2008 Pitcher Projections

Pitcher Projections

I'm not very happy with them, but here it goes anyway. I've projected just about everyone who faced at least 100 batters in a full season league last year, with at least one Japanese player, and a few guys who were hurt last year.

They were reasonably accurate last year, by accurate meaning maybe 1/10th of a percent more accurate than other projections or so (see my October archives for comparisons to other systems). I'm projecting a lot more this year, last year I just did major leaguers and top prospects. Under the theory that it takes me as much time to select "top prospects" than to just do everyone, I projected everyone.

That's where the problem comes in. I don't think numerical estimations are the best way to evaluate minor league pitchers, even if they are complete, holistic, and objective. The conversion factors are just brutal, especially for the low minors. There are a lot of pitchers with so-so stuff that can get people out in the minors, but get killed in the majors. Even after regressing the samples for convertion factors to the mean, there are very few pitchers who put up pretty MLE's. Perhaps that is correct (If they're so good, they should be in the majors, not putting up MLE's in the first place). But I think you're far better off buying a copy of Baseball America's prospect handbook than taking these numbers too seriously for minor leaguers.

For major leaguers, I'm much more confidant, as confidant as you can be in projecting such a fluky, oft-injured creature like an MLB pitcher. Their batted ball stats (grounder, fly, pop, and liner) feature prominently in the projection, for minors they do not - I could look up one pitcher on minorleaguesplits.com, but I'm not looking up 1800 of them.

So what's the value of the minor league projections? If you are into detailed sims that include the minors then this should be a treasure mine for you. It can give you a rough idea of how your team's organizational depth stacks up. It might even be a reasonable estimation of how the pitcher will perform now, but I don't think the system has any clue how a pitcher will develop. If Nick Adenhart's projection is for a high 5's ERA, don't take it to say he can't pitch. Just that he's probably not ready yet, after a mediocre K/W rate in a pitcher's park in AA ball. If Red Sox fans think Clay Buchholtz is going to do better than a 4.46 ERA, they are probably right. Its just the limitations of the system.