As the first half of the 1998 World Cup final drew to a close, France’s star midfielder Zinedine Zidane headed the ball past Brazil goalkeeper Claudio Taffarel for his second goal of the game. “Lightning has struck twice!” one of the television commentators declared. “What a time to score a second.”
The moment echoes an old sentiment in the soccer world: that a goal scored shortly before halftime is a particularly valuable contribution toward winning (and indeed, France went on to triumph 3-0, with Zidane named the game’s biggest star). The belief that goals scored just before halftime make a win more likely has persisted for decades, but it has rarely been put to the test, and attempts to do so have done little to reveal the truth of the adage. In previous studies, some researchers found evidence contrary to the myth, others found it might be true, but it remained unclear whether the timing of a goal could affect the outcome of a game.
In a study published in the journal PLOS ONE, Yale SOM professor of operations management Nils Rudi and crime novelist Jo Nesbø set out to test this conventional wisdom. They analyzed data across several hundred thousand games and found that there was some truth to the myth: scoring just before halftime proffered an advantage, in particular, to the home team.
The findings could have implications beyond sports, including the management of teams in the corporate world. “Understanding how the timing of success affects a team’s response and future performance can have implications for other areas,” Rudi says.
The study emerged out of a long personal history. Rudi and Nesbø are friends who grew up in the same small town in Norway. Rudi’s brother Petter and Nesbø both played soccer professionally, but a knee injury ended Nesbø’s soccer career at the age of 18.
Nils Rudi became a scholar of operations research, whose research has sometimes focused on sports analytics and soccer contracts. He says that the business world can learn from sports. When studying a firm’s performance, he points out, the data at hand are often coarse—sometimes just in the form of end-of-year summaries that don’t capture mid-year changes in strategy, leadership or other aspects that could affect performance. But in sports, the rules remain the same across all games, and second-by-second data about play is often available. “Sports analysis is great if you want to study the psychological effects of events during the game on the final outcome,” he says.
Rudi and his brother introduced Nesbø to their interest in analysis. “At some point, Jo, Petter, and I started discussing soccer analytics,” Rudi remembers. “And as a closet quant geek, Jo got hooked.”
When the conversation came around to the effect of late first-half goals, Rudi and Nesbø disagreed. “Jo believed it was likely to be a significant effect, while I thought that any effect would be so small that it would not show up as significant,” Rudi says.
To investigate, Rudi and Nesbø, together with researchers Henrich Greve of INSEAD in Singapore and Yale SOM postdoctoral researcher Marat Salikhov, began by gathering information from the World Football data website, which has the timings of goals and game outcomes for more than 300,000 games. For parts of the study, they combined this data with games for which betting data were available, to create a log of 72,426 games played by 27 top national leagues between the years 1998 to 2016.
A difference of five percentage points in the odds of a win is “a significant percentage,” Nils Rudi says. “To increase your odds by that much, you need to practice a lot—or get some better players.”
They homed in on games where the score was tied at one goal for each team at halftime, because it meant the teams’ performance differed only in the order and timing of goals. They found that the home team won 41% of games when the team had scored a goal after the 44-minute mark of the first half. But when the away team scored in that same time frame, they only won 36% of games. (Under both circumstances, there was a tie about 36% of the time.) Overall, goals right before halftime were correlated with a greater difference in the scores of the two teams, and with greater odds of a home team win. The researchers found a similar, albeit weaker, correlation in games where only one team had scored before halftime.
“Admittedly I was very skeptical that we would find anything,” says Rudi. A difference of five percentage points in the odds of a win is “a significant percentage,” he adds. “To increase your odds by that much, you need to practice a lot—or get some better players.”
What explains this effect? In interviews with managers of professional teams, Rudi and Nesbø would often hear that scoring late in the first half boosted team morale. “When we asked their thoughts, they’d say yes, this timing matters, and give reasons like, ‘you have a good feeling and carry that back into the game,‘” Nesbø says. “But when we asked if they thought these goals were objectively more valuable, they’d say no.”
The study doesn’t offer an explanation as to why goals just before halftime seem to be more important, but Rudi suggests a few possible mechanisms for the effect.
One possibility is that tactical discussion in the locker room at halftime may focus more on offense for the team that just scored, and more on defense for the team that gave up a goal; evidence suggests that focusing on scoring is more effective.
For the player who scored, Rudi suggests, “a break may give time for more reflection, which may make the effect stronger than if the player needs to keep chasing the next ball.”
Or the managers could be right and the difference is about morale. The team that has been scored upon may be focused on blame, with negative effects on its play in the second half.
“The halftime break is when dialogue happens as a team,” he says, “The recency effect of a goal just before it may turn the focus of the scoring team towards more positive elements, and for the conceding team it may turn it more negative.”