BP Unfiltered: A Bad Framer, a Bad Call, and an Encouraging Stat About Umpires

If you weren’t watching the World Baseball Classic on Thursday night, you missed a memorable moment with one out in the top of the ninth, when Dominican Republic pinch-hitter Erick Aybar broke a 1-1 tie with a single off US closer Craig Kimbrel, driving in Nelson Cruz from third. The go-ahead run proved to be the winning run, sending the 5-0 Dominicans to the semifinals and the 3-2 Americans to an elimination game against Puerto Rico on Friday.

The single was exciting, but there's more to say about the preceding pitch. Here’s Kimbrel’s 1-1 offering to Aybar at regular speed:

And slowed down:

The pitch was way outside, home plate umpire Angel Hernandez ruled it a strike, and Aybar was understandably upset. As was the Dominican bench:

Although the blown call didn’t end up hurting the DR, it had the potential to be a pivotal moment. After 1-2 counts last season, Kimbrel’s opponents hit .071/.114/.107 and struck out 65 percent of the time, so the strike put Aybar in an almost impossible spot (not that Kimbrel’s opponents hit him much harder after 2-1 counts).

It’s hard to say how it happened. According to Harry Pavlidis, the “strike” was 1.37 feet from the center of the plate, or roughly eight inches outside, when it crossed the front edge. The average strike zone against left-handed hitters is shifted a couple inches farther outside than the average zone against righties, and Hernandez has a big zone—the average strikeout rate with him behind the plate last season was the seventh-highest out of 82 umpires—but this call was unusually ugly. Hernandez watched it all the way into the glove:

He just didn’t notice that the glove was well outside the strike zone.

I want to point out two things. First, the pitch was actually breaking back toward the plate when it was caught, but it’s difficult to tell because of catcher J.P. Arencibia’s exaggerated slide away from Aybar. According to Max Marchi’s framing stats, Arencibia was the worst framing catcher in baseball last season (in a cumulative sense), costing Toronto 23 runs. Watch Jonathan Lucroy, an excellent framer, demonstrate and describe how he receives pitches in this video at MLB.com: he sets a low target and tries to “catch the ball and stop it where it is,” without making excess movements. Granted, Kimbrel missed his target by a considerable margin, but Arencibia does the opposite of that. It’s even more surprising that the pitch was called a strike in light of how poorly Arencibia received it.

And second: Maybe one of the reasons that this strike stands out is that we’re not seeing calls of this caliber as often as we used to.

Here are the counts of regular-season pitches captured by PITCHf/x in each of the last six years:

2007: 322,812
2008: 688,609
2009: 701,358
2010: 697,475
2011: 698,089
2012: 694,014

And here are the number of pitches at least as far outside as the 1-1 to Aybar that were called strikes:

2007: 1656
2008: 1200
2009: 561
2010: 480
2011: 375
2012: 317

You wouldn’t know it from last night, but umpires aren’t blowing calls that badly as often as they used to. And we probably have PITCHf/x to thank for it. The technologies that enable us to watch a blown call over and over and evaluate exactly how awful it was are the same ones giving umpires both the incentive and the tools to be better at their jobs.

Thanks to Harry Pavlidis for research assistance.