Major League Baseball released an internal memo last month indicating they were making a change in the physical construction of the official baseball. Their goal was to deaden it by increasing its drag, thus decreasing the distance balls travel when hit, with the ultimate effect of reducing the number of home runs.
In Part 1 of this series, I described a probabilistic model designed to estimate how a deadened ball would affect team-level offense as well as how it would affect individual hitters. I then zoomed in on the Dodgers, specifically. Of course, every home run taken away from a hitter by a deadened ball is also a home run not allowed by a pitcher. With only a few minor tweaks, the same model can be used to predict the effect of a deadened ball on pitchers. If you’re interested in how the model works, read part 1.
Pitchers have to be happy about reduced home run rates, right? Of course they are! Dodgers pitcher David Price voiced his … um … feelings in a tweet.
Frustrated with the ever-changing ball as pitchers may be, deadening it in 2021 will likely lead to tangible run suppression. How will the Dodgers pitching staff benefit relative to other teams? Which individual Dodgers pitchers will benefit the most? Those are the questions I hope to answer here.
As a team, the model thinks Dodgers pitching staff would allow 9.38% fewer home runs with a deadened ball than they would with the baseballs used in 2019-2020. That’s the 5th largest predicted percentage drop in the majors. For individual pitchers, I’ll break down the analysis by starters and relievers.
The Dodgers have 3 pitchers on their 2021 roster who threw at least 200 innings between 2019 and 2020: Clayton Kershaw, Walker Buehler, and Trevor Bauer. All three of them stand to allow at least 8% fewer home runs with a deadened ball (according to the model). I’m not sure why, but there’s an assumption out there in the Dodgers-baseball-observing universe that Kershaw will see an outsized benefit from deadening the ball. Although the model predicts Kershaw will see a decent sized benefit, allowing 8% fewer home runs, it thinks Buehler (9.94% reduction) and Bauer (10.9% reduction) give up more marginal contact that a deadened ball will convert from home runs to fly outs at the warning track.
Though Kershaw, Buehler and Bauer will surely represent the top of the rotation in 2021, the Dodgers have enviable depth behind them in Julio Urias, Dustin May, Tony Gonsolin and Price. In the plot below, I’m showing the number of home runs allowed that a deadened ball would have taken away between 2019 and 2020 (x-axis) as well as the percentage reduction (y-axis) for each of the seven likely 2021 Dodgers starters.
Between 2019 and 2020, Bauer allowed a team-leading 43 home runs. He also threw a team-leading 286 innings. Of course, more innings pitched means more opportunities for opponents to hit the ball over the fence, so the total home runs reduced is not quite as informative as the percentage reduction. As you can see, the deadened ball is predicted to affect Bauer, Buehler, Price and Urias at a similar rate — between 10% and 11% reduction in home runs.
Kershaw, May and Gonsolin are in somewhat distinct tier from the other starters, as a deadened ball is predicted to subtract between only 7% and 8% of their home runs allowed, which is lower than the team average. Why is this? When Kershaw slips up and allows good contact, it tends to be better contact than the good contact Buehler (for example) allows. In other words, the home runs he allows are rarely just over the fence, they’re usually crushed. The same is true of Gonsolin and May … for now. With younger starters, there tends to be more room for development. Both Gonsolin and May have worked quite a lot in the offseason and I’m curious to see how this manifests itself in terms of contact quality allowed.
Predicting reliever performance is a fraught affair. There’s a lot of variability season-to-season and because they pitch fewer innings than starters, one bad outing or one month of a nagging injury can ruin an entire stat line. Those same limitations are present here, when predicting who benefits from a deadened ball, so take these predictions with an extra grain of salt as I offer them with a dose of humility. The plot below represents exactly the same data as the plot above, but now only shows relievers on the 2021 roster.
Kenley Jansen had a fly ball rate allowed of 45.1% between 2019 and 2020, which is 41st most out of the 416 pitchers who threw a minimum of 50 innings. He tends to limit quality contact, but (of course) doesn’t always. Given the type of contact Jansen tends to allow, the model thinks he stands to give up 11.3% fewer home runs with a deadened ball than he otherwise would. Over the course of a season, that’s probably only 1 home run, so although these changes are important to bullpen performance in aggregate, it will be difficult to notice in any individual outing.
I’m not sure why, but I was under the (very, extremely, amazingly) false impression that Joe Kelly and Brusdar Graterol were fly ball pitchers. Probably psychological scars from the postseason. Anyway, they are decidedly not fly ball pitchers. Both have a ground ball rate over 58% and both have a fly ball rate under 22%. Nonetheless, both of them stand to benefit about as much as Jansen does. How is this possible? Because when they do give up a fly ball, the contact tends to be good. Sometimes it’s only marginally good, though, and a deadened ball is predicted to convert some of those batted balls from home runs to fly outs.
Victor Gonzalez had a phenomenal rookie season in 2020. He went on to post a 1.67 FIP over 20.1 innings pitched, allowing a grand total of zero home runs. The model thinks a deadened ball will save him from 7.53% of home runs allowed. You can’t allow negative home runs so how does the model project a reduction from zero? Well, the model is designed to examine expected home runs (xHR) from different baseball constructions. Remember, xHR is a probabilistic model and it thinks when you add up the home run probabilities of contact against Gonzalez, he would have allowed 0.69 home runs in 2020, but with a deadened ball he would have allowed 0.639 home runs. The advantage of this model is that it allows you to make predictions from only a small quantity of data. Gonzalez is set to become an important piece out of the pen and as he accumulates more innings, the model thinks a deadened ball might just save him from a home run or two.
Blake Treinen was Dave Roberts‘ go-to high leverage reliever in 2020. He also pitched more innings out of the bullpen last season than anyone on the Dodgers’ roster. His home run reduction profile is similar to Gonzalez at 7.7%. That may seem like a small number but it matters when you’re accumulating innings the way Treinen did in the 2020 shortened season, and matters even more in a full season.
Scott Alexander is a left-handed specialist known for inducing ground-ball double plays. Though much like Kelly and Graterol, when the batter he’s facing does launch a ball into the air, it tends to be on good contact. If he can’t induce a double play, a deadened ball might just get him out of a sticky situation. He’s projected to allow 9% fewer home runs.
You may have noticed that I chose to omit several pitchers (Jimmy Nelson, Corey Knebel, Brandon Morrow, and a few others) because they were significantly injured, are unlikely to make the major league roster for significant innings or did not have enough recent batted ball data to make confident projections. I chose to focus on these players because they are all probably going to get a lot of innings in 2021.
In conclusion, projections are difficult and can never be perfectly precise, but I do think the use of a model can help shape expectations. In general, I do believe Bauer, Price, Jansen, Kelly and Graterol will benefit from a deadened ball more than the rest of the pitching staff. Others will also benefit, but probably not quite as much or as often.
It remains to be seen just how accurate the model is, but I’m excited to test against on-the-ground results. More than anything though, I’m just excited for the 2021 season to get underway so we can stop looking at probabilistic models and start looking at real baseball.