Clutch gene? The hot hand? What the numbers say about old-school playoff beliefs

In one corner, fighting with the force of their convictions and the accumulated knowledge of decades of historical examples, are the eye test and the old school. In the other, armed with spreadsheets and cold, hard numbers, are the analytics and the new school. Let’s get ready to rumble.

That’s a bit overstated. Every team uses scouting and analytics in their quest to win. But the balance between the two is not the same for every team. The same could be said at this point for any pundit, analyst, or writer, really. Numbers are so ever-present in player development and coaching processes that there aren’t many players left who are strictly old-school, at least not in major-league clubhouses.

Advertisement

Still, there are battlegrounds where the old-schoolers and the numbers don’t see eye to eye. Those skirmishes are bubbling up in places like San Francisco, where an aggressively analytics-based coaching approach may have suffered a setback with manager Gabe Kapler’s firing, and Toronto, where the players expressed surprise about their dealing pitcher being taken out of a playoff game early. And then there are the countless broadcasts and social media spaces where deeply held beliefs about baseball are being put up against what the research finds about those ideas.

Let’s look at four of those long-held beliefs in baseball and what the numbers reveal about them.

A clutch gene exists

Philadelphia’s Bryce Harper is clutch. We can see the homers, the panache, and those gaudy postseason numbers that make him one of the three or 10 best of all time, depending on where you place the playing time cutoff.

“There’s a reason you get the nickname, ‘The Showman,'” Bryson Stott said about Harper’s growing October legacy. “He’s just one of those guys where it feels like any time you need a run, you need a hit, you need a home run, he’s always up. And then he always does what he’s supposed to do.”

According to the best research out there, though, “clutch” is not a repeatable thing. It’s clear on the team level that situational hitting comes and goes — you have several hitters, all going in different directions, dealing with various issues and facing teams of varying quality.

This is every team’s BA w RISP by month and if it looks like a total incomprehensible mess, yes, exactly pic.twitter.com/w8Vq3om7bK

— Mike Petriello (@mike_petriello) September 19, 2023

There’s no relationship between a team’s batting average with runs in scoring position over the last month and their batting average with runs in scoring position in the future. And it doesn’t help if you look at the last two or three months. It’s all just chaos.

Advertisement

More surprising to some might be that the numbers have yet to show any relationship between a specific hitter’s ability to be clutch and their future clutchiness. Yes, there is a “clutch” stat, which looks at how players perform in close and late (high-leverage) situations compared to how they perform in less critical situations, and yes, Harper was first in that stat this year. He was 52nd in 2022. He was 176th out of 188 in 2021.

These wild swings in year-to-year numbers were replicated in the research. Different analyses have looked at how “sticky” clutch is year to year and have found virtually no correlation between one year and the next. One found that it does not exist, but if it did, it would be on the order of single-digit points of batting average. Even the studies that found an effect didn’t find a large effect. Statistician Nate Silver found that clutch hitting was maybe two percent of a player’s overall quality (which would agree with the single-digit batting average hypothesis). Another study found that clutch exists, but you need 7,600 plate appearances in the clutch before you can believe it. Harper has had 651 high-leverage plate appearances in his career.

There’s a big difference between proving it doesn’t exist and not being able to show it exists, though, and that’s an important distinction. Our understanding of the physics of ball flight has changed with the improvement of technology; just look at the concepts of seam-shifted wake and how that led to the sweeper pitch boom as evidence. Sometimes, further study results in new findings.

What if we had a better way to study clutch?

“I firmly believe that if put heart monitors on us in the games, you’d find that some people are just more chill than others in big moments,” outfielder Brett Phillips told me this year.

Advertisement

Right now, wearable tech is not allowed in game situations, but theoretically, if it were, we might find that some hitters have different cardiac characteristics in high-leverage moments. Finding that biological processes were ultimately different between players in clutch situations would be a good way to prove it exists, but that is also fraught with implications when it comes to measuring those processes, who owns that data, and how front offices could use that information when it comes to paying players.

For now, clutch exists in our hearts but not really in our spreadsheets.

Pitchers get into grooves

This one is similar to clutch but on the pitching side. A pitcher is dealing, hitting his spots, cruising through the innings, and up comes a crucial decision for the manager. Blue Jays starter José Berríos struck out five while allowing just three hits to the first 11 batters he faced in Game 2 of the Wild Card Series against the Twins, and his manager took him out of the game. When asked later if he understood the reasoning behind the move — ostensibly, the thinking was that the righty was about to face a lefty for the second time in the game, and pitchers are progressively worse against hitters the more they see them, and it was a must-win game — the pitcher said:

“Honestly, I don’t know,” Berrios said after the game. “But other than that, I can’t control that. So, like I say, I did my best … So that’s what it is.”

Other managers have made different choices and been praised for it.

“If you’re making pitches, you’re going to be out there.” How Jordan Montgomery validated Bruce Bochy’s trust in Game 1: https://t.co/11fs44E9cF

— Ken Rosenthal (@Ken_Rosenthal) October 16, 2023

But history is littered with people yelling about taking a pitcher out too early or too late. What do the numbers say about the likelihood that a dealing pitcher will continue doing what he’s doing? Mostly, they say: “You’re only hot till you’re not.” The most robust study about this showed that there was virtually no difference in seventh-inning production between pitchers who dealt for six and pitchers who were not dealing and that managers haven’t, as a group, demonstrated the ability to make that decision successfully.

This flies in the face of many pitchers’ lived experiences.

Advertisement

“We can all see it. We know it. I know it when I’m dealing and when I’m not,” said one veteran starting pitcher when asked about the feeling.

Again, maybe more research will change the message. There’s already been a new and surprising study on the subject that suggests that this might be a legit thing. Writer and analyst Rob Arthur found that starting pitchers can have the “hot hand”  and that the best way to measure it is the radar gun; when pitchers have their best velocity, they perform much better than when they don’t, and they go through somewhat predictable streaks. A pitcher may be fundamentally different when they have a better fastball, and it’s not really a hot hand, but that’s semantics.

The study that’s yet to be done (publicly, at least) is to take Arthur’s work and see if, by using radar gun readings, managers could make better in-game decisions on when to take a pitcher out. Berríos? He’d thrown just seven fastballs and they were all a tick higher than his seasonal average, and he ended his short start with the second-best fastball velocity of the season.

It’s totally possible that the numbers say he should’ve been left in.

A hitter can “own” a pitcher

Rangers starter Nathan Eovaldi dispatched the first three Astros in Game 2 of this year’s ALCS in 11 pitches. Then Yordan Alvarez came up and demolished a high fastball. Eovaldi dispatched eight of the next nine and then … struck out Alvarez on five pitches, none of them fastballs. In their last matchup, Alvarez walked. In all three cases, the batter seemed to be stepping in confidently against the pitcher. With good reason:

Yordan Alvarez is 8-for-11 against Eovaldi.

— Brian McTaggart (@brianmctaggart) October 16, 2023

But we’ve also seen a good example of the opposite: a great hitter stepping in against a nasty pitcher, with a trail of defeat in their rearview mirror.

Another example of how batter vs pitcher numbers are devoid of meaning! https://t.co/a5iZaPVckU

— Eno Sarris (@enosarris) October 13, 2023

Saying that the numbers are entirely meaningless is not quite right. The numbers between a batter and pitcher tell the story of what happened in the past, so there’s meaning there. It’s part of telling the story of Turner’s night and how good he was. In terms of the narrative, batter versus pitcher numbers can help spin the yarn. But how impactful are those numbers, even in telling a story, if they have no predictive quality?

Advertisement

The research suggests that, almost always, past results between a specific hitter and a specific pitcher are much too small a sample size to have any statistical meaning. Tom Tango, MLBAM’s Senior Data Architect, once co-authored a book called “The Book: Playing the Percentages in Baseball” and after looking into the issue mathematically, the text had this to say about batter versus pitcher numbers:

… having twenty to thirty PA against an opponent is a drop in the bucket, and it tells you almost nothing about what to expect. The player has a long history, say 1,500 PA, against the rest of the league. Any way you slice it, you can’t equate, or even compare, twenty-five PA against one opponent to 1,500 PA against the rest of the league. … When a particular batter has faced a particular pitcher two hundred or three hundred times, come back and we’ll talk. Maybe. 

Put this way, it makes intuitive and statistical sense, but that isn’t to say that other research doesn’t provide us with some asterisks.

Back in 2013, Andrew Koo at Baseball Prospectus noticed that the Oakland Athletics were loading up with fly-ball hitters and linked that with a passage in “The Book” that pointed out that fly-ball hitters are especially good against ground-ball pitchers to suggest that the A’s were countering the league’s ground-ball pitchers with hitters who had swing paths that could lift those low pitches.

I talked to then-Mariners starter Chris Young ahead of a 2014 start against those Athletics and told him throwing his high-80s fastball up in the zone was daring, considering the risk of homers up there.

“That’s your opinion,” he told me. “I’ll show you a chart on every hitter that shows you that most hitters have a hole in the zone up.”

He then scrolled through heat maps on an iPad to show me he wasn’t worried about high fastballs against any of the A’s hitters other than Coco Crisp. Which was interesting, because despite being a high ball hitter, Crisp was only 3-for-15 with three singles against Young at the time. In any case, the current GM of the Rangers was an early adopter of how many analysts are now thinking about batter versus pitcher matchups.

Teams have gotten even more sophisticated with this type of analysis. They have linked swing paths to pitcher strengths and even started incorporating biomechanical markers in their batter versus pitcher models. Imagine something along the lines of: this hitter’s hips commit early, but his hands are adjustable so he’s theoretically better against this pitcher’s changeups than breaking balls with bigger movement profiles. Once you start getting into process versus process comparisons, you can potentially start to model that batter versus pitcher matchup and say something predictive about a future interaction.

Advertisement

But that’s a lot more complicated than saying this guy is 3-for-10 off that pitcher.

Hitters get cold after four days off

After a divisional round in which three of the four teams with byes lost — and did so while seeing their bats go ice cold — this belief is perhaps the most contentious. It’s definitely true that Atlanta, after being one of the best offenses of all time during the regular season, hit .186/.255/.264 in their short series against the Phillies, and that the Dodgers hit only .177/.248/.250 against the Diamondbacks. It’s also true that baseball is the most chaotic sport.

From April 18 to April 21, Matt Olson, Ronald Acuña Jr., Mookie Betts and Freddie Freeman combined for 58 plate appearances and a .143/.260/.262 line.

In the playoffs, that foursome combined for 56 plate appearances and a .137/.241/.157 line.

poop happens!

— Eno Sarris (@enosarris) October 13, 2023

Anything can happen in a short series, too. One study estimated that for MLB to match the NBA’s rate of the better-ranked team advancing in the playoffs, they’d have to play a best-of-75 series, so it’s possible even the American League East’s 11-game losing streak could all be due to chance.

Therefore, studying whether players get cold during those layoffs is fraught with issues. Ben Clemens at FanGraphs found that teams with four-plus days of rest are 24-11 against teams with two or fewer days of rest, and that number is 21-10 in the Wild Card era — and that is through this year, so it doesn’t suggest that teams do poorly overall with byes. That makes sense because it gives their pitchers time to rest and recover, but it also doesn’t look at hitters specifically.

“I’m a firm believer that more than three days off of facing a live pitcher can only hurt you,” said Giants outfielder Mitch Haniger this week. “We just never do it other than the All-Star break, and most guys talk about how they feel like crap in the box coming out of the break. Multiple days off can’t help your timing. It may help your body or mind.”

Again, the numbers don’t back the subjective experience. Thanks to STATS Perform, we have baseball’s OPS on the day after the break, compared to the season overall, in each of the last three years.

The August OPS across this sample was .731, so it doesn’t look like hitters suffer as a group coming out of the All-Star Game. Plus, teams with byes have ways to keep their hitters fresh.

Advertisement

“The teams that had the Wild Card off were working out and likely getting live at-bats during the downtime. And if they weren’t, that was a mistake,” said recently retired hitter Jed Lowrie earlier this week. Tommy Pham echoed the sentiment, saying it was on the coaches to help keep the players sharp during downtime. 

This one is obvious, though: there’s more research required, and the players talking about getting cold should be considered. Those who work on how the body adapts to its peak outcomes suggest that some adaptations — like peak sprint speed — can disappear within as soon as five days. But again, can training keep those adaptations without the benefit of the best live pitching? Do you have to face someone as good as Zack Wheeler live five days before you face him, or can a lesser pitcher throwing live BP help you keep your timing?

“No good answer,” said Diamondbacks third baseman Evan Longoria.

And that’s an important thing to remember: the numbers only give us their best possible answer right now. As data and tech improve, the answers might change. There’s an asterisk on each of these arguments that points to future research and more nuanced truths. The collective beliefs of the players engaged in the game should never be discounted fully. Instead, these disagreements should prove that there’s more work to be done here, more research to be explored and that these arguments should continue evolving.

(Top photo of Bryce Harper: Elsa / Getty Images)

ncG1vNJzZmismJqutbTLnquim16YvK57k3JtcHFoZnxzfJFsZmpoX2aGcLnLm2SppJGuvKey0maYp5mcrsGqr9JmmqWtpJi1br%2FTmqusZw%3D%3D