Thursday, April 16, 2015

Having a Hot Hand Increases Confidence, But Not Success


One of the great things about doing research is that you can actually test the beliefs that people take for granted.  And sometimes, those beliefs are shown to be false.  A classic example of this approach comes in the belief in a hot hand in basketball.  When you watch a basketball game, a player will make a couple of shots, and the announcers will decide that player is “on fire” and that he ought to take the team’s next shot.

Back in 1985, though, Tom Gilovich, Robert Vallone, and Amos Tversky actually analyzed data from the Philadelphia 76ers.  They found no evidence for a hot hand.  The hot hand would say that if a player makes one shot, then they should be more likely to make a second.  Gilovich, Vallone, and Tversky found that the probability that a player would make a second shot was independent of whether they made the first one, suggesting that there is no hot hand.

An interesting question, though, is whether the belief in the hot hand influences the behavior of the players themselves. That question was explored in analyses by Yigal Attali reported in the July, 2013 issue of Psychological Science.  He analyzed all of the data from every game in the 2010-2011 National Basketball Association season.  Modern transcripts for games include lots of information including who took each shot, whether it was made, and the distance of the shot. 

Attali found evidence that the belief in a hot hand did affect the behavior of players.  When a player made one shot, it affected whether they would take the team’s next shot.  When the shot was from a short distance (a dunk or layup), then players took about 20% of their team’s next shots regardless of whether they made or missed the shot.  However, when they made a shot that was longer than 4 feet, they were much more likely to take the team’s next shot than if they missed that shot.

That’s not all.  When players made a shot, the next shot they took was generally further from the basket than when players missed their last shot.  Because longer shots probably reflect that a player has more confidence in his ability, this suggests that making a shot increases a player’s confidence. 

Paradoxically, though, this confidence has a cost.  Longer shots are more likely to be missed than shorter shots, so when a player takes two shots in a row, he is much more likely to miss the second shot than to make it, because the second shot is probably taken from further away following a hit than following a miss.  (Indeed, Attali re-analyzed the data from the Philadelphia 76ers that Gilovich, Vallone, and Tversky used, and found a similar effect that when a player makes one shot, they are actually less likely to make the second shot than when they missed the previous shot.)

Finally, Attali explored the effect of making a shot on the behavior of coaches.  He found that players were much less likely to be taken out of a game following a made shot than following a missed shot.  So, coaches are also acting as though they believe in a hot hand.

What does all of this mean?

In lots of domains (including basketball), we have theories about the way the world works.  Those theories influence our actions.  However, it is important to know whether our theories about the way the world works are actually true.  Sometimes, as in the case of the hot hand in basketball, not only is the theory false, but acting based on the theory also makes people’s performance worse than it would be if they did not believe in the theory.