To follow up on my article at the Hardball Times, here’s another look at the strike zone- this time taking it pitcher by pitcher (hitters come next). When I started crunching all this data I thought that this would be much more interesting than just looking at the quirks from umpire to umpire, and go towards proving (and/or quantifying) another widely held baseball assumption- that umpires tend to give the best hitters and pitchers favourable calls.
However, it’s not that easy to draw conclusions this time because there are many factors (what hand he throws with, how hard and with what movement, even how much deception is in a pitchers delivery) that could affect the data and one year of (incomplete) data isn’t really enough to go on. I’m pretty confident that the umpire results are accurate because most of them are using 2000+ calls. But for pitchers the average number of calls available is 500, and there’s likely to be a lot of variance in them since they’re dealing with a few umpires each (who we now know can be drastically biased one way or the other) rather than a random sampling of players from the entire league.
It seems plausible that Cy Young types would dominate the top of the rankings based on reputation alone (and rookie pitchers would have to earn respect) but that’s not the case. The following table shows the top and bottom 10 pitchers in terms (here is the complete list) of how many extra strikes they got per game. The only thing I’ve changed is the Strikes Above Average (SAA) rate is now expressed in terms of extra strikes per 75 pitches, because that’s roughly how many a ump makes for a pitcher over 9 innings. So these numbers are are +/- of how many favourable calls these pitchers got per game last season, according to pitch f/x:
|Bottom SAA||Top SAA|
|Derek Lowe||6.26||Frank Francisco||-5.90|
|Jeremy Accardo||5.24||Gavin Floyd||-5.23|
|Joe Beimel||5.31||Brandon Morrow||-5.07|
|Roy Halladay||4.14||Randy Wolf||-4.90|
|Tony Armas||3.93||Kameron Loe||-4.50|
|Bronson Arroyo||3.87||Ryan Feierabend||-4.39|
|Cole Hamels||3.70||Vincente Padilla||-4.29|
|Josh Towers||3.60||Joaquin Benoit||-4.10|
|Yovani Gallardo||3.58||Jamey Wright||-4.05|
|Joe Kennedy||3.32||Jon Lester||-3.98|
The only thing that really jumps out from for me is that there are three blue Jays in the top 10. (Scott Downs was over 7 but just missed the cut for minimum pitches). Perhaps a sign of good framing behind the plate? But in general well-known aces don’t seem to be highly favored- at least not enough to overcome the noise and randomness fluctuation in the data.
Here’s where it gets interesting, though. To see if there was a correlation in the long run between better pitchers and better calls, I graphed the opponent’s batting average against the number of extra strikes per game and ran a trendline. There is none- it’s dead level. That would seem to confirm that the quality of the pitcher on the mount doesn’t matter (or isn’t particularly significant). As is probably blindingly obvious, charting BB/9 against extra strikes BB/9 against extra strikes shows that pitchers who get better calls walk fewer batters.
But here’s the kicker- the same is true for strikeouts: pitchers that receive a higher number of favorable calls strike out fewer batters, not more.
Of course that makes zero sense, so it has to be the other way around. Pitchers who strike out a lot of batters tend to have electric stuff rather than impeccable control, and those kind of pitchers don’t get as many calls going their way. They still manage to strike out far more batters by getting swinging strikes and overpowering hitters down the middle, but it’s pitchers like Tom Glavine, Josh Fogg, Jarrod Washburn and Roy Halladay that are the umpire’s darlings and manage to expand the strike zone, even though they don’t reduce the “ball” zone.
That doesn’t exactly come as a big surprise to anyone who has watched a control artist pick establish the outside corner, but now we know that control has a significant impact on umpires, while a pitcher with superior pitchers may actually have a negative effect. And it can be a significant difference- between the two extremes of Roy Halladay and Roy Oswalt, it comes out to about 7 calls a game, although on average the range is more like 3 or 4.