Are Walker Buehler’s Run Suppression Gains for Real?
Walker Buehler has been on a tear in the second half, leading all pitchers in WAR since the All-Star break at 1.8, just a hair ahead of Adam Wainwright, Frankie Montas, and Max Fried. For the season, he now ranks fourth in WAR among qualified starters (fifth if you include Jacob deGrom’s 92 preposterous innings), toting a 26.9% strikeout rate and a park-adjusted ERA 47% better than league average. His 2.11 ERA is more than half a run better than any other season in his career, and the park adjusted figure is his best by 15 points. It’s another great season from one of the consistently best pitchers in the majors; since becoming a full-time starter for the Dodgers back in 2018, Buehler has posted the 11th-most pitching WAR with the 19th-most innings pitched.
I am not breaking any news by pointing out that Buehler has been and continues to be excellent. The surface-level numbers indicate he has never been better. What caught my eye, however, was how he has gone about doing that. There have been some noticeable changes under the hood. For example, at age 26, he has lost over a full tick of velocity on his fastball compared to any prior season; based on my own research, you would expect a player of his age to lose only about 0.15 mph.

Buehler has also lost almost two percentage points on his strikeout rate. On its face, missing less bats and losing velocity is never something you want to see in a pitcher. (That said, he has brought his ground-ball rate back to pre-2020 levels, going from 35.5% last season to 43.8% this year, similar to his ’19 rate as well as a tick above the rest of the league in 2021.)
It is fair to wonder if Buehler has taken a small step back despite his stellar results, as players that lose velocity year over year generally also lose effectiveness in terms of run suppression and strikeouts. I took every season in the pitch tracking era for starting pitchers who threw at least 100 innings, then paired the seasons (each season had to have 100 innings pitched), took the changes in velocity, ERA, FIP, and strikeout rate, and charted them. Here’s where Buehler falls in terms of his velocity decrease:

Buehler’s velocity change falls on the wrong end of all three curves. These curves describe the starting pitching population in general, however, so he could be attacking hitters differently to compensate for his velocity dip. Looking at thousands of pitchers is not exactly conducive to potential edge cases.
This is not to say he is not among the league’s elite; he most definitely is. What I wanted to figure out is if Buehler is being propped up by some batted ball luck or if he has changed his approach at the expense of strikeouts. To that end, he has made some tweaks to his pitch mix, possibly in response to the lack of velocity, and is leaning more on his slider and cutter, as well as incorporating a changeup (mostly to lefties), at the expense of his four-seamer.
| Season | CH | FC | FF | KC | SI | SL |
|---|---|---|---|---|---|---|
| 2018 | 3.7 | 8.0 | 40.7 | 13.8 | 18.6 | 15.2 |
| 2019 | 0.5 | 13.0 | 53.2 | 12.0 | 7.0 | 14.3 |
| 2020 | 0.0 | 12.8 | 53.9 | 17.4 | 7.7 | 8.2 |
| 2021 | 3.4 | 16.4 | 45.6 | 13.6 | 7.4 | 13.6 |
This change makes sense; his fastball has not been as effective an offering as it has been in the past.

The swinging-strike rate on Buehler’s fastball has steadily decreased over time, hitting a nadir in 2021, even with more swings (as a percentage of pitches) and more chases than last season. The curveball has also seen a noticeable decline in performance across the board, and I wonder if this is a result of using his fastball less often and to lesser effect, as the curve tunnels off of it. The real boons to his arsenal have been the slider and changeup, the latter of which he did not use at all in 2020 but has gotten swinging strikes out of almost 33% more than average. The slider, meanwhile, has been especially potent in inducing chases.
Buehler is also varying the locations at which he throws each of his pitches. His zone rate is down a fraction of a percentage point from 2020, and his chase rate has increased by 1.5%, a promising trend. He has also yielded strikes on the first pitch 65.7% of the time, going from slightly below-average last year to firmly above the league mean. Using the Gameday zones defined by MLBAM (which you can find in the Baseball Savant search tool), you can see Buehler is throwing a higher percentage of his pitches up in the zone (zones one, two and three), mostly with his four-seamers (a 5.7 percentage point increase year over year) and cutters (10.6 percentage points), and has thrown more pitches down and off the plate, mostly with his slider.
One thing that cannot be ignored and that is likely behind both using the four-seamer less and the pitch’s precipitous fall in whiff rate is a change in spin rate following enforcement of the sticky substance ban. Spin rate is partially a product of velocity, so Buehler’s decrease in that stat this season should not come as a surprise. But even when you account for this relationship (by dividing spin rate by velocity, which I will refer to as spin-velocity ratio, or SVR), the change is still visible.
| Time Period | SVR |
|---|---|
| 2018 | 25.1 |
| 2019 | 25.5 |
| 2020 | 26.4 |
| 2021 Before June 3rd | 27.6 |
| 2021 Since June 3rd | 25.1 |
You can see a sharp decline in SVR right after reports emerged that the enforcement of sticky substances would begin in earnest on June 3:

That date is marked with the red dashed line; both the ranges and averages by start began to fall after that point.
What is most surprising is that Buehler seems to have improved despite this. After a brief dip around the date where the enforcement was reported, his strikeout rate has steadily climbed as the season has gone on:

Buehler’s adjustment to his repertoire post-June 3 is even more stark than the year-over-year changes, as he has accelerated his use of his breaking pitches and changeup:
| Time Period | CH | FC | FF | KC | SI | SL |
|---|---|---|---|---|---|---|
| 2021 Before June 3rd | 0.7 | 12.4 | 50.8 | 15.3 | 8.8 | 12.0 |
| 2021 Since June 3rd | 5.4 | 19.4 | 41.7 | 12.4 | 6.4 | 14.8 |
The same holds true when split by handedness, where he has relied on his changeup more against lefties (6.3% versus lefties versus one lone changeup to righties) and increased his usage with the slider against right-handers to roughly 35%.
Buehler has improved as the season has gone along, even in the face of his spin rate degradation, and while losing velocity is never a positive development, he has compensated with increased use of his excellent breaking pitches as well as the return of his changeup. He’s also been aided by a .239 BABIP and a 9.9% HR/FB rate, compared to a 13.0% mark pre-2021. That mostly explains the 1.02 gap between his ERA and FIP, the largest among qualified starters this season. Some of this batted ball fortune should be attributed to Buehler, though; he has allowed below-average figures in terms of both barrel rate and hard-hit rate, as well as a .340 wOBAcon versus a .370 rate for the rest of the league.
That .239 BABIP is a bit low even for a pitcher displaying great command, as is the home-run rate, which is about a third lower than the major league average. But Buehler has been locating in such a way that we should expect his results on contact against to at least be a little below average. Pair that with a great strikeout rate, and you have one of the best pitchers in the league. Is he a 2.11 ERA pitcher? Probably not, but he has shown the ability to adjust. That will be essential if the Dodgers hope to avoid the dreaded wild-card game, win the NL West, and defend their title.
Carmen is a part-time contributor to FanGraphs. An engineer by education and trade, he spends too much of his free time thinking about baseball.
“…just a hair ahead of Adam Wainwright…” Whoa, hold up. Isn’t he 39? Does that guy age at all? By ERA and WAR, the last time he had a season this good, they were still making Boardwalk Empire, Kitchen Nightmares, How I Met Your Mother, and the Colbert Report.
He looked like he might wash out of the league, then this. I think a lot of pitchers have found new success with the new baseball.
Buehler is the NL Cy Young right now as far as I’m concerned, and only Wheeler can dispute him. The most impressive thing about him this season, imo?He’s pitched at least 6 full innings in all but one of his starts this season, and is currently either on pace for 200+ IP or just below it. The velocity decrease kinda jumps out, but part of me thinks he’s simply pacing himself through the season, and I bet we’ll see triple digits when October rolls around.
Buehler can show high velocity variance inside a start (if you go watch the pitch-by-pitch of his NLCS game 6 vs the Braves last year, he routinely added 2-3 mph in the high leverage PAs)
He has talked this year a few times about how he’s trying to pitch at ‘95%’ in order to be more efficient with less deep counts.
With that said, in his high leverage PAs you see a bit less peak velocity (it’s hard to find 98-99 like we saw just last year in the playoffs at times, for example) and the spin changes are real.
The Dodgers as a staff and their starters in particular have a long-running, multi-year trend of beating their FIP by a decent margin: as a pitching staff, the Dodgers have beaten their FIP by 0.36 runs *since the beginning of 2017*, and their starters have done so by even more than that – 0.42 runs over that span. That’s over 5 years and innumerable starting pitchers. I don’t exactly know what to attribute that trend to, but I don’t think you can attribute it to luck over a 5 year, 3500 inning, 15+ player sample.
The over-performance of their FIP is really interesting. At first, I assumed it was related to strong defensive performance (FIP is “Fielding Independent Pitching”, after all), but what started as a quick glance revealed something more interesting.
By UZR, the Dodgers have had a roughly average team defense since the start of 2017 (-0.8 UZR/150, good for 17th in the league, so actually slightly below average). However, by DRS, they are ranked 2nd in the league over that time, trailing only the Astros and well ahead of the Cubs in 3rd. Even more interestingly, the Astros have an even worse UZR over that timespan, with a UZR/150 of -1.1, 21st in the league, and have also outperformed their FIP fairly meaningfully over that timespan, beating it by 0.13 runs. For whoever is curious, the Cubs rank 3rd in the league in both DRS and UZR and have also outperformed their FIP substantially, so that one makes sense.
In short, both the Dodgers and Astros have beat their FIP by a decent margin over the last 5 years, a large sample size, and rank head and shoulders above the league in DRS over that time. However, UZR thinks that both were not only not top-of-the-league good, but actually below average defensively over this same span. I’m not sure what to take away from this, but would be curious if anyone has any theories.
Efficient shifting perhaps? Iirc UZR doesn’t take shifts into account no?
Yeah, I think it’s shifting to compensate for some dodgy (you’re welcome) defense. This season, especially, the defense in LA has been subpar (no doubt related to the injuries).
Yeah. this is something I’ve commented on before, and fwiw the Dodgers (and Astros to a lesser extent) also tend to have some of the lowest BABIP allowed in MLB year after year, suggesting they really are very good at turning balls in play into outs, regardless of how it shows up in individual defensive metrics.
Its probably a micro-positioning thing more than anything else, which is something they have long been at the forefront on back to 2015 and the controversies about marks in the outfield to use their positioning cards. They lead the industry in that area, and the circumstantial evidence is that they still do.
Well the obvious inference is DRS is more meaningful than UZR.
The inference is that DRS is more meaningful to TEAM defense than UZR.
The Dodgers have the lowest BABIP during that period, but that’s probably just another reflection of the same thing which is causing the low ERA compared to FIP. It isn’t good HR luck or some ballpark effect because the xFIP/ERA difference is even greater. The only thing I can see if they have the highest infield fly % in the majors during that time period, which will help ERA relative to FIP (because infield flies are essentially Ks although aren’t counted as such) and lower BABIP, but I don’t know if that’s enough such a big FIP/ERA difference.
Okay, here’s something weird: The Dodgers’ ERA/FIP overperformance is almost completely at home. For 2017-21 the home difference is .60 while on the road it’s only .12. That kinda looks like some kind of ballpark effect (if we assume it isn’t luck). Or could the Dodger stadium scorer be giving a lot of errors with the result that ERA is unrealistically low? We could check that by seeing if the Dodgers have better fielding stats on the road, but unfortunately it seems that Fangraphs doesn’t give home/away splits for fielding (does it?).
Or because of their familiarity with the park they are better at positioning at the home park.
How would that work?
Just throwing out a theory. They tried, but were ultimately not allowed, using marks in the outfield at visiting parks so the could be more accurate in the positioning. Perhaps they have landmarks at home that make it easier for them to be more accurate. Is it a stretch? Maybe, but I’m just throwing out ideas.
Not only has he gone 6 innings in all but 1 start, but he has a QS in all but 3 starts, highlighting incredible consistency. Roberts is even letting him go in the 110 pitch range, so clearly the kid gloves are off. There definitely seems to be something deliberate about trying to go deeper into starts, whether it be innings or pitch counts, and I think Carmen’s analysis highlights some data that could be seen as evidence towards that hypothesis.
“Buehler is also varying the locations at which he throws each of his pitches. His zone rate is down a fraction of a percentage point from 2020, and his chase rate has increased by 1.5%, a promising trend.”
I’m interested to hear from someone who knows more (never having been a pitcher myself), but could the decrease in fastball speed be a tactical choice? Perhaps he and the coaching staff identified that if he shaves a bit off the speed he gets a more control and that’s more valuable to his style, and the above quote supports this hypothesis.
Also put some legends on those plots!
Buehler’s talked about this in interviews: his goal this year was to pitch into the 7th inning of every game, and if that meant losing some Ks to ground balls, he was fine with it (as Jaffe and Fabian Ardaya have mentioned, as well). So it’s definitely intentional.
You also see it in his pitch selection. He faced the Mets back to back recently, and in his second start, used the CH and CU far differently. He’s definitely got a gameplan this season, and while this might be the 90th percentile outcome of it, I think you have to give him credit for it.
I also wonder if he looked around at Flaherty, DeGrom, etc. and thought, “I’ve already had one TJ; maybe if I don’t throw *quite* as hard but still get outs, I can avoid another surgery?”
Hopefully down the stretch, in anticipation of the playoffs, Buehler will get a day off.
Seeing ggplot2 graphs makes me happy. Nice work–it’s interesting to see how he’s making up for lower velo by changing his pitch mix and location. My only suggestion is with the K% by date graph–is that some kind of GLM you fit? Why not just use a moving average plot to show increasing K%? But nice work overall!