PHOTOGRAPH BY MATT SLOCUM, AP
Published February 20, 2014
American ice hockey fans are thinking one thought: Go for the gold, USA men's Olympic hockey team. You've got the "Big Mo."
Only it turns out, according to a new study by Kevin Kniffin, a behavioral scientist at Cornell University, that there's no evidence of momentum—the idea that once a team starts winning, there's no stopping 'em—playing a role in the outcome of sports games. Kniffin and colleague Vince Mihalek examined men's college hockey games played over a six-year period in the Western Collegiate Hockey Association. The schedule was ideal for a study: Teams play each other in back-to-back games, one day after the next, in the same location. Conclusion: "We find that neither victory nor the margin of victory in Game 1 of a two-game series is predictive of the outcome of Game 2."
Why is it important to study momentum in sports?
[People] look to sports as a model for drawing lessons for work and life. Why do people watch bobsledding? It's not because we're going to go bobsledding. It's because partly I think they're distilling and drawing generalizable lessons.
Like: The key to success is loading up on carbs and doing lunges?
Exactly. Or hard work pays off. Or, at a team level, if a team gets along with each other, and they perform well, people will accurately or inaccurately draw the inference that that's probably true for my work team as well. People perceive momentum to be important in sports and draw the conclusion that momentum is important for nonsports activities, whether it's ... stocks or company performance or politicians.
And sports fans really do like to talk about momentum.
People perceive momentum as an observer or participant in activities because we have an evolved instinct to look for patterns.
So it's part of human evolution? What's the value of seeing patterns?
The idea is that we look for signals, for patterns, to help anticipate the future. If you can anticipate the future, you can see how that would have benefits.
What did your analysis of the scores show?
On the surface we found a correlation between the margin of victory in game one and game two. A reasonable observer would draw from that that there is momentum.
I sense a "but" coming.
What's really happening is just good teams winning more frequently. Winning Friday night or running up the score [in the first game] doesn't influence the probability of winning Saturday night when you control for the quality of the teams, using the end-of-season record.
So clearly the coach of the men's USA Olympic ice hockey team shouldn't talk to players about the Big Mo.
Coaches believe momentum might help propel them. But momentum can cause overconfidence. The conclusion of this study is that focusing on fundamental performance is what's most valuable.
Maybe the coach should talk about "FOF"—"focus on fundamentals."
Focus on fundamentals—that flows from the study. When you do that, streaks will happen.
And while running up the score isn't going to give you momentum to win the next day's game, some people think it's bad—you're using up your juju.
There's some belief among some sports fans that teams can use up their performance. We found no negative influence [from running up points]. The idea of fans being concerned about wasted effort, achievement, we find no evidence of that.
Any predictions about whether the men's USA Olympic team will get the gold?
I'm a fan like everybody else, but the sample size [of previous matches that can be used to make predictions] is too small.
How to Feed Our Growing Planet
National Geographic explores how we can feed the growing population without overwhelming the planet in our food series.
The Innovators Project
After achieving nuclear fusion at age 14, Taylor, now 19, is working with subatomic particles for solutions to nuclear terrorism and cancer.
Larvae attract more larvae, but not if they don’t have any bacteria. by Ed Yong
Latest News Video
The nation's most complete Tyrannosaurus rex specimen is taking a 2,000-mile road trip from Montana to its new home in Washington, D.C.