a photo of a basketball net

Why do NBA players who aren’t living up to their high salaries manage to get so much time on the court? Perhaps their coaches have fallen victim to unconscious bias known as the “sunk-cost fallacy,” according to a new University of Guelph study.

Published in the journal Empirical Economics, the study by two U of G economics researchers found that professional basketball players being paid high salaries tend to get slightly more playing time than their on-court performance suggests they deserve.

“Intuitively you would think the amount of time a player spends on court would solely depend on how well they perform,” said Prof. Yiguo Sun, who worked on the study with master’s student Alexander Hinton. “But it doesn’t work that way in practice. We found that salary does have a positive impact on how long they play.”

The study revealed that, on average, higher-paid players got 42 to 48 seconds more court time per game than players earning less than half their salary who were equally good. 

Someone earning $27 million playing alongside a comparable player earning $10 million got almost a minute more court time per game.

Prof. Yiguo Sun, Department of Economics and Finance

“The differences were not stark – just one or two minutes per game – but they were significant. A few extra minutes on the court per game is a lot,” said Sun.

In the NBA, salaries are a prime example of a “sunk cost,” because once a player has agreed to a contract, he gets paid no matter how much he plays. If a player is not living up to his salary, he should theoretically spend more time warming the bench.

But this study found that even players who were playing below a level reflecting their high salaries still got plenty of court time.

For the study, Sun and Hinton used spatial econometric modelling to analyze the effect of both performance and salary on the playing time of hundreds of NBA players over four years.

They analyzed 465 NBA athletes between the 2013/2014 and 2016/2017 seasons, looking at key markers such as free-throw percentages, points-per-minute and blocks-per-minute, as well as age and experience.

Their study also accounted for “unobservable player characteristics,” such as a player’s popularity or leadership, which can also affect playing time, as well as periods when players perform better alongside certain teammates.

And finally, they used spatial econometric methods to control for the performances of a player’s teammates. This unique approach allowed the researchers to capture the interdependencies in playing time between players, to account for when players perform better when alongside certain teammates.

Alexander Hinton

“To our best knowledge, our paper is the first to explore the within-team player time dependence in the sunk-cost fallacy literature,” said Sun.

They found that when two players performed equally well, the higher-paid player ended up with slightly more playing time.

Previous research in this area has focused on a player’s draft pick order, rather than salary. But the authors say players’ draft numbers can be forgotten over time; salary, on the other hand, is relevant each season.

Hinton said the research idea came to him while watching NBA playoffs and noticing that some teams used players who didn’t seem to deserve more playing time.

“I wondered if part of the reason was because the teams didn’t want to seem wasteful and to have their highest-paid players sitting on the bench.”

In these days of instant player statistics and timely performance feedback, it would stand to reason NBA managers would focus only on performance data, not sunk costs, particularly given the intense media and fan scrutiny.

But Hinton and Sun wonder whether it’s precisely that scrutiny that makes it hard for teams to admit to mistakes in salary decisions, particularly if a player is a new hire.

“We know managers, if they contract a player themselves, they hesitate to throw away the bad seeds. That’s the sunk-cost fallacy at work,” said Sun.

If the fallacy exists in a data-driven environment, such as the NBA, it likely affects decisions in other industries, where performance feedback is less accurate and timely, said Hinton.

“It can be hard to ignore the fact that you spent a lot of money and time on something,” he said. “You see it all the time, whether it’s in relationships or situations where people have invested a lot of time or effort. It’s just human nature.”


Prof. Yiguo Sun