Hot Hand Fallacy: Is the Shooter Really on Fire?

The crowd is on its feet. A basketball player has just nailed three shots in a row. The commentator calls it: "He's on fire — keep feeding him the ball!" Coaches agree. Teammates agree. The player himself agrees. And for decades, cognitive science agreed with one verdict: you're all wrong. The "hot hand" is an illusion, a pattern-detecting brain tricking itself into seeing streaks in random coin flips. Then, in 2018, everything got more complicated.

The 1985 Study That Changed Sports Psychology

In their landmark 1985 paper, Thomas Gilovich, Robert Vallone, and Amos Tversky set out to rigorously test the hot hand belief in the Philadelphia 76ers. They analysed every shot attempt across a full season, then looked at whether a made shot was actually followed by made shots more often than chance would predict. Their finding was unambiguous: a player's shooting percentage after a hit was statistically indistinguishable from his shooting percentage after a miss. The sequences were consistent with independent random events — like coin flips — not streaky, dependent processes.

They also tested free throws (where defensive pressure is constant) and conducted a controlled shooting experiment with Cornell players. Same result every time. The hot hand, statistically, wasn't there.

The authors weren't just making a statistical point. They were demonstrating something profound about human cognition: we are constitutionally unable to perceive randomness as random. A sequence of coin flips like HHHTHHT looks suspicious to us; we expect alternation because that "feels" more random. Give us HHHHHH and we're certain the coin is rigged. This is the same machinery that produces the clustering illusion and fuels conspiracy theories — the mind compulsively finds signal in noise.

Why the Belief Persists

Gilovich and colleagues also surveyed basketball fans, coaches, and players, finding that nearly everyone believed firmly in the hot hand. Fans estimated that after a hit, a player is roughly 6–10 percentage points more likely to make the next shot. Players agreed. Coaches structured their offences around it.

Why is the belief so stubborn? Several mechanisms conspire:

• Selective memory: We remember the three-shot streak vividly; we forget the missed shot that followed it or the isolated makes sprinkled through a cold quarter. Confirmation bias filters the evidence.
• Narrative drive: Humans are story-making animals. A streak becomes a story — the player "found his rhythm," "got in the zone." Random alternation has no story.
• Gambler's fallacy reversal: Interestingly, the hot hand belief is the opposite of the gambler's fallacy. The gambler thinks after a streak of reds, black is "due." The hot hand believer thinks after a streak of makes, another make is more likely. Both are wrong in their way — but the hot hand version feels intuitively supported by real athletic phenomena like momentum, confidence, and physical rhythm.

The Miller & Sanjurjo Bombshell

For thirty years, the hot hand fallacy stood as one of behavioural economics' clearest examples of cognitive bias. Then in 2018, Joshua Miller and Adam Sanjurjo published a paper that sent shockwaves through the field: there was a subtle but consequential statistical bias in the original analysis.

The issue is counterintuitive. Imagine you flip a coin 100 times and look at every time heads comes up, then record what happened on the next flip. You'd expect the next flip to be heads 50% of the time — but it isn't. Due to a sampling bias in finite sequences, heads-after-heads occurs slightly less than 50% in a fair sequence. This means that the original Gilovich analysis — which compared post-hit shooting percentages to baseline — was using a biased baseline. The methodology slightly underestimated the hot hand effect.

When Miller and Sanjurjo corrected for this bias and reanalysed the original Cornell shooting experiment data, they found evidence that the hot hand is real — players who had made their last three shots were indeed more likely to make the next one. The effect size was modest but statistically significant. They also reanalysed professional basketball data and found similar results.

This is, by any measure, a stunning reversal. One of the most cited findings in cognitive psychology — taught in every decision-making course, used as the canonical example of pattern-seeking bias — turned out to contain a flaw.

Where the Debate Stands

The Miller-Sanjurjo critique has itself been debated. Other researchers have pointed out that:

• Even granting the statistical correction, the hot hand effect in actual game conditions is small and inconsistent across datasets.
• In-game shooting attempts are not random samples — players who are "hot" may take harder shots, face tighter defence, and shoot from less favourable positions. These confounders could suppress an underlying effect.
• Conversely, "hot" players who take harder shots and still outperform their baseline might actually provide stronger evidence for the effect than analyses of controlled shooting drills.

The current scientific consensus, if there is one, is: the hot hand belief is no longer simply a fallacy. There is plausibly a small real effect — driven by genuine within-game factors like muscle memory, defensive attention, and confidence — that the original analysis was not equipped to detect and that the human mind wildly exaggerates.

What the Debate Reveals

The hot hand saga is instructive beyond basketball. It illustrates several deep principles:

Science corrects itself. A canonical finding was scrutinised, a methodological flaw was identified, and the conclusion was revised. This is how it's supposed to work — even if it takes three decades.

Effect size matters as much as direction. Even if the hot hand is real, the human intuition wildly overshoots the magnitude. We treat a modest probabilistic elevation as a near-guarantee ("keep feeding him!"), restructuring entire game plans around a small statistical signal buried in noise.

Base rates matter. Whether you believe in the hot hand or not, the rational response is to know a player's baseline shooting percentage and update it cautiously in light of recent performance — not to throw strategy out the window because of three consecutive makes. This is precisely where base rate neglect enters.

Streaks happen by chance, frequently. In any long sequence of events, runs of consecutive successes are mathematically inevitable. The clustering illusion tells us that when we encounter these runs, our brain insists they must mean something. Sometimes they do. Usually the effect is smaller than our intuition demands.

Practical Implications

Sports fans and coaches aren't the only ones affected. Hot hand thinking appears in:

• Finance: Fund managers with recent outperformance are rewarded with inflows, even though short-term performance is a poor predictor of long-term skill. Regression to the mean is chronically underestimated.
• Hiring and promotion: Candidates on a "hot streak" — recent wins, high quarterly numbers — are over-weighted relative to longer-term performance records.
• Gambling: Players on winning streaks at roulette or poker feel "in the zone" and raise their bets, not recognising that random processes have no memory and past wins don't shift future odds.
• Creative fields: The belief that artists, writers, or directors are "on a roll" after recent successes can lead to rushed green-lights and inflated expectations that feed directly into disappointment when regression to the mean reasserts itself.

Sources & Further Reading

• Gilovich, T., Vallone, R., & Tversky, A. "The Hot Hand in Basketball: On the Misperception of Random Sequences." Cognitive Psychology 17, no. 3 (1985): 295–314.
• Miller, J. B., & Sanjurjo, A. "Surprised by the Hot Hand Fallacy? A Truth in the Law of Small Numbers." Econometrica 86, no. 6 (2018): 2019–2047.
• Kahneman, D. Thinking, Fast and Slow. Farrar, Straus and Giroux, 2011. Chapter 10.
• Thaler, R. H., & Sunstein, C. R. Nudge. Yale University Press, 2008.
• Wikipedia: Hot hand