It is one of the most talked about topics among Yankees fans and baseball fans alike. Which players seem to thrive when the game is on the line, and which ones wilt under pressure? I am talking, of course, about performance in the clutch.
Clutch, by one definition, is above-average performance in a high-pressure situation. We, as fans, know it as being consistently able to “deliver” or “get the job done” with the game hanging in the balance. Within these definitions, I believe there are two related, yet distinct, scenarios that attempt to capture the meaning of clutch.
The first scenario describes a player who performs better in a high-pressure situation than they would in any other. This player might experience a surge of adrenaline that sharpens their focus and quickens their reaction time. He would also be exerting maximum effort and concentration in this high-pressure situation in order to maximize chance of success.
The alternate definition of clutch posits that while performance may not increase, it also does not falter in high pressure situations. It is for this scenario that one might hear the cliches “cool hand,” “heartbeat slows,” “ice in his veins,” “everything slows down,” or “trusting one’s instincts.” Instead of necessarily trying harder, all of the preparation one puts in allows that player to treat the high-leverage situation as any other.
FanGraphs has attempted to quantify clutch, and its formula adheres much more to the first description of clutch. You can find an introduction into their calculation of clutch here. Basically they look at how much a player increases their team’s probability of winning in high-leverage situations versus in every other situation.
In my opinion, FanGraphs’ clutch calculation is flawed. It assumes that one has to perform better in high-leverage situations than their context-neutral average in order to be considered clutch. This is a very narrow and incomplete measure of a player’s ability to perform under pressure.
Additionally, FanGraphs’ formula only compares a player to himself and not to other players in the league. Therefore, even if said player produced far above league average in high-leverage situations, they could still be assigned a negative clutch value if they did as well or better in all situations. Would we say that such a player was not clutch? I think not.
Take, for example, the fact that Mariano Rivera has a negative clutch value for his career. He was so good in just about every situation that the variation took effect from the small sample size of his most high leverage situations to bring down his overall score. In other words, he is penalized by the formula for being so consistently excellent.
It is understandable that FanGraphs used this methodology when formulating a “Clutch” statistic, as it is easier to compare players using a more distilled analysis, and statistics cannot adequately capture a more big picture, zoomed-out approach that tells us that the Jeters and Pettittes and Riveras of the world may indeed have been clutch players.
What evidence is there to support the existence of clutch? In a 2017 study, Swann et. al. listed 12 defining characteristics of clutch states based on self-report surveys of athletes. The authors detailed how athletes who succeeded in clutch situations experienced factors such as deliberate focus and intense effort leading to their successes, which would support the first description of clutch above. They also found that just as important were the automaticity of skills and increased alertness, which would support the second description of clutch above.
Interestingly, the authors also noted the influence of confidence, perceived control, and enhanced motivation in clutch moments. This has an interesting correlation to several other theories in cognitive and behavioral psychology which analyze the ability to perform a task. Chief among these include self-efficacy theory, which explains how belief in one’s own abilities modulates performance; expectancy theory, which describes the effect of motivation on performance, and the Interpersonal Expectation Effect, or self-fulfilling prophecy, which states that the belief in an outcome can increase its probability.
I also noticed the intersection of clutch with two other psychological phenomena in sports: the hot-hand and choking, both of which dovetail nicely with both the findings of the Swann et. al. study and the effects described by the various psychological theories.
Former University of Chicago professor Sian Beilock detailed in her book Choke: What the Secrets of the Brain Reveal About Getting it Right When You Have To how overthinking in high-pressure situations crowds our working memory, interrupts our motor cortex, and inhibits the performance of intuitive, muscle-memory actions. This offers satisfying affirmation of the second clutch scenario, as choking is essentially the opposite of clutch.
With regard to the hot-hand phenomenon, recent studies have reignited the belief that the hot-hand is more than mere fallacy, and in fact is an internally-generated effect stemming in part from self-belief and repetition of motion. This provides an interesting connection to Swann et. al. study, as it appears confidence is operating in both situations.
On the flip side, there is ample evidence to suggest that clutch performance is not a concrete phenomenon, and instead is the outcome of random chance. These chance-induced fluctuations in performance are then modulated by the observer’s perception and skewed to support a belief that a player is or is not clutch.
Take the aforementioned Derek Jeter and compare him to Alex Rodriguez. Jeter is the paradigm of a clutch hitter among Yankees fans, while A-Rod has been construed as a choke artist. The binomial probability distribution provides convincing evidence that Jeter’s perceived successes and A-Rod’s perceived failures in the postseason are the result of chance. As fans, we remember Jeter’s successes more vividly because of his near-universal veneration among the fan base. As for A-Rod, the postseason slumps are more readily recalled as he did not enjoy nearly the amount of support as his teammate.
Ultimately, the most important thing to take away from any debate about clutch ability is that our perceptions trump all. Whether being clutch is something that can be measured or the consequence of chance, it is our previously held opinions and evaluations of a player that inform the assignment of a clutch or non-clutch label to that player. Like Tyler said a little over a year ago, when the game is on the line, you want your best player at the plate or on the mound.