The Too-Much-Talent Effect

A few years ago, the psychologists Adam Galinsky and Roderick Swaab began working on a study that looked at the relationship between national levels of egalitarianism – the belief that everyone deserves equal rights and opportunities – and the performance of national soccer teams in international competitions like the World Cup. It was an admittedly speculative hypothesis, an attempt to find a link between a vague cultural ethos and success on the field. But their logic went something like this: because talented athletes often come from impoverished communities, the most successful countries in the highly competitive World Cup would find a way to draw from the biggest pools of human talent. Think here of the great Pele, who was too poor to afford a soccer ball so he practiced his kicks with a grapefruit instead. Or the famous Diego Maradona, born in a shantytown on the outskirts of Buenos Aires. These men had talent but little else. It is a testament to egalitarianism that they were still able to get the opportunities to succeed.

It’s a nice theory, but is it true? After controlling for a number of variables, including GDP, population size, length of national soccer history and climate, Galinsky and Swaab found that egalitarianism was, indeed, “strongly linked” to better performance in international competition. It also predicted the quantity of talent on each team, with more egalitarian countries producing more players under contract with elite European clubs. In short, the most successful soccer countries don’t necessarily have the most innately talented populations. Instead, they do a better job of not squandering the talent they already have. 

It’s a fascinating study with broad implications. It suggests, for one thing, that much of the national variation in performance – and it doesn’t matter if we’re talking about the soccer pitch or 8th grade math scores – has to do with how well countries utilize their available human capital. What T.S. Eliot said about the excess of literary geniuses during the Elizabethan age (Shakespeare, Marlowe, Spenser, Donne, etc.) turns out to be a far more general truth. “The great ages did not perhaps produce much more talent than ours,” Eliot wrote, “but less talent was wasted.”

So far, so interesting. But as often happens in science, answers have a slippery way of inspiring new questions; the scientific process is a perpetual mystery generating machine. And it’s this next mystery – one utterly unrelated to egalitarianism – that most interests me.

While analyzing the soccer data, Galinsky and Swaab noticed something very peculiar – at a certain point, having more highly talented players on a national team led to worse performance. It was an unsettling finding, since people generally assume that talent exists in a linear relationship with success. (More talent is always better.) Such logic underpins the frenzy of NBA free-agency – every team is begging for superstars – and the predictions of bookies and commentators, who believe that the most gifted teams are the most likely to win. It’s why an already loaded Barcelona team just spent more than $100 million to acquire Luis Suarez, a player who has become as famous for biting as he has for striking.

And so, armed with this anomaly, Galinsky, Swaab and colleagues at INSEAD, Columbia University and VU University Amsterdam, decided to continue the investigation. After confirming the result among soccer teams competing at the 2010 and 2014 World Cup – too much talent appeared to be a burden, making national teams less likely to win – the scientists decided to see if their findings could be extended to other sports.

They turned first to basketball, looking at the impact of top talent on NBA team performance between 2002 and 2012. They coded talent by looking at the Estimated Wins Added (EWA) statistic, a measure that reflects the approximate number of wins a given player adds to a team’s season total. (In the 2013-2014 season, Kevin Durant led the league with an EWA of 30.1. LeBron was second with 27.3.) Once again, talent exhibited a tipping point: NBA teams benefited from having the best players unless they had too many of them. While most general managers assume the link between talent and performance is linear – a straight line with an upward slope – the scientists found that it was actually curved, and teams with more than 60 percent top talent did worse than their less skilled competition. Swaab and Galinsky call this the “too-much-talent” effect.

 The relationship between team talent levels and team performance in the NBA

The relationship between team talent levels and team performance in the NBA

What accounts for the negative returns of excessive talent? The problem isn’t talent itself; there’s nothing inherently wrong with gifted players. Rather, Galinsky and Swaab argue that too much talent can disrupt the dynamics required for effective teamwork. “Too much talent is really a metaphor for having ineffective coordination among players,” Galinsky says. “Sometimes, you need a hierarchy on a team. You need to have different roles. But if everyone thinks they should be the one with the ball, then you’re going to run into problems.” Galinsky, et al. documented this drop-off in coordination by tracking various measures of “intra-team coordination,” such as the number of assists and defensive rebounds per game. (Both stats require teammates to work together.) Sure enough, the-too-much-talent effect was mediated by a drop-off in effective coordination, as teams with too many top-flight athletes also struggled with their chemistry. The egos didn’t gel; the players competed for the spotlight; all the talent became a curse.

When I asked Galinsky for an example of a team undone by their surfeit of talent, he cites a 2013 quote from Mike D’Antoni, the head coach of a gifted Lakers team that woefully underperformed. (The starting five featured four probable Hall of Famers: Kobe Bryant, Steve Nash, Dwight Howard and Pau Gasol.) “Have you ever watched an All-Star game? It's god-awful,” D’Antoni said to reporters. “Everybody gets the ball and goes one on one and then they play no defense. That’s our team. That’s us. We’re an All-Star team.” The 2012-13 Lakers were swept by the Spurs in the first round of the playoffs. 

Likewise, the LeBron era Miami Heat only succeeded once their talented stars learned how to work together. “When Dwayne Wade got hurt [in 2012), the Heat became a less talented team,” Galinsky says. “But I think his injury also made it clear that he was subordinate to James, and that James was the true leader of the team. That helped them play together. Having less pure talent actually increased their performance.” This suggests that the too-much-talent effect might explain a bit of the The Ewing Theory, which occurs when a team performs better after the loss of one of its stars.

Of course, if athletic talent exists in a tensioned relationship with teamwork, then the effect should not exist in sports, such as baseball, that require less coordination.  “If you have five starting pitchers, those pitchers don’t need to like each other, because they all start on different days,” Galinsky says. “Too much talent shouldn’t be a big problem.” (The scientists quote Bill Simmons in their paper, noting that baseball is “an individual sport masquerading as a team sport.”) To test this hypothesis, Galinsky, et. al. used the Wins Above Replacement stat, or WAR, to assess the talent level of every MLB player. Then, they looked to see how different levels of team talent were related to team performance. As predicted, the relationship never turned negative: for baseball clubs, having more highly skilled players was always better. “These results suggest that people’s lay beliefs about the relationship between talent and performance are accurate, but only for tasks low in interdependence,” write the scientists.

 The relationship between team talent levels and team performance in MLB

The relationship between team talent levels and team performance in MLB

These findings aren’t just relevant for sports teams. Rather, the scientists insist that the too-much-talent effect should apply to many different kinds of collective activity. While organizations place a big emphasis on acquiring top talent – it’s often their top HR priority – the importance of talent depends on the nature of the task. If success depends on the accumulation of individual performances – think of a sales team, or hedge fund traders – then more talent will lead to better outcomes. However, if success requires a high level of coordination among colleagues, then more talent can backfire, especially if the group lacks a clear hierarchy or well-defined roles.  And that’s why the best basketball teams, Galinsky argues, feature talented athletes who focus on different aspects of the game. “No one would argue that the Jordan era Bulls teams weren’t incredibly gifted,” he says. “But Jordan, Pippen and Rodman all understood their roles.  They knew what they needed to do.”

There is, I think, one final implication of this paper. In a world of moneyball GMs and SportVU tracking, it’s easy to dismiss the importance of team chemistry as yet another myth of the small data age, an intangible factor in a time of measurable facts. But this paper provides fans and coaches with a useful way of thinking about the importance of player chemistry, even if we still can’t reliably quantify it.* We’ve always known that team coordination matters, that a group of talented athletes can become more (or less) than the sum of their parts. But now we have empirical proof – a lack of chemistry is the one problem that more talent cannot solve.

*We might not be able to quantify player chemistry, but there does seem to be some consensus among players as to who has it. Talented athletes take big pay cuts to play with LeBron – he makes his teammates better - but Houston couldn't convince any superstars to play with Dwight Howard and James Harden. 

Swaab, Roderick I., and Adam D. Galinsky. "Egalitarianism Makes Organizations Stronger: Cross-National Variation in Institutional and Psychological Equality Predicts Talent Levels and the Performance of National Teams." Organizational Behavior & Human Decision Processes (forthcoming)

Swaab, Roderick I., et al. "The Too-Much-Talent Effect Team Interdependence Determines When More Talent Is Too Much or Not Enough." Psychological Science (2014)