| well what WAR does it try to figure out as much as possible how much is the individual pitcher's responsibility, and tease out the league factors, the defense behind him, etc. The problem is that a lot of traditional stats are highly flawed. For example, WHIP - walks are more or less 100% up to the pitcher, but hits are VERY luck-driven. Once a ball is put in play, you are more or less at the mercy of random chance (he can hit it right on the nose but right at someone, he can hit a weak dribbler but beat it out for an infield single) and your defense (maybe Ozzie Smith makes the tough play, maybe a lousy shortstop doesn't). WHIP is less useful in terms of evaluating how good (not lucky!) an individual pitcher is, than looking separately at BB% (walks issued per plate appearance-against), K% (K per PA), and BABIP (batting average on balls put in play - this is a measure of how lucky a pitcher is because once a ball is put in play, it is basically random chance). WAR takes these stats, figures out how each measures up to league, ballpark and defense, figures out how much each walks, strikeout, home run surrendered, and ball put in play allowed either prevents or allows a run to score, then figures out how many runs it takes to equal a win in the context, and that's your WAR. So if you are set on using traditional stats, I'd say ERA+, K% and BB% are probably good ones to hang on to. |