, Glen Lawrence 17, Ian Lebowitz 16. Heckscher, with its spacious outfield (and some claim, its difficult hitting backdrop), took away everyone's "power", though they were really the same hitters. Look in the lifetime stats-- the HR explosion is 1998-2001, and 2002-2005 is the dead ball era.

So, obvious conclusion #1-- some ballparks make certain hitters look good. Singles hitters do not benefit from Dobbs Ferry. Big fly-ball hitters do, as do line-drive gap hitters.

In baseball predictions, age is a big factor, but I don't think that would be the case in softball. Most baseball players are finished by 35, and softball hitters are never "finished", and you could make a case that the decline phase only even begins somewhere between 40 and 60. I would guess that how often a player plays, and how good their conditioning is, has a greater effect. So, I might look at how players perform after 200-AB seasons. Do they improve, or at least decline more slowly? Complicating this-- many of our players get at bats elsewhere, with other teams and pickup games. How can we measure for that?

Well, that's all the stats I have in me for today, and my girlfriend will probably kill me if I don't get off her laptop soon. Nate Silver, one of the guys who designed PECOTA, later said that ballplayers perform pretty much like they did in past seasons. That's good enough for me. Maybe.