New metric tracks how well (or poorly) pitchers prevent basestealing

Running game prevention starts on the mound

November 11th, 2023

If the new rules put into place in 2023 were intended to help return some action to the basepaths, then they succeeded. Runners attempted to steal 1,072 additional bases in 2023 than they did in '22, and they were successful 80.2% of the time, the highest mark in AL/NL history. (They were even more successful in the playoffs, at 82.6%.)

There are plenty of reasons to think that this is just the start. That much of Ronald Acuña Jr.’s incredible season can be chalked up to his willingness to run. That we’re going to see a whole lot more of this next year. If so, that means skills that had declined in importance – catchers' ability to throw runners out, and pitchers' ability to keep them from running in the first place – have suddenly regained importance.

We know, using Statcast data, who the best catchers are at preventing stolen bases given the opportunities presented to them. Today, let’s talk about the players helping to create those opportunities. Let’s talk about the pitchers who do a great job of holding runners on, and those who do not.

“The pitcher, if he doesn't get the ball to the catcher quickly,” as legendary basestealer and long-time coach Davey Lopes once said, “the catcher doesn't have a chance.” It’s been 10 years since FanGraphs wrote that “surface value says that a pitcher’s quickness to the plate is a whole lot more influential than a catcher’s arm in the battery dynamic.

Let’s find out. The pitcher running game prevention leaderboards and associated data are now available at Baseball Savant (as of now, this only shows steal attempts from first base to second with no other runners aboard). Here are nine things we’ve learned.

1. What are we measuring here?

It’s a skill to keep runners close, and it’s a separate skill to prevent them from running wild while you release the ball because you’re doing it so slowly. The best skill of all is to do both.

On a pitch-by-pitch basis, we can take into account a variety of information, including the speed of the baserunner and how often he steals. Based on those inputs, you can output a given situation's estimated likelihood of success before you even know if the catcher throws out the runner.

With a number on each included pitch, we can break down pitcher performance against runners advancing from first to second (on steals or balks), how much they contribute to outs (on caught stealings or pickoffs), or from preventing runners from trying to go at all (holds). Combine all that, and we can output a season-long number of base advances allowed, or prevented.

So, for example, take this Gleyber Torres steal off of Bobby Miller in June. Miller didn’t bother to hold Torres, so Torres was already off to a full sprint before the pitcher even moved. (Torres was a wild 38 feet off first base by the time Miller released the pitch.)

Dodgers catcher Will Smith was presented with an impossible opportunity, and didn’t even make a throw. From his point of view, it was a 0% chance, and so he’s not penalized at all for the stolen base allowed. But for Miller, he did so little to prevent the steal, he allowed 99% of the base, or -.20 of a run towards his seasonal total, as an advance allowed is given a run value of -.20, and he allowed nearly all of that base. (These will not always add up to exactly 100% for a variety of reasons.)

Sometimes, a pitcher can do a great job and still get a positive review here even if the runner is successful, because preventing a steal is a team effort. Consider this pitch from Washington’s Jake Irvin, with Chicago’s Nico Hoerner running. Hoerner was successful without a throw, and that counts as a steal against Irvin, except that he did his job well, and catcher Keibert Ruiz simply dropped the ball. Despite the steal, Irvin gains +.20 runs here, because the catcher's mistake has nothing to do with the pitcher.

Pitchers also get credit for runners not going, because that’s a skill, too. We’re talking tiny, tiny fractions of value, but they can add up. In 2023, White Sox starter Dylan Cease threw the most pitches with a runner on first without the runner going, 731.

2. Who was good – and not – in 2022 and ’23?

Data goes back to 2016; let’s combine the last two years for a more recent leaderboard. The best, in terms of base advances prevented in 2022-'23, is an interesting and somewhat eclectic group.

  • +14 David Peterson
  • +10 Zack Greinke
  • +10 Patrick Sandoval
  • +9 Kyle Bradish
  • +8 Zac Gallen, Dylan Cease, Cole Irvin, several others

We’ll dig further into some of these below. The bottom of the list, pitchers who allowed more bases than expected, is made up of names you may have guessed. Any list without the notoriously weak (in this area) Syndergaard and Ottavino at the bottom would be surprising.

  • -26 Noah Syndergaard
  • -18 Adam Ottavino
  • -16 Michael Kopech
  • -15 Logan Webb
  • -15 Aaron Nola
  • -15 Kevin Gausman

We can go deeper, too, with a variety of looks at runner-lead distance. As you’d expect, there’s lefties at the top.

Who holds runners to the largest (or smallest) leads at the first move on steal opportunities?

  • Best: Wade Miley, Jesus Luzardo, and Brent Suter: 8.7 ft
  • Worst: Ottavino: 10.4 ft

Who holds runners to the largest (or smallest) leads at pitch release on steal opportunities?

  • Best: Wade Miley, Max Fried: 9.6 ft
  • Worst: Ottavino, 16.8 ft; Syndergaard: 15.8 ft

A gap of seven-plus feet! Here’s what that looks like, using individual pitches against each pitcher as an example.

But it’s not just about the lead distance, either. Because some pitchers are fast to the plate, and others quite slow, we can look at the distance that runners can travel between the first move of the pitcher and the pitch release.

Who allows – or prevents – the most/least lead distance gained between first move and pitch release on steal opportunities?

  • Best: Brooks Raley, Fried: 0.5 ft allowed
  • Worst: Ottavino: 6.4 feet allowed; Syndergaard: 5.7 ft allowed

We talk about steal attempts coming down to fractions of a second or mere inches between success and failure. But between the best and worst here, we're not talking about inches or seconds. We're talking feet.

3. What makes David Peterson so good at this?

The Mets lefty is the best over the past two years, and the best over the past four years, too. Unfortunately, he recently underwent hip surgery, and is likely to miss some of the 2024 season.

It’s partially because even when he allows steals, he’s accruing some positive value. When Brendan Drury stole a base on him in July of 2022, Peterson did just fine on his end of it, earning +.32 of a stealing run. His catcher … did not:

It’s partially because he gets more outs than expected (+4 bases prevented on outs), or fourth-best, and because he does OK on holds, too (+9), though he’s more good than great there. It basically comes down to the fact that when runners do go, they do poorly – likely in part because he’s got one of the smallest lead distances allowed on steal attempts in the game.

4. Was it really that easy to steal against Syndergaard & Ottavino?

Yes. Syndergaard has long been notoriously vulnerable, and Ottavino’s issues are a long-known weakness of his that he’s been open about discussing. For one thing, going back to 2016, Syndergaard has five of the seven weakest seasons in this regard. In 2023, he had the highest attempt rate of any starter on record; only reliever Kenley Jansen, who’s long had his own issues in this regard, motivated more attempts.

At one point this year, Syndergaard allowed former teammate Mookie Betts to successfully steal on a pitchout. Because Betts was so far off the base when the ball finally got to catcher Bo Naylor, he only faced an 11% chance to throw the runner out (which is to say, very difficult), while Syndergaard had to eat -0.13 of a run towards his total.

5. A special note for Johnny Cueto, the legend

Cueto, 37 years old, got into 13 games for the Marlins in 2023, with a 6.02 ERA that may or may not spell the end of a long and successful career. Despite allowing a ton of baserunners, he didn’t allow a single steal. No one even tried to steal against him, though he did pick off Houston’s Chas McCormick in August.

Our data goes back to 2016. Put those eight years together, and make a leaderboard out of it. No pitcher in that time has rated better than Cueto, who prevented 30 base advances, worth 8 runs, or nearly 1 win based just on this one skill, which validates an ESPN study from nearly a decade ago that suggested his skill was real. The shimmy works!

It’s not hard to see why, either. Cueto allowed runners a lead of 8.8 feet at his first move on steal chances, tied for the lowest of 269 qualified pitchers. On steal attempts, it was only 9.4 feet, also the lowest.

“[Johnny is] an Olympic-quality baseball player. The things he does beyond throwing the ball are so top of the line,” said then-Cincinnati manager Bryan Price in 2014. “Like holding runners, or his move to first base.”

Considering that he made 226 starts before our data begins in 2016, it’s likely that we’re underselling him by a lot here.

6. What was really happening with Jon Lester?

Remember back in the glory days of the Cubs, when a huge topic of conversation was Lester’s all-but-total inability to complete a pickoff throw? In 2017, he attempted only one throw all year, making it a huge deal when it actually resulted in an out against Tommy Pham. Even at the time, we knew the leads were enormous, yet he never did get abused by runners as much as you’d have thought. What would this data have said about Lester, who retired in 2021 after a successful 16-year career?

It would have said exactly what you’d have thought, but maybe not for the reasons you’d have expected.

Lester, from 2016-21, was poor at this, preventing 29 fewer base advances than expected (which works out to -4 runs), tied for eighth-weakest in that span. The leads he allowed were indeed huge.

  • 11.2-ft. lead at first move on opportunities was most, by a whole foot
  • 13.2-ft. lead at first move on steal attempts was most, by half a foot

But the trick here is that runners really just didn’t go wild by as much as you’d think, and if they’re not going, then by how much they got off the base to not go doesn’t really hurt him – like when Enrique Hernandez was taking 20-foot leads and then staying put.

Lester’s -29 breaks down, to start, with -10 on outs, which is the worst of any qualified pitcher, which sounds weird – a negative score on outs! – but is telling you that his catchers saved his bacon more than a few times, like this 2016 play on which David Ross threw out a runner he had absolutely no business throwing out, from his point of view (8% estimated CS). Runner Jake Elmore was out, but Lester got a negative number anyway.

He also got -36 bases by allowing advances, which is bad but not awful (13 pitches rated weaker), and then gained +17 bases back by all the times those runners held and didn’t try to go. He was below-average at all this, quite obviously. It just wasn’t quite in as game-changing a way as you’d have thought.

In addition, for as much talk as there was about what would happen after his personal catcher Ross retired following 2016, runners took the exact same leads, to the inch, against Lester regardless of whether it was Ross or Willson Contreras behind the plate.

7. How did Greinke improve so much?

Greinke, one of the most interesting and intellectual pitchers of his generation, posted a 5.06 ERA in 30 games for the Royals in 2023, his 20th Major League season. It wasn’t a great season, overall, except for this: He suddenly got really, really good at controlling the running game. From 2016-’22, he rated as above-average, posting +12 base advances prevented. In 2023 alone, he was +9. He’d had six pickoffs in the previous four seasons. He had five in 2023. 

Early in the season, he picked off Acuña, of all people.

In true Greinke fashion, this does not seem to have been an accident. Remember back in Spring Training, when Greinke was hit hard by the D-Backs, then implied he wasn’t really mad about it?

“I’ve been complaining, actually, that people haven’t been getting on so I can’t work on pickoff moves all spring,” Greinke said. “So wanting to work on pickoffs. That’s what Spring Training’s usually for, stuff like that.”

It sure seems to have worked.

8. Taijuan Walker is the king of pickoffs.

Ten years ago, “excellent pickoff move” even made its way into Baseball Prospectus’ writeup of Walker as a rising Seattle prospect. That was a tremendous call, and it’s proven to be true. Over the last two years, he’s created more outs (18) than base advances allowed (17) in these qualifying situations, in large part because he leads all pitchers in pickoffs here.

Yet: Overall, Walker rates as more ‘solid’ than ‘great.’ Why? Because he just gets more of everything, good and bad, than most everyone else. Yes, Walker contributes to more outs on the bases than most, but at the same time, he allows plenty of advancements, too. (Sometimes on a balk, like this example.) He’s strong on holds, not elite. The pickoffs are incredible, yet they're not the only thing that matters here.

9. Is there any strategy in what happens after multiple disengagements?

In the first year when pitchers were limited to two penalty-free disengagements before facing a scenario where they could have a third, but it would be considered a balk if the out wasn’t made, we got a chance to see what would happen

What we learned, to some extent, is that we didn’t even get to two disengagements very often; there were something like 1% as many pitches thrown with two disengagements (in our runner on first, no other runner scenario) as there were with zero. But from what we can show, there’s some clear movement in the direction you’d expect. As runners get deeper into disengagements, they take bigger leads. They attempt to steal more often, and advance more successfully, though by a very small margin.

  • 0 disengagements
    • 1.4% attempt // 9.6 ft lead at first move // 77.7% advance rate
  • 1 disengagements
    • 5.3% attempt // 10.4 ft lead at first move // 79.4% advance rate
  • 2 disengagements
    • 10.4% attempt // 11.1 ft lead at first move // 79.0% advance rate

The full leaderboards, based on the work of’s Tom Tango, Jason Bernard, Dana Bennett, David Adler, and others, are available at Baseball Savant.