The science of competitiveness
What can the labor market tell us about sports?
In March 2017, a Dutch behavioral economist named Thomas Buser began a two-year experiment, conducted in two phases. In the first phase, he sent out a survey to a representative sample of 5,255 working-age people in the Netherlands with a questionnaire consisting of a single question: “How competitive do you consider yourself to be?” Respondents were told to choose a value between 0 (”not competitive at all”) and 10 (“very competitive”)
In the second phase, Buser sent out another questionnaire to the same group a year later asking if they’d like to participate in a simple math tournament. They could choose to be paid for their participation by a basic hourly fee. Or, they could opt to collect a larger sum if and only their score was better than another randomly selected participant.
Only 27 percent of the participants chose the tournament-style compensation. But there were some noteworthy characteristics among those who did:
They were more than twice as likely to be in the highest quintile of income compared to the other participants, and much less likely to be in the lowest two.
They were more highly educated on average, and more than twice as likely to have graduated from one of the highest paying majors (medicine or STEM).
Their responses to the first questionnaire were also strongly correlated to their incentivized responses to the second. For both men and women, the individuals who considered themselves to be competitive were much more likely to have graduated from a university, chosen a lucrative major, and earned a high-paying job. Competitive people were more likely to have chosen to work in medicine, business, or law.
On their surface, these results probably aren’t exactly shocking or illuminating. And, on the surface, the design of the study seems deserving of some scrutiny. What characterizes somebody as competitive vs. simply somebody who likes taking risks? Surely we all know people who might be competitive in some ways but risk-averse in others (like gambling).
But subsequent research by Ernesto Reuben and Lina Lozano at NYU Abu Dhabi has found that, among the subjects they tested, individuals could exhibit a strong tendency toward competitiveness irrespective of their attitudes toward risk. Work done by Linda Kamas at Santa Clara University and Anne Preston at Haverford College has explored gender differences in competitiveness and found that working-age women “who exhibit a taste for competition” earn, on average, 7 percent more than other women.
And so over the past decade it seems that the field of behavioral economics is at least beginning to chip away at the question of “intangible” traits and whether they can be predictive of success in the labor market, an important question as we grapple with the question of whether college is useful and how to close the racial and gender gaps in earnings.
And it was only a matter of time before somebody adopted this line of inquiry and applied it to sports.
Measuring competitiveness in athletes
A paper presented at the MIT Sloan Analytics Convention earlier this month by Spanish researchers Ander and Julene Palacios-Saracho was called “Measuring Individual Competitiveness and Its Impact on Sporting Success.” It’s not a peer-reviewed paper, let’s get that out of the way first. I’m not suggesting this work is gospel.
But I was struck by a line in the piece that I couldn’t believe to be true:
“We are not aware of any empirical study in the scientific sports literature that measures and studies competitiveness.”
That’s always a good place to start.
I was also struck by Ander and Julene’s approach. Absent any comparable work in sports science, they went to where the existing research could guide them. They took their cue from the work that have economists have done on competitiveness in the labor market.
Elite soccer stars
Here’s the other thing about this paper. It was a longitudinal study (meaning it was conducted over a long period) that involved repeatedly observing soccer players ages 10-23 from an elite soccer academy in the one of the world’s best leagues (Spain’s La Liga) over the course of a decade from 2011 to 2023.
That’s another good place to start.
Here’s how their “competitiveness” was measured.
First, they were told to throw a tennis ball into a bucket that is placed 3 meters away. They were each given 10 chances to score as many buckets as possible.
Each player is told he is randomly matched with another player who is performing exactly the same task at the same time in another room.
Before starting, the player can decide to receive 1 euro per successful shot regardless of the performance of the other player. Or, he can decide to receive 3 euros per successful shot if he outperforms the other player (a tie gets 1 euro per shot).
In addition to the players, the researchers also quizzed the coaches on their perceived competitiveness and their willingness to engage in a competitive situation — exactly how Buser gauged the competitiveness of working-age adults in the Netherlands. I’ll get to why this was informative in a moment.
RESULTS
First off, the players got progressively more competitive as they aged. At age 10, around 55 percent of the players chose their compensation based on competition. By age 18, its 92 percent. The steepest jump in percentages came at ages 14 to 15.
To begin with, the elite athletes even at a young age were much more likely to choose the competitive compensation than non-players of the same age from Spain and Germany, suggesting there is some selection effect that differentiates elite athletes starting as early as age 10.
There were some positional differences. Goalkeepers and defenders appeared to be more risk averse than midfielders, and midfielders were more risk averse than forwards.
And there were differences in how some players responded in the presence of a competitive coach. But these results weren’t uniform across all ages. When 13 to 15-year-old players were paired with a highly competitive coach, the player’s competitiveness would in turn also tend to spike, while the opposite was true for a low-competitive coach, suggesting a malleability in competitiveness as a direct correlate to coaching at a certain age. This didn’t happen at all with the youngest kids, and the types of coaches also appeared to be meaningless for players after they turned 16.
Lastly, as with the working-age Dutch, the same was true of Spanish soccer players: The researchers were fairly convinced that those who chose to be competitive in the bucket-toss game were more likely to reach a higher level of ability on the soccer field.
They wrote:
We have found that differences in competitiveness do help account for differences in performance and successful outcomes, that players’ preferences for competition are malleable, but only up to a certain age (more precisely, within a certain age), and that coaches may have a causal impact on this intangible (typically positive, but not always).
That’s a good place to start.
Competitiveness and the 10,000 Hours Rule
One of the many misconceptions about the famous 10,000 hours rule is that it is a rule. This would suggest that, upon practicing something deliberately for 10,000 hours, at the moment you reach that threshold you will be bestowed as an “expert,” as if earning a degree. This is ridiculous, and it also misses the point.
The point is that 10,000 hours is a really, really long time. If you practiced for five hours a day, all 365 days of the year, it would still take you 5.5 years (sportswriter math here) to accrue 10,000 hours of practice time. This might be a bigger factor in who becomes elite than just the practice: the willingness to devote this much time and deliberate focus to one pursuit at the sacrifice of almost literally everything else. Not everyone is willing to do this.
What fuels that drive to succeed is, quite often, competitiveness. I’m not saying anything earth-shifting here! But here again is an example where skill alone isn’t the end-all determinant of excellence. Expertise, in the clinical sense of the word, requires unusual drive and competitiveness, and identifying that trait early might be a better way to identify the next Kobe Bryant or Simone Biles.
We’ve also seen how malleable that trait can be, and how receptive it is to nurturing at a young age. Therein lies another lesson for coaching. Perhaps the youth coaches could dial it down just a smidge on the “just win, baby” philosophy for the 10-year-olds. There’s scant evidence that it’s making any difference. In fact, it may be detrimental.
Wait until middle school to stoke the competitive juices. By age 16, they’ll be tuning you out again. Competitiveness, at that point, is self-induced.
Measuring intangibles
My other big takeaway from this work is that sports too frequently gives up on measuring things that are considered out of reach. Analytics has come a long way — plenty of people would say it has gone too far. But that’s because it has become too granular in many cases, while leaving wide swathes of the science of skill (TM) unexplored.
We can pinpoint the spin attributes of a baseball traveling 100 mph, but we have little way of measuring the impact that motivation, pressure, stress, fatigue, or competitiveness had in also determining the ball’s eventual location. Some of this can be blamed on technological shortcomings, but I think some of it also has to do with a lack of imagination in the investigatory approach.
So kudos to Ander and Julene Palacios-Saracho for looking outside the box for a way to measure the “intangible.”
Links
Heading a soccer ball does cause brain damage … Sports psychology has an evidence problem … Ingesting sodium citrate improved sprint performance of soccer players, but also caused gastrointestinal problems … Five experts (3 data scientists, 1 video analyst, and 1 coach) preferred TacticAI’s suggested positional improvements at corner kicks … AI-powered tennis ball machine can learn your skill level and simulate live play … Cranberries can help improve performance of distance runners … The physics behind Caitlin Clark’s pure shot … Athletes likely to have higher levels of PFAS after playing on artificial turf
References
Buser, Thomas, Muriel Niederle, and Hessel Oosterbeek. Can competitiveness predict education and labor market outcomes? Evidence from incentivized choice and survey measures. No. w28916. National Bureau of Economic Research, 2021.
Kamas, Linda, and Anne Preston. "Competing with confidence: The ticket to labor market success for college-educated women." Journal of Economic Behavior & Organization 155 (2018): 231-252.
Lozano, Lina, and Ernesto Reuben. Measuring preferences for competition. No. 20220078. 2022.
You have found a niche area with these pieces, sort of a sports expert on stuff that most people don't pay any attention to. Really interesting.