Saturday, August 17, 2013

Smoke and Mirrors

If I asked you last night who Pittsburgh's best starting pitcher is, there's a good chance you'd say Jeff Locke. Locke led the Pirates with a 2.43 ERA, behind only Clayton Kershaw and Matt Harvey for all pitchers. You'd be quite wrong.

A shiny ERA is easy to comprehend and an appealing way to quickly assess a pitcher. The metric does a good job of basically measuring what it says it will and isn't hiding behind any gory mathematics. Given this, Locke's ERA is very impressive, to be sure, but it doesn't take much digging to realize that it is headed for a correction the size of the subprime mortgage bubble.

According to DIPS theory (Defense Independent Pitching Statistics), a pitcher essentially can produce three outcomes and the rest of the play is independent of that work. These three outcomes are strikeouts, walks and batted ball type.

These three components come together to produce a predictive statistic called Fielding Independent Pitching or FIP. FIP has been proven to be more predictive of future ERA than ERA itself (if you care to check this, it's easily google-able). Knowing this much, a basic rule of thumb for superficial use of ERA and FIP could be that ERA tells a story about what happened and FIP predicts what will happen going forward.

For example, Pitcher A throws 6 innings and surrenders 3 hits and 4 walks while striking out 2 en route to allowing 2 runs, that pitcher's day would be better described by ERA but to say his true talent is that of a 3.00 ERA would lose a gambler many bets.

All that to say that what Jeff Locke has accomplished over his first 23 starts is significant. He's produced tons of value for his team and is a huge reason they're sitting atop the NL Central. According to FanGraphs he's been worth 4 wins if you simply use results.

All that also to say that Jeff Locke and his mediocre 3.74 FIP will not continue to produce at a level that rivals Matt Harvey. This stat is fueled by below average strikeout (6.64 k/9 or 17.8% of batters) and walk (4.37 BB/9 or 11.5% of batters) rates. Truthfully, I would argue his FIP may undersell just how mediocre Locke has been as it credits him with keeping fly balls in the park, which he hasn't demonstrated as a skill in the past and therefore is probably not sustainable given his skill set. Given this paradigm, Locke has only been worth 1.4 wins according to FanGraphs.

How is this disparity between ERA and FIP created? A large portion can be explained by babip; his, .256, sits well below the standard average, .300. Over the course of a season that's a difference of about 30 hits. It's been demonstrated time and again that it's nearly impossible for a pitcher to control his babip and therefore should be expected to regress to the mean, .300. The Pirates exceptional defensive positioning and general skill will likely depress his babip below the league average (the team babip is .274) sympathetically hedging Locke's regression to some degree.

Locke has been an okay pitcher but he's produced at exceptional levels in 2013. There's a distinction there that's evident if you've read this far and what he has been will be more predictive of what he will produce going forward.

The regression monster showed himself to Locke tonight in a game which he allowed 8 runs in 2.2 innings against the Diamondbacks. His ERA jumped to 2.90. In all likelihood it will continue to rise; and if there's any gamblers out there, I'd be happy to bet on it.

No comments:

Post a Comment