Earlier this season, Nationals' starting pitcher Jordan Zimmermann sat down with Fangraph's David Laurila and spoke about the deliberate attempts he has made to work deeper into games by 'pitching to contact'.
"Before I had Tommy John, I was striking a lot of guys out. I would throw a lot of pitches. Once I came back from it, I told myself, ‘Let’s pitch to contact and let’s not strike as many guys out.’ That way I can stay in the game longer."
Far too often the concept of a 'pitching to contact' has been used as a euphemism for 'can't strike anyone out', but for Jordan Zimmermann that is likely not the case.
It is certainly true that his K% has noticeably dropped in each of the last four seasons:
And yet despite this drop in strikeouts, Zimmermann is getting better results than he ever has. By lowering his walk rate along with the strikeouts, Zimmermann has managed to maintain his FIP while simultaneously pitching deeper into games. He has accomplished this by throwing far fewer pitches against the opposition than he ever has.
From 2009 to 2010 Zimmermann's pitches per Batter Faced (or P/BF) was a rather pedestrian 4.04, but from 2011 to 2012 that number is down to a very efficient 3.71. To put this into context, here are the largest drops in P/BF for starting pitchers (with at least 150IP) since retrosheet began tracking pitch counts in 1989.
BIGGEST YEAR-TO-YEAR P/BF DROPS SINCE 1989
Zimmermann's quaint balancing act between performance and efficiency prompted a discussion over at The Book Blog. Readers wondered, at what point do the gains from an increase in efficiency outweigh the losses of a declining FIP? Tom Tango suggested searching for a set of comps for Zimmermann by meeting the following criteria:
1. Look for good young starting pitchers, say a FIP that is no worse than 90% of league average, over at most a 2 year period, with at least 500 batters faced
2. For those pitchers, only keep those that maintained a FIP of no worse than league average over the next two years, min 500 batters faced
3. Only keep those pitchers who dropped their K per PA rate by at least 10%
I used my own discretion to define "young pitchers" and kept only those that were still younger than 28 years old by the time the four-year sample was complete. This returned 30 pitcher-periods, although two of these returns were overlapping periods from the same pitcher (Barry Zito and Tim Hudson). If interested, you can view the results as a google doc HERE.
(Along with the pitcher-efficiency metrics I've included a few other 'ci' or 'change in' measurements including K/BB/HR rates, wOBA against, BABIP, and non-park-adjusted FIP+ and ERA+ to measure against league averages per Tango's suggestion. IP, BF, and pitch-count data are tallied from outings as a starter only.)
The group did, in fact, improve it's Pit/BF as expected by .07 on average. This did not, however, lead to more efficient innings or deeper outings for the group as a whole. As it turns out, P/IP went up (+.21), BF/IP went up (+.13) and IP/GS went slightly down (-.03). The problem is that same age-old dilemma for pitchers, perhaps even for young ones in particular: While pitching to contact will allow you to throw fewer pitchers per batter, by increasing the amount of balls in play you naturally will allow more hits. This means more baserunners and fewer outs, which means fewer innings pitched and shorter outings altogether.
So what if we disregard the age-restriction and just try to find pitchers who had dropped their K-rate, maintained their FIP, and still managed to go deeper into games? Is it possible by observing pitchers that have increased their efficiency that they can provide a clue as to how they accomplished it? I ran a second query just for kicks.
Of those pitchers that increased their IP/GS by at least .30 you'll see a dramatic 10% decline in their walk rates to accompany the loss in K's. This led to a far more tolerable slip in FIP+ which provided better results across the board. There was a general increase in HR-rate of about 5%, but overall the group fared well by increasing their ERA+ by a substantial margin. Certainly, it seems, half the battle for a pitcher cutting down on his K-rate is keeping that walk rate down just as much.
But the elephant in the room in these cases is always BABip and Zimmermann is no exception with a .264 BABip in 2012. Similarly, two thirds of the IP/GS Improvers saw a lower BABip in their second two-year sample and the group on average saw a 3% decrease. I'm sure it's a tempting argument for many to suggest that 'pitching to contact' involves inducing weaker contact and that a lower BABIP is intentional, deliberate, and part of that plan. But keep in mind that the original group showed no improvement in BABip, and in fact, the group on average saw a slight 3% increase.
It's probably impossible to completely separate pitchers who have deliberately decided to pitch to contact from pitcher's who have simply lost their ability to strike batters out. And a fair analysis on the matter is not something I would hold my breath waiting for (though I am certainly open to any suggestions). And while I'm willing to applaud Jordan Zimmerman for proving he can maintain a reasonable FIP while cutting down his pitch count and working deeper into games, I'm still cautious as to whether the overall success of his experiment may be exaggerated by a simple case of plain old BABIP-deflation.
Follow the dude who wrote this @dangerwhatevs and talk about baseball, math-rock, or dinosaurs.