Goatriders of the Apocalypse

Taking stock, part II

If you haven't already, go back and check out part I.

You back? Good. Now we're going to do Wins Above Replacement, but for pitchers. The basics of it is this. You figure out a pitcher's win percentage, and compare that to a hypothetical "replacement level" pitcher, a guy like Les Walrond or Ryan O'Malley.

A pitcher's win-loss record isn't very useful for this purposes, for two reasons: it includes a lot of factors beyond a pitcher's control, and because no decisions don't tell us anything about whether or not the pitcher gave the team a chance to win.

So, first we estimate a pitcher's defense (and luck) neutral Earned Run Average, through FIP. Remember: we've already credited the team's defensive performance to the fielders via Zone Rating, and so we need to avoid crediting pitchers for a fielder's performance or the numbers will be off. FIP considers home runs, strikeouts and walks and estimates a player's ERA, given an average defense and average luck.

From there, we calculate a player's win percentage versus an average pitcher. (Or, looked at another way, given a league-average offense behind him. Remember: we calculate wins generated by the offense seperately.) Then we compare that with our replacement level pitcher to find out how many additional wins that pitcher contributed.

For relievers, they are also given a "bonus" for their leverage, or the importance of the game situation they are used in. Kerry Wood has the team's highest leverage - he pitches in the most tight spots - and recieves credit for it.

Remember: what we are measuring is the value of a player's performance, not his true talent level. Over a half-season of baseball, a player can perform above or below his underlying talent level. FIP is a better indicator of a pitcher's true talent level than ERA, but that's not why we're using it here. Hopefull tonight I'll get projections for the rest of the season, which will be an estimate of true talent.

Now for the tables. I may come back and replace the EditGrid tables with prettier HTML tables, depending on how industrious I feel. I may not.



I have a feeling I'll be saying a lot more about Ryan Dempster tonight. He, along with Z and Wood, are the real standouts of this pitching staff so far. The rest of the bullpen seems astonishingly mediocre following Marmol's blowup, and the rest of the rotation was solid but unspectacular.

You may be wondering - how meaningful are these numbers, anyway? I added everything up, and came up with (roughly) 27 WAR - 16 wins for hitting, 7.5 for starting pitching, 3.5 for relief pitching. A replacement level team is right around a .300 win percentage, which would be 29 wins. Add it all up and you get 56 wins. The Cubs have won (wait for it) 57 games going into the All Star Break. I can live with that level of accuracy, I really can.

Chicago Tribune's Chicago's Best Blogs award