Giants’ wOBA Against
Quick post time:
Batting average against is commonly cited in mainstream analysis as a means of identifying which pitchers have been hit harder than others. This is blatantly incorrect. Batting average tells us absolutely nothing other than the fact that the pitcher gave up x amount of hits. Well, were they predominantly bloop singles? Were they hard hit doubles, or were they home runs? Furthermore, what about other baserunners allowed not by hits? We might as well just look at on-base percentage against, then. Right?
Well, that’s definitely a step up. We’re now looking at walks and hits allowed per plate appearance. But that only solves part of the issue, as we’re still weighting singles as much as home runs like we do in batting average against. How about slugging against? Well, we’re still looking at some issues. A home run isn’t worth four singles, nor is a triple worth a double and a single. That’s just awkward weighting. OPS against? Still not what we’re looking for. We’re adding inaccurate coefficients from slugging percentage to on-base percentage, and we’re mixing two rates with different denominators. That doesn’t do us any good. What then?
In short, what we’re looking for is wOBA. Weighted On-Base Average, as I’ve written about extensively before, is a rate statistic derived from linear weights, so it accurately weights the impact of each event. Problem solved. Since we have batting lines against available on Baseball-Reference, we’re able to generate wOBA against for all pitchers. Naturally, the only pitchers that I really want to look at are…well, the Giants, of course. I’ve generated the coefficients based on linear weight values derived from a relatively complex Baseruns formula, which you can find here (goodness, that is quite a mouthful):
This model incorporates ROE, along with intentional walks just for the sake of completeness. Steals and caught stealing against work as a proxy for how good the pitcher is at holding runners on. Anyways, the results:
I’ve also park-adjusted the wOBA against based on batter handedness (pretty cool, huh? In this case, it makes virtually no difference though). Anyways, nothing too surprising here. Big Time Timmy Jim is leading the pack, as per usual, followed by Cain, Zito, and Sanchez. Looks like Zito’s been doing slightly better than Jonathan, who isn’t doing much better than Brian Wilson, despite more than twice as many innings pitched. The Giants could really use some improvement from Affeldt, in addition to another lefty reliever- Runzler’s been doing all right, but we could do much better in terms of left-handed relief. And maybe, just maybe, we shouldn’t let Todd Wellemeyer do anything more than mop-up duty…and we should give Romo some more chances in higher leverage situations. Just sayin’.
*If you’re familiar with wOBA and LWTS, you’ll notice that it doesn’t make sense to see Tim Lincecum with a wOBA against of .300, which translates to -12.5 runs. I multiplied by -1 in order to switch it around, to make the runs above average figure from the pitcher’s perspective, not the hitter’s.