Chasers Gonna Chase, But Where?

It was the third inning of the All-Star Game, and John Smoltz and Joe Davis were marveling at Freddie Freeman. That part’s not particularly surprising. Who among us hasn’t marveled at Freddie Freeman? The Dodger first baseman worked a full count off Sonny Gray, then resisted a tempting fastball at the letters for ball four.

As Freeman jogged to first, Smoltz shared some of the wisdom he picked up during his Hall of Fame career: “The biggest thing pitchers want to find out with all the information that’s given: Where is the guy more likely to chase? And rarely do they chase up and down. It’s usually down or up. And then you try to pitch accordingly to the strength. And where Sonny Gray’s strength is, is down. He’ll surprise you up, but he wants to get you out down.”

Whenever I hear Smoltz articulate a hypothesis about the game, I find myself torn. The man has seen an awful lot of baseball, and I don’t want to discount that experience, but he’s also made some pretty outlandish-seeming claims before. Best to be a bit skeptical and double check his math, as it were.

I get that each player has their own hot zone, their own swing plane, a particular area that looks inviting to them. But I also think that some players are just more swing-happy than others. Takers gonna take, chasers gonna chase (chase, chase, chase, chase, chase).

Luckily, this particular Smoltz theory is an easy one to test. I pulled data on every batter in the pitch tracking era. Our sample includes the 720 batters who have seen at least 4,500 tracked pitches. Next we focus on the pitches that are above or below the zone. I used the intersection between Gameday zones and attack zones to look only at pitches in the flashing areas of the diagram below:

With all the data in one place, I calculated the correlation coefficient between chase rates above the zone and below the zone: r = .22. It’s definitely a positive correlation, which means that players who chase more above the zone tend to chase more below the zone too. Takers gonna take, chasers gonna chase. Shake it off, Smoltzie.

However, .22 isn’t the world’s strongest correlation coefficient. Let’s take a look at the scatterplot of our 720 players. The overall trend is clear to see. Of our 720 players, 453 have less than a 10-point difference between their chase rates above and below the zone:

But that still leaves plenty of outliers. 148 of our hitters have a difference between 10 and 15 percentage points; 119 of them have a gulf of more than 15 points. I don’t know exactly where the cutoff is, but if you told a pitcher that the batter they’re facing chases above the zone half the time and below the zone a third of the time, I have to imagine that might affect their approach. The 15-point group makes up 16.5% of the batters in our sample, which means that in any given lineup, there are one or two batters from whom a pitcher should definitely be trying to induce chases either high or low. Smoltz was wrong when he said batters usually chased either down or up, but his point was true for a smaller number of players.

At this point, I should remind you that there was more to the quote. Smoltz next mentioned that pitchers don’t just throw the ball wherever the batter is likely to chase. They factor that knowledge in, alongside knowledge of their own strengths. So I pulled the exact same data, but this time for pitchers. That left me with a sample of 707 players. Take a look at the scatterplot of their chase rates:

This time, there is a negative correlation: -.34. Pitchers are more likely to be either good at inducing chases above the zone or below it, and less likely to be good at both. There are a few outliers at the top right, but not many toward the bottom left. That’s because a pitcher who can’t get chases anywhere isn’t likely to last long enough to make our 4,500 pitch cutoff.

This is pretty intuitive stuff. 66.6% of the chases above the zone come on four-seamers. Sinkers and changeups make up 32.6% of the chases below the zone. There aren’t that many pitchers who feature a great swing-and-miss four-seamer as well as a great swing-and-miss sinker or changeup. For the most part, the human arm tends not to work that way.

However, I’d like you to note that the scatterplot for hitters is much more spread out than the scatterplot for pitchers. Not a single one of our 707 pitchers has an overall chase rate above 50% or a chase rate above the zone that’s lower than 18%. Largely, this is because chases (and their old friend, swinging strikes) are more important to pitchers. There are plenty of hitters who can compensate for poor plate discipline through elite bat-to-ball skills or light tower power. Pitchers who can consistently induce weak contact in the zone are much harder to find.

Lastly, let’s take a look at whether pitchers are actually acting in the way that Smoltz indicated. Are pitchers throwing more pitches above the zone to batters who chase more there? Or are they worrying less about the batter and just playing to their own strengths? I calculated the correlation coefficient between chase rate in each location and the percentage of pitches thrown in that location:

That’s pretty clear-cut. The batter’s chase rate is a small factor, but pitchers are much more likely to go with whatever works best for them.

Interestingly, for pitchers, chase rate isn’t always the stat that correlates most strongly with the percentage of pitches thrown in a particular location. Above the zone, there’s a .45 correlation to swinging strike rate. (Quick reminder: whiff rate is the percentage of swings that result in whiffs, but swinging strike rate is the percentage of overall pitches that result in whiffs). This makes sense. If a batter puts the ball in play above the zone, they’re more likely to elevate it and hit it out of the ballpark. But below the zone, any kind of chase will do. Even if the batter puts the ball in play, they’re likely to hit it straight into the ground. Balls in play below the zone have a wOBA of .279, as opposed to .303 on pitches above the zone.

I’m sure that the pitchers are looking at the data Smoltz was talking about. At the very least, for a batter or two a night, it could prove to be very useful. Ozzie Albies chases 59% of pitches below the zone and 19% of pitches above the zone. It would be downright negligent not to take that kind of split into account. However, it’s pretty clear that more pitchers are dancing with the one that brung them.

That may well have been different in Smoltz’s time. Think of all the pitching trends we’ve been talking about over the last several years. Pitchers de-emphasizing their fastball and throwing their best pitch more often. Pitchers trusting their stuff and daring hitters to hit it. It makes sense that today’s pitchers might be less likely to tailor their games to certain batter tendencies than Smoltz and his peers did. I wish I had access to this kind of data going back to the 1990s. It might well show us yet another quiet way that the game has evolved.