Update on pitch location data

Building on the last post

In the last post I showed the expected FIP based on pitch location only, for fastballs and non-fastballs. For the most part it looked as expected. Throwing pitches outside the strike zone or right down the middle is bad. Throwing to the edges of the strike zone is good, and throwing to the corners of it is even better.

The question I had in mind when I started this was how much of a pitcher’s success can be attributed to his location alone. The answer is not all that much.

Let me back up and show how to calculate location-based FIP. Here are the zones as defined by Baseball Savant.

Zone 10 is missing. It represents an extradimensional changeup first perfected in 1946 by Bugs Bunny in a game against the Gashouse Gorillas. It has rarely been thrown since, although both Greg Maddux and Pedro Martinez occasionally made use of the pitch. It was last thrown on October 16, 2009, by Martinez to strike out Manny Ramirez.

For each section of the strike zone, I look at how many pitches are swung at and missed, taken for strikes, fouled off, taken for balls, or put into play. I also look at a subset of those put into play that are especially hard hit, the barrels. For this exercise, I am defining a barrel as any hit that has an expected WOBA of .800 or greater.

Based on where a pitcher throws all of his pitches, I get an expected total of swings and misses, balls, etc. Then I came up with this formula for turning the expected pitch result data into a stat that is on the same scale as FIP, or ERA.

LocationFIP = (barrels * 67 + balls * 3.5 – swing&miss *10 – strike taken *4)/Pitches +3.20

This is a rough estimate. It does not take into consideration what the count is. For example, a slider low and away is a much better pitch to throw when you are ahead in the count than when you are behind.

With the league average of 4.18, the very best pitchers have a location-FIP around 4.00, and the worst are around 4.40. There isn’t all that much difference between the top and bottom pitches in the overall mix of zones they throw to. The differences in pitcher effectiveness will have to be found in 1) difference in the quality of their stuff and 2) matchup and count-specific ability to locate pitches.

I can use the same formula above, and call it pitchFIP, to create a comparison. This does not look at the location of the pitch, just the pitch result. How many swings and misses, called balls, fouls, barrels, etc. do you get? And what is the expected runs from that?

The correlation between locationFIP and pitchFIP is 0.127 with an average of 400 pitches thrown. It’s something, but a small part of what makes pitchers good or bad.

Here are the leaders in pitchFIP:

Miller is no surprise, he’s striking out 2 batters per inning. Misiorowski and Ohtani are guys we all know about. The others are less known, but are the latest guys putting up impressive numbers in one-inning bursts out of the bullpen.

Great to see Sam Bachman in there, one of the few bright spots in an atrocious Angel bullpen. He can be a bit on the wild side but has done a great job of striking people out and avoiding hard contact.

Here are the trailers:

Walker is no surprise, he wasn’t getting anyone out before the Phillies cut him. Scherzer has looked like father time has caught him. Or maybe he was just hurt and has a bit left - we’ll find out soon.

Sugano is a real surprise towards the bottom because he has pitched very well in Colorado. This could indicate that he’s been really lucky so far. He gives up a lot of homers and rarely strikes anyone out. He is not pitching in a place that is very forgiving of traits like that.

Here are the pitchers with the best location FIP:

Overall, a solid group of pitchers. The trailers:

Some of these guys are struggling, but there are some good pitchers in there too. Alvardo is a weird one. He’s not hitting the most effective zones, but his walks are not too high and he’s getting a lot of swing and miss. Throwing 98 MPH certainly helps. His FIP (3.55) is in line with his pitchFIP. Alvarado is not getting any help in the field though, and a .429 BABIP has stuck him with a 5.64 ERA.