Econ Sense: Streakiness and Hot Hand in Sports

One of the most thrilling elements with regards to professional sports is their unpredictability. One team can lose as many times in a row as one can count only to bounce back and win just as many. The same thing applies to professional players in all sports; a player can thrill the audience by collecting many hits or making many shots in one glorious stretch, only later becoming a pitcher’s best friend or being booed off the court after missing several easy lay-ups. Indeed, phenomena such as streaks and hot/cold hand have mystified fans of all ages. But are such phenomena only phenomena? Is streakiness part of the standard norm or is it only unique to a certain player?

There have been many attempts to uncover the mysteries behind such concepts. Albright (1993a) attempted to detect streakiness in the hitting data of professional baseball players. He found no evidence that streakiness, for the general player, is anything more than chance. For some specific players, however, there was evidence that they were what we would call ‘streaky’ players. Albert and Bennett (2003) performed statistical tests on winning and losing streaks of Major League Baseball teams. They concluded that streaks like Oakland’s 20-game winning streak in 2002 are rare and only happen once in every 25 years.

These are just a few of numerous studies on the streakiness of professional players covering a variety of sports and games including bowling, tennis, and even blackjack. Wong (1993) conducted an exhaustive simulation of hands to determine whether or not the result of the previous hand has an effect on the current hand. Interestingly, he found the effect to be statistically insignificant

All the aforementioned studies echo a similar conclusion. For most players, streakiness is no more than chance. If you toss a coin ten times, it is probable (though rare) that all 10 attempts will show head. If we attempt a lot more tosses, we’ll likely see many streaks—big and small. But no valid inferences can be drawn from these random streaks.

I recently conducted my own empirical research on the topic of streakiness in sports, professional golf in particular. Golf has always been considered to be a “feast or famine” game, where a player can play incredibly well for one round or for a couple of months only to cool off considerably in the subsequent rounds or months. The research was an attempt to examine whether or not streakiness is a general trait of professional golfers by observing any effect a player’s performance on the previous hole has on his performance on the current hole.

The results reconfirmed the conclusions reached by previous studies. I found no positive link between a player’s performance on the previous hole and that of the current hole. Digging deep into the empirical results, however, led to a surprise discovery. If you performed well on the previous hole, the probability of a bad result on the current hole is higher than the probability of a positive result. The converse is also true. A bad result appears to motivate people to increase their effort level; a good result induces lower effort level.

I’ll add my research paper on this topic to the author page in a few days. I want your opinion. What defines a great professional athlete? The number of major titles or championships won? Stats? Clutch play? Sustained excellence? Explain your choice.

[Simon N.]
~~~