Hot Hand Fallacy (Streak Shooting Bias)
The hot hand fallacy is inferring that success itself raises the chance of further success when the outcomes do not support that inference. The historical basketball analysis here tested whether made and missed shots clustered more than expected under independence.
THE IDEA
Analysis of Runs
The table discussed below is not included in this excerpt. The reported results concern this sample and this test; they do not establish that every player’s shot outcomes are always independent.
Table 33.2 displays the results of the Wald–Wolfowitz run test for each player (Siegel, 1956). For this test, each sequence of consecutive hits or misses is counted as a “run.” Thus, a series of hits and misses such as XOOOXXO contains four runs. The more a player’s hits (and misses) cluster together, the fewer runs there are in his record. Column 4 presents the observed number of runs in each player’s record (across all 48 games), and column 5 presents the expected number of runs if the outcomes of all shots were independent of one another. A comparison of columns 4 and 5 indicates that for five of the nine players the observed number of runs is actually greater than the expected number of runs, contrary to the streak-shooting hypothesis. The Z statistic reported in column 6 tests the significance of the difference between the observed and the expected number of runs. A significant difference between these values exists for only one player (Dawkins), whose record includes significantly more runs than expected under independence, again, contrary to streak shooting.
When to use it
Evaluating player performance in sports analytics to avoid overvaluing streaks.
Investment decisions: avoiding the belief that a fund manager on a winning streak will continue to outperform.
In hiring, distinguishing several strong answers from evidence about performance on other skills that the interview has not yet tested.
In everyday life, recognizing when a series of successes is likely just random variation.
FROM THEORY TO EVERYDAY LIFE
In practice
Basketball Shooting Streak Analysis
In a study of nine professional basketball players over 48 games, the observed number of runs (consecutive hits or misses) was compared to expected runs under independence. Only one player (Dawkins) showed a statistically significant deviation from independence, and that deviation was more runs than expected, meaning his hits and misses actually alternated more than random, not less. This test did not find the clustering predicted by the streak-shooting hypothesis in this sample. It does not rule out every possible change in shooting probability.
A MOMENT TO REFLECT
Put it to work.
Can you recall a situation where you or others believed in a 'hot hand'? How would you design a simple test to check if the streak was real?
Why might our brains be prone to seeing patterns even in random sequences? How does this bias affect decision-making in areas beyond sports?
Consider the concept of 'regression to the mean.' How does it relate to the hot hand fallacy?
References
- Nonparametric Statistics for the Behavioral Sciences — Sidney Siegel (1956)
CONTINUE THE CONVERSATION
Source texts
- Heuristics and Biases The Psychology of Intuitive Judgment (Thomas Gilovich, Dale Griffin, Daniel Kahneman)