Getting Strikes on the Edge

The last time I wrote about Edge% it was in the context of the Tampa Bay Rays using it to get their pitchers into more favorable counts on 1-1. But now I want to take that topic and drill a little deeper to understand how often edge pitches are taken for called strikes.

Overall, pitches taken on the edge are called strikes 69% of the time. But that aggregate measure hides some pretty substantial differences. Going further on that idea, I wanted to see how the count impacts the likelihood of a pitch on the edge being called a strike.

Here are the results:

Just as a refresher, here is what the Edge looks like visually as Jeff Zimmerman and I have defined it:

Edge for LHH and RHH

I broke down the percent of pitches by count that were taken by a batter, that were thrown to the edge and that were called strikes. Here’s the league averages for 2013, ordered from pitcher-friendly counts to hitter-friendly counts. I also listed the overall called-strike percentage, as well as the percentage for all two-strike counts:

Overall Edge% 17%
Overall Edge Strike % 69%
0-2 Edge Strike % 53%
1-2 Edge Strike % 55%
0-1 Edge Strike % 59%
2-2 Edge Strike % 60%
1-1 Edge Strike % 66%
0-0 Edge Strike % 73%
1-0 Edge Strike % 73%
2-1 Edge Strike % 67%
3-2 Edge Strike % 63%
2-0 Edge Strike % 76%
3-1 Edge Strike % 68%
3-0 Edge Strike % 81%
2 Strikes Edge Strike % 57%

While Edge pitches are generally called strikes 69% of the time, the most friendly pitcher’s count (0-2) only results in a strike 53% of the time. Compare that to the least pitcher-friendly (3-0) at 81% and you are looking at a 28-percentage-point swing. Generally speaking, as the count runs more in the batter’s favor, umpires appear to expand their strike zones. If you split the counts roughly in half (with 0-0 essentially being even in terms of above/below league average for run values), pitcher-friendly counts generate an edge called strike 59% of the time, compared with 71% for hitter-friendly counts.

It’s possible this effect is partly due to an umpire’s expectations. On two-strike counts, umpires are likely assuming pitchers will avoid the strike zone to the extent they can — since hitters will have to expand their own zones to protect against strike three. Assuming  pitchers are doing that can lead to umpires assuming pitches aren’t as close as they really are, which would lead to fewer called strikes in the zone. The opposite effect could hold for 3-0. Just a thought.

I also ran the data and grouped the percentages by pitcher and by count. I included all pitchers in 2013 and then broke out just those with at least 500 pitches. You can sort and filter in the web application or just download the data:

The first set of columns shows the raw percentages, by count, of called strikes for pitches thrown to and taken on the edge. The second set of columns shows what percent of league average those percentages are. So, for example, Justin Verlander gets a called strike on the edge 90% of the time in 1-2 counts. That is 162% of league average (55%).

Needless to say, there is a lot of data here. But I’ll throw out a few highlights in particular and then let the readers dig around.

For pitchers with between 1,000 and 2,100 pitches, who’s the the best 0-2 edge artist? Arguably it’s been Kevin Slowey. So far in 2013, he’s thrown 21 pitches to the edge on 0-2 that have been taken — and every time they’ve been taken for strike three. The most called strike threes on the edge? That title goes to Cliff Lee, who has 35, but he’s only getting those calls 73% of the time. Of course, that’s still 138% of league average.

In terms of the overall title, that goes to David Price who is getting called strikes at a 83% clip, although Kyle Kendrick is right behind Price at 82.2%. And that’s with  roughly 700 more pitches thrown. But even those these two pitchers have nearly the same outcome when batters take their pitches on the edge; their performance varies considerably by count.


David Price Kyle Kendrick
Total Pitches 1052 1763
Overall Edge% 20% 18%
Overall Edge Strike % 83% 82%
0-2 Edge Strike % 60% 25%
1-2 Edge Strike % 100% 75%
0-1 Edge Strike % 83% 60%
2-2 Edge Strike % 86% 100%
1-1 Edge Strike % 63% 73%
0-0 Edge Strike % 84% 90%
1-0 Edge Strike % 100% 90%
2-1 Edge Strike % 80% 71%
3-2 Edge Strike % 50% 100%
2-0 Edge Strike % 100% 100%
3-1 Edge Strike % 100% 100%
3-0 Edge Strike % 100% 100%
2 Strikes Edge Strike % 79% 71%

Price has a much higher percentage of called strikes in the most pitcher-friendly counts, including a commanding 60% to 25% advantage in 0-2 counts. Price’s performance in 0-2 is 113% of league average, while Kendrick’s is just 47%. Kendrick starts to turn the tables in 2-2, 1-1 and 0-0 counts and really dominates in 3-2 counts, where he’s managed to get called strikes on the edge 100% of the time. For those keeping score, that’s 13 3-2 punch outs for Kendrick on the edge this year — though that’s far from the highest number (Gio Gonzalez has 27). Still that fact is quite useful.

This leads to a question I want to look at next: What effect does batter and pitcher handedness have on whether umpires will give pitchers calls on the strike zone’s edge? Do umpires generally like lefties more than righties? Do they give pitchers more calls when they’re in a platoon situation? I’ll tackle that in the next installment.


PITCHf/x data used covered all games from the beginning of the 2013 season through July 10.

Bill leads Predictive Modeling and Data Science consulting at Gallup. In his free time, he writes for The Hardball Times, speaks about baseball research and analytics, has consulted for a Major League Baseball team, and has appeared on MLB Network's Clubhouse Confidential as well as several MLB-produced documentaries. He is also the creator of the baseballr package for the R programming language. Along with Jeff Zimmerman, he won the 2013 SABR Analytics Research Award for Contemporary Analysis. Follow him on Twitter @BillPetti.

Newest Most Voted
Inline Feedbacks
View all comments
10 years ago

Bill – great stuff. One thought your speculation concerning possible confirmation bias (expecting a ball so calling balls). If you have made this argument somewhere else and I haven’t see it, my apologies.

Isn’t the most likely explanation is a systematic umpire preference for nonintervention? While umpires are obviously charged with calling balls and strikes in all circumstances, I would assume that umpires, all else being equal, would prefer a player to hit a home run or ground out to short because it is harder to “blame the umpire” for that result. In the case of walks or strikeouts (most likely in 3-0 and 0-2 counts), even if the umpire calls a perfect strike zone, there is going to be some allegation that the umpire influenced the game (however unfair). Trying to avoid this allegation probably won’t lead an umpire to totally shirk their duties (i.e. calling a pitch down the middle a ball), but for border-line (or Edge) calls, perhaps this motivation could be a tipping point?

Just a thought. Anyhow, really interesting stuff. Will be curious to see how this dovetails with things like catching framing.

10 years ago
Reply to  Jeff

I think it’s likely that both factors are at play here. I don’t see how anyone could conclusively say that either is more prominent than the other.

10 years ago
Reply to  Jeff

Just thinking about human nature, this strikes as far more likely. To make a decision like calling a player out on a questionable 0-2 pitch, when the alternative is just to let it go to 1-2, an umpire would have to be pretty convinced. This could lead an umpire to make the less “controversial” call in situations where he might not feel certain.

The “expectation” argument doesn’t make sense to be at all. What’s the effect at play? He expects a pitch out of the zone, the pitch comes and it is in the zone, and he thinks “Oh shit I never considered that that might happen.” The argument suggests his perception of the pitch is altered. I think it’s just his judgment of the pitch that varies.

10 years ago
Reply to  Jeff

I agree with you. Everything I’ve read about how the mind works (and I have read a lot) supports your explanation more. To really prove either side (or some other hypothesis) would require an extremely difficult experiment–if one is possible at all.

10 years ago
Reply to  Jeff

Great work! To expand on the idea of being “non-controversial”, it would be interesting to see the data for Home and Away. Human nature could lead the umpires to call more at bat deciding pitches in favor of the home team. It would also be interesting to see if certain umpires have significant trends.