Hot-Hand Fallacy
Understanding the Hot-Hand Fallacy
The hot-hand fallacy is when people expect a random series to continue. For instance, a basketball player who has made her last three shots is expected to make her next shot because she has a “hot hand.” Here, it appears that the streak of success prompts individuals to overestimate the degree of skill, and hence overestimate probability of ongoing success.
The hot-hand fallacy refers to the intuition of basketball fans and coaches that players tend to get “hot” after having scored a few times in a series. The intuition is that after a series of n hits, the player’s probability of another hit will increase and that of a miss will decrease.
This fallacy was attributed to the representativeness heuristic Opens in new window. The reason given was that “even short random sequences are thought to be highly representative of their generating processes” (Gilovich, Vallone, and Tversky, 1985, 295). People predict another hit because a hit is representative of the previous series of n hits.
Gilovich, Vallone, and Tversky (1985) found that both basketball fans and professional players suffered from an erroneous belief concerning the chances of scoring. They tended to believe that a player’s chances of hitting a shot are greater following a hit than following a miss. They termed this a belief in the “hot hand” or “streak shooting.”
However, a detailed analysis of the shooting records of two professional teams and a controlled experiment with a university team provided no evidence of a correlation between outcomes of successive scoring attempts.
Classic Explanation of the Hot-Hand Fallacy
The hot-hand fallacy is a version of the general notion of ‘form’ in sports performers: that during a period of good form, a player is more likely to hit following a hit compared to following a miss than would be expected by chance.
People strongly believe in form, and will swear that this elusive property is causing a higher than normal probability of hits. Having conducted several studies of players’ actual performance when they looked at the conditional probabilities of hits following hits and misses, Gilovich and colleagues found no evidence for this effect.
What is happening, they say, is that people are oblivious of the statistical effect of regression to the mean Opens in new window, which posits that each player will, over the long term, produce a personal average hit rate.
In the short term, there will be deviations from this mean that are essentially random, and these deviations will sometimes occur in lumps. Thus, people with huge experience of playing the game, or who at least attend enthusiastically to the causality of the game, were liable to produce a causal hypothesis, “form,” to ‘explain’ the effect for what were, in fact, chance outcomes.
The hot-hand fallacy is almost always studied along with its close relative, the gambler’s fallacy Opens in new window.
