Aaron Nola, one of the best free-agent pitchers available, has a problem. Or, at least, he had a problem. But we’ll get back to that.

The problem is easy to identify. Over the past eight seasons, only four pitchers were weaker at holding at first base  those baserunners who had a chance to steal second. That's according to the newest Statcast metric, which tracks how well a pitcher prevents runners from advancing from first to second (on steals or balks) and how much they contribute to outs (on caught stealings or pickoffs), or from preventing runners from trying to go at all (holds).

It’s a short list. There was the notoriously vulnerable Noah Syndergaard, who allowed a whopping 64 more bases than would have been estimated given the speed and lead distance of the runners. There was reliever Adam Ottavino, at 45 more than estimated, a long-known weakness of his that he’s been open about discussing, and then starters Yu Darvish and Tyler Glasnow, who each allowed 41 more total bases than would have been expected. Again, no surprises: here’s Glasnow way back in 2018 openly talking about his weakness.

And then, tied with Aníbal Sánchez and just ahead of Kenley Jansen and Jon Lester – each infamous for their inability to hold runners on – is Nola, at minus-32.

So what’s happening here? Let’s focus on the past three seasons, when Nola was fourth weakest.

All the bases taken against Nola in that time began to pile up – 33 of them in those runner-on-first, no-one-else-on situations, when only 16 would have been expected given the specifics of the runner and other Statcast inputs. That gap of 17 between actual and expected is sixth largest, which is a big problem, and they look like you’d expect, be they balks or stolen bases against.

But the bigger problem, really, are the outs, which we accept is a statement that seemingly makes no sense.

No pitcher over the last three years earned less credit on their outs – minus-6 in Nola’s case – and if you’re wondering how that’s possible, realize that catching a runner stealing is a team effort, and sometimes, even if the pitcher doesn’t do a good job holding the runner on, an elite effort from the catcher can still rescue the opportunity. Like, for example, if you’re pitching to J.T. Realmuto, who is merely the deadliest throwing catcher of his generation.

You can see that on this José Azocar attempt from last season, where Nola’s slow release and inability to hold Azocar close resulted in an opportunity that, from Realmuto’s point of view, was given just a 7% chance of being successful. That it was indeed an out is a tremendous credit to Realmuto (+0.93 run), while still counting as a negative on Nola’s ledger (-0.15 run.) There are others like it.

So: Why?

Look at all the pitchers who have thrown 1,000 pitches that were steal opportunities over the last three years, and you can see how far each pitcher lets the runner get off the base. That runners on steal attempts have taken a 9.4-foot initial lead on Nola on steal opportunities is not terribly noteworthy; it is essentially average. But by the time he releases the ball, the runners are an average of 14.6 feet off first base. That means that during the time it takes him to start moving and finish throwing, runners have gained more than five feet of distance toward second – and that is the largest number among any regular starting pitcher.

It’s similar when the runner is actually going. The extra 12.4 feet allowed during Nola’s slow delivery to the plate is the eighth most.

When we say “bases are stolen off the pitcher, not the catcher,” this is what we’re talking about. For example, when Aaron Judge stole second base off of Nola and Realmuto in the first week of the season, the broadcast video makes it look like a close play, one that might possibly have been an out if second baseman Bryson Stott could have cleanly received the throw.

But what you can’t see there is that Nola allowed Judge to get off to such a massive jump – Judge gained nearly 16 feet of distance between Nola’s first move and releasing the pitch – that barely any credit was subtracted against Realmuto for failing to convert an opportunity that was considered nearly impossible (11% est. caught stealing) given the particulars of the situation when he received the ball. Nola, meanwhile received -0.13 runs.

So that’s all bad, but: We said he had a problem, right?

Nola pitched four times in the 2023 playoffs. He allowed a single stolen base. That’s partially because in August, Nola re-introduced the slide step he hadn’t used in many years.

“I feel like once the pitch clock came,” he said before the NLCS, “it made things a little more challenging, because those [previous] years, I really did rely on holding the ball a lot and picking over a lot. So, obviously guys are stealing and bases are bigger. And I knew I could incorporate what I used to do back in 2017. I just had to do it and get comfortable with it again. The most important thing is being able to repeat the slide step while making a good pitch and focusing on making a good pitch.”

It was notable that in the Phillies' NLCS against the run-happy D-backs, Arizona was unable or unwilling to be aggressive. Corbin Carroll, who had been successful on 54 of 59 (92%) steal attempts this year, never ran.

Nola began using the slide step on Aug. 21, and while he didn't switch to it on every pitch with runners on after that, when he did, it was extremely noticeable. That's clear in this side-by-side view with runners on first from an April game (when Jazz Chisholm Jr. steals second) and NLCS Game 2 (when Carroll stays put).

The numbers back it up, too.

There was barely any difference between the leads taken against Nola in steal opportunities before and after he began using the slide step, as runners were 9.8 feet off first at his first move before the slide step, and 9.4 after. But look at how much less distance they gained while he was in motion:

Through Aug. 16 (leads from first)

• At first move, 9.8 ft
• At pitch release, 15.4 ft
• Lead distance gained, 5.6 ft

After Aug. 21 (leads from first)

• At first move, 9.4 ft
• At pitch release, 14.2 ft
• Lead distance gained, 4.8 ft

That’s a foot less in distance gained, which is a big deal. That's not the same thing as saying he's elite now, because he's hardly that, but it's better, and may still yet be improved because, as we said, he was sometimes still using the windup in steal situations:

Which, naturally, means runners are trying less, too.

“He was slide-stepping every pitch to keep him there,” said Arizona manager Torey Lovullo after the second game of the NLCS. “I didn’t want to run into an out. I want to be aggressive. It’s definitely in our DNA, and we know how to do it. But he was slide-stepping the majority of the time.”

Through his first 25 starts of the year, Nola faced 22 steal attempts, or 0.88 per game. In his final 11 starts of the year, including the postseason, after the slide step came back, he faced six steal attempts, or 0.55 per game.

“In my eyes and in the eyes of the coaches,” said Carroll, “it didn’t seem like there was much opportunity to run.”

That's one of the foremost basestealers in the Majors speaking, and it's not something you'd have heard about Nola for most of the last half-decade. As the rules changed, so did he. It's the kind of adjustment interested teams in free agency will be very excited to see.