Research: We Take More Risks When We Compete Against Rivals



What Not to Do When You’re Trying to Motivate Your Team

Rivalry is everywhere. We see rivalries in sports, business, school, and basically any arena where there is competition. Whether it is rivalry between people (Bill Gates versus Steve Jobs; Roger Federer versus Rafael Nadal), between organizations (e.g., Ford versus GM), or between nations (e.g., should U.S. soccer fans have rooted for rival Mexico in the World Cup?), there is something uniquely powerful about rivalry that differentiates it from others forms of competition and relationships.

There are many stories highlighting the benefits of rivalry, from how it makes competitors more motivated to how it helps them perform better. For example, the rivalry between the Beatles and the Beach Boys is thought to have pushed both groups to greater levels of creativity and performance in their song-writing. Similarly, the rivalry between Intel and AMD is thought to have helped advance computer chip technology.

Yet despite the wealth of anecdotal evidence, scientific research into such rivalry relationships is scarce. We’ve spent the last few years studying what makes rivalry unique, how it affects our behavior, and how it impacts organizations.

Our past findings support the idea that rivalry can motivate competitors to perform at higher levels, and they reveal that rivalry is sparked by experiences of similarity, repeated competition, and closely-decided contests. In our most recent paper, we ran two studies to examine the effects of rivalry on decision-making and risk-taking. Our goal was to determine whether, how, and why feelings of rivalry between competitors influenced their propensity to take risk.

Our first study looked at a field with fierce and well-known rivalries: the National Football League (NFL). With annual budgets and revenues in the hundreds of millions of dollars, hundreds of employees structured into teams and hierarchies, and high stakes for performance, NFL teams have a lot riding on coaches’ and players’ decision making. We analyzed thousands of in-game decisions to see whether teams take more risks on the field when they’re competing against rivals.

We measured rivalry between teams by averaging a few different ratings — those made by NFL experts (e.g., the NFL’s yearly list of “top 10 NFL rivalries”), those made by 56 NFL fans we surveyed who demonstrated expertise, and the count of Google search results for terms such as “New England Patriots versus New York Jets rivalry.” We then collected nine years of play-by-play data (2002 – 2010), totaling almost half a million unique plays.

We examined teams’ decisions in two critical in-game scenarios — conversions after touchdowns, and whether to “go for it” on fourth down. Both decisions involve a choice between a low risk (reliable but low-reward) option and a high-risk (less reliable but high-reward) option. Take conversions after touchdowns. After a team scores a touchdown (and receives six points), they can choose to either 1) attempt to kick the ball between the goal posts for one additional point, or 2) attempt to move the ball into the end zone from two yards out for two additional points. This is a classic choice between a low-risk, low-reward option and a high-risk, high-reward option. In our sample, teams successfully converted one point attempts 98.56% of the time and successfully converted two point attempts 45.68% of the time.

Fourth down decisions are slightly more complex, but they similarly involve choosing between a safer option (punting or attempting a field goal to three points) and a riskier one (“going for it,” or attempting to advance the ball in order to maintain possession and preserve the possibility of scoring a touchdown), where the potential rewards are higher, but so is the possibility of loss.

After controlling for a wide range of factors that might also influence teams’ decisions — including teams’ general propensities for risk-taking, time left in the game, score, week of the season, yardline and yards to go, the relative ability levels of the teams (determined from ESPN rankings), and whether the teams were in the same conference or division — we found that rivalry between teams predicted greater risk-taking. Specifically, in one analysis we found that teams were 37% more likely to attempt two point conversions, and 7% more likely to attempt fourth down conversions, in games against their rivals as opposed to games against non-rivals. We didn’t find that this increased risk-taking led to better or worse overall performance or outcomes.

We next ran an experiment to corroborate our results in a controlled environment and examine the reasons for why rivalry increases risk-taking. We invited 149 students at the University of Arizona to take part in a lab study of decision-making. Participants were told that they would be competing against another participant on a decision-making task that mimicked Blackjack. Unbeknownst to them, their competitor was actually a member of our research team, who posed as either a recent University of Colorado graduate or a recent Arizona State University graduate by wearing a school shirt. (We also told participants to wear clothing featuring their home university, as a way to activate their in-group identity.) As is true of many pairs of large U.S. universities located in the same state, Arizona and Arizona State are fierce rivals, with a history of competition that dates back to 1899 and a trophy that goes to the victor of their annual “Territorial Cup” football game. So participants were randomly assigned to compete against a member of either a rival or non-rival institution.

We found that participants who were matched with a rival Arizona State graduate took substantially higher levels of risk, on average, than participants matched with a Colorado grad. Specifically, they turned over more cards in the task, which was a high-risk, high-reward move, when competing against rivals. They were more likely to risk losing everything to accumulate more points.

In this experiment, we also investigated why rivalry increased risk-taking. We did so in two ways: First, we measured participants’ physiological arousal, via their heart rate, before they encountered their