Shafty Phat-C Diablogs Astros Baseball Fantasy Sports The Arts

Thursday, July 13, 2006

Pitching Win Shares, at the Break

Having handed out grades for the Astros' pitching staff yesterday, I thought I'd come back this morning and try to offer some additional statistical rationale for yesterday's rankings. There may be some disagreements between the two rankings, because yesterday's was what I perceive as a rabid fan (with a few traditional statistics thrown in); these rankings, however, are entirely statistics-based, without regard to whether I think somebody's been clutch, etc.

Bill James devised the concept of "Win Shares," a rather complicated statistic that tries to take into account all of the contributions each player makes towards his team's win total, and then cranks out one number that represents the number of wins contributed to the team, times three. Rather than calculate all of the numbers myself, I've taken those listed yesterday in The Hardball Times, which I highly recommend to other stats geeks. Wonderful site. They have tweaked Bill James's original formula somewhat. (Without getting into the specifics, each player has a calculation for batting, pitching, and fielding. Playing time is factored in; bench players are not expected to contribute as much as starters.)

Although their win shares are expressed by multiplying times three, I'd rather keep the original number so that it gives a clearer picture of approximately how much each player has contributed to the Astros's 43 wins at the moment. Keep in mind that win shares do not reflect "loss shares." That is, where a player actively contributed to a loss, that isn't reflected here, except to the extent to which that player failed to contribute to a win. (Also, even though guys have contributed to many wins--such as Oswalt, who I said should be 11-4 by now--this number reflects only those games that they essentially won by themselves.)
Pitcher Win Shares Above/Below Expected
Oswalt 3.33 +1.67
Qualls 1.67 +0.67
Nieve 1.33 +0.33
Pettitte 1.00 -0.67
Wheeler 1.00 +0.67
Buchholz 1.00 -0.33
Rodriguez 1.00 -0.33
Lidge 0.67 -0.67
Borkowski 0.67 +0.33
Clemens 0.67 +0.33
Springer 0.67 +0.33
Backe 0.33 +0.33
Sampson 0.33 +0.33
Miller 0.33 Even
Astacio 0.00 Even
Gallo 0.00 -0.33

I didn't do too badly, after all. According to Win Shares, the two biggest disappointments have been Pettitte and Lidge (whom I graded as D- and F-, respectively). Not surprisingly, Oswalt has contributed much more to his team's wins than has the average starting pitcher. Qualls (whom I graded B+) and Wheeler (C+) also performed better than expected; perhaps I should have been less critical of Wheeler.

I still wouldn't downgrade Buchholz and Rodriguez, based on Win Shares. They're being compared to the average pitcher with their innings pitched and, after all, they're rookies.

One final note: when compared with last year's Win Shares, Oswalt has contributed even more to his team's success this year than last year (when he won twenty games). Pettitte has contributed only a third as much as he did last year.

Lidge has clearly been the biggest disappointment. At this time last year, Lidge had contributed 2.19 Win Shares to his team's success. This year, less than one game. Like Pettite, he's been two-thirds less effective than last year. In other words:

He sucks. As if we needed statistics to tell us that..


Post a Comment

<< Home