One simple idea to quantify fastball command

Fastball command is a good thing for a pitcher to have. In 2016, pitchers across the league threw their four-seam fastballs 36.4% of the time, according to Pitch F/X. They also threw their four-seamers, two-seamers, or sinkers 56.6% of the time. Naturally, if a pitcher plans on throwing a certain category of pitches over half of the time, it is important that his ability to locate is up to par.

The problem is that quantifying command is extremely difficult. Nonetheless, Jeff Long, Jonathan Judge, and Harry Pavlidis recently gave it a try for Baseball Prospectus. They use a statistic called CSAA (Called Strikes Above Average) to quantify a pitcher’s command, calling upon a stat that is often used to quantify a catcher’s pitch framing ability.

The basic idea behind their theory is that a pitcher who locates well will not force his catcher to move his glove as much. This will make it easier for the catcher to frame pitches on the border of the strike zone, which will lead to more called strikes. But one other idea is to look at how often hitters chase a pitcher’s fastballs out of the strike zone.

Applying a similar idea, pitchers with better command will be better at locating their fastball just a bit outside the zone, forcing hitters to pull the trigger. Pitchers with poor command will either miss down the middle or well out of the zone, and won’t fool hitters.

According to data from Baseball Savant (the search query can be found here), hitters across the league swung at 28.9% of four-seam fastballs outside the strike zone. Against two-seam fastballs and sinkers, they had an O-Swing% of 30.6%. Here are the top five pitchers in both categories, with a minimum of 200 fastballs thrown out of the strike zone:

There are a handful of confounding variables to consider. For one, pitchers who have better overall deception may be better at getting hitters to chase outside the strike zone with their fastballs. However, two members of the 2016 Yankees dispute this idea. Michael Pineda, whose slider makes him a very deceptive pitcher, had an below average O-Swing% of 25% on his fastballs. On the other hand, Nathan Eovaldi was at 30.1%.

I also ran some linear regressions to see if there was a relationship between a few other metrics and Fastball O-Swing%. Again, sticking with the minimum of 200 pitches thrown outside the zone, the results were encouraging. Neither average fastball velocity nor extension had a statistically significant effect relationship with a pitcher’s Fastball O-Swing%.

For four-seam fastballs, spin rate did have a statistically significant relationship with O-Swing%. However, it was fairly weak, with an R-squared of 0.07. The model predicted that for every additional 118 average RPM a pitcher has on his fastball, we would expect his O-Swing% to increase by one point. This makes sense, as pitchers whose fastballs have a high spin rate are often told to try to get hitters to chase fastballs above the strike zone.

In the never-ending quest to quantify fastball command, looking at a pitcher’s ability to get hitters to chase heaters outside the zone might be a starting point. As metrics continue to evolve, we will probably hear more and more ideas. Hopefully, the sabermetrics community will figure out a way to solve this problem once and for all.

Data is courtesy of FanGraphs and Baseball Savant.