Aaron Sanchez: Strikethrower?

A week ago, I wrote about some concerns I had digging through Sanchez's 2014 bullpen stats. Chief among those was a propensity to fall behind in the count significantly more than league average, which is generally not a recipe for pitching success in the long run. Generally, it's also a sign that a pitcher is not able to consistently throw strikes. And it wasn't just an issue upon promotion to MLB; from the second half of 2013 to his promotion, Sanchez walked 81 in 144 innings.

At the State of the Franchise, GM Alex Anthopoulos answered a question stating (I'm paraphrasing) that Sanchez had become "a strikethrower" following mid-season adjustments made around the time he was promoted to Buffalo. Having already looked at some of the data points I included last week, I found this claim curious to say the least. He didn't walking as many as in the minors, but underlying issues that lead to walks were there.

An important caveat is that looking at ratios of balls and strikes isn't exactly the same thing as strikethrowing ability, if that's defined as consistently pitching in the strike zone. Pitchers can induce strikes on balls out of the zone. Catcher framing ability is another factor, and one that seems especially relevant for the Jays. In a smallish sample of 33 innings, even umpire factors could move the numbers.

The best proxy for strikethrowing is Zone%, which happily is available on leaderboards (under plate discipline) at Fangraphs, making it easy to see how Sanchez stacked up and to evaluate Anthopoulos' claim. The league average Zone% for relievers was 45%, at 41.9% Sanchez ranked 168th of 209 relievers with 30 relief IP in 204. That puts Sanchez at the 20th percentile, roughly a standard deviation below average (assuming a roughly normal distribution). Not exactly what one would expect of a bona fida strikethrower.

So, case closed, right? I thought so initially, but not so much. FanGraphs actually provides two sets of plate discipline statistics. The first, referenced above, is from Baseball Info Solutions (BIS), who use trained video scouts watching the games to compile their data. The second is based on PITCHf/x data captured by the cameras installed in each stadium.

In general, the data from these two sources is very consistent, though not exactly the same. For Zone%, FanGraphs' PITCHf/x data records about 3-4% more pitches in the strike zone, a difference that holds up over time and for most players. But not Sanchez. His PITCH f/x Zone% was 51.6%, almost 10% higher than BIS reported. With a 2014 the league average for reliever Zone% of 48.5%, Sanchez ranked 51st of 209, or at the 75th percentile. This would indicate he was indeed filling up the strike zone.

Of the aforementioned 209 relievers, Sanchez had the second highest difference between the two Zone%, trailing Yusmeiro Petit and ahead of Neftali Feliz. The chart below shows how tight the relationship generally is, with Sanchez highlighted as the red outlying circle:

So why is there such a big difference? It seems to me the source of the data is significant. Trained video scouts, while undoubtedly very good, are humans and subject to bias, whereas the PITCHf/x cameras are much more objective (though they will have calibration errors). Especially for Sanchez, whose two seam fastball had so much movement and who was throwing mostly to a catcher who had issues receiving the ball, it's plausible there was a material issue converting pitches around the edges in the zone into called strikes rather than balls. In which case, just having Russell Martin behind the compared to Dioner Navarro could go a long way towards resolving one of the major issues identified last week as a potential headwind to MLB starting success.

And it wasn't just Sanchez either. At 5%, The Blue Jays had the biggest differential in MLB between BIS and PITCHf/x Zone% (50.3% v. 45.3%), albeit with a number of teams very close behind. Of the 140 pitchers with at least 100 starting innings, the Jays placed three in the top 20 largest differentials: Buehrle (9th, 6.7%), Stroman (11th, 6.2%), and Happ (20th, 5.9%).

To try and answer the strikethrower question definitively, I pulled Sanchez's data from a third source, Baseball Savant, which (in addition to being totally awesome) pulls all the PITCHf/x data from MLBAM used in GameDay, including their own zone classifications. Here was another surprise. By their classifications, only 177 of 465 (38%) Sanchez pitches ended up in the strike zone, with a league average in the low-40 percent range. This would indicate problems throwing strikes.

This directly contradicts the other PITCHf/x data putting Sanchez above league average. Frankly, I have no idea what to make of it, other than classifying pitches is not as neat and tidy as it often seems on various plots and Sanchez apparently had a really high number of pitches around the edges that could be either way.

One final data point. With yesterday's outing, Sanchez has now thrown 18.1 innings and faced 75 batters, with Martin doing most of the catching. So far, his ball percentage is down about 2% from last year at 36.3%, just slightly above the 2014 league average. His rates of called strikes, fouls, and whiffs are mostly in line, with balls being put in play 21.5% of the time, up 3%. Perhaps a sign of progress, perhaps hitters just looking to get their hacks in. However, he has continued to fall behind on the first pitch, going to 1-0 counts 45% of the time compared to 46% last year. He's only struck out nine batters (12%), but held the walks down too (5.3%, 6.6% free passes including a HBP).

So did Sanchez actually become a strikethrower? I don't know. But this could well be a key to the Blue Jays season. If Sanchez throws a reasonable amount of strikes and limits walks, even with a lowish strikeout rate and strong regression on balls in play, he will be a productive MLB starting pitcher. Something like 16% strikeout rate, 8% walk rate, and 55% GB-rate works out to roughly league average expected run prevention.

And if he actually fills up the zone in a big way, assisted by Martin, watch out.