Basketball’s hot hand and Wall Street’s momentum effect get filed together constantly, in the same breath, as twin examples of humanity’s stubborn urge to see meaningful streaks in what might just be noise. Both involve the same surface question — does recent success predict near-term future success — and both have decades of academic argument behind them. Treat them as the same dispute, though, and you’ll miss something more interesting: they’re actually different kinds of disputes, sitting at different, and quite unequal, stages of resolution, with one genuine technical thread quietly connecting them underneath.
Scientific Foundation
The hot hand’s modern history begins with a 1985 paper by Thomas Gilovich, Robert Vallone, and Amos Tversky, which examined shooting data from the Philadelphia 76ers, free-throw data from the Boston Celtics, and a controlled shooting experiment with Cornell basketball players, and concluded that a player’s chance of making a shot showed no meaningful dependence on whether their previous shots had gone in. The finding became conventional wisdom fast; Daniel Kahneman, Tversky’s longtime collaborator, later called the hot hand “a massive and widespread cognitive illusion.” For roughly three decades, that verdict held. Then, in the 2010s, researchers Joshua Miller and Adam Sanjurjo identified a genuine, subtle flaw baked into the original analysis’s method: a systematic statistical bias, sometimes called the “law of small numbers” effect, that causes the naive way of estimating “probability of a hit given a recent streak of hits” from a finite sequence to be biased downward — the correction is mathematically real, independent of basketball or psychology, and applies to any finite binary sequence analyzed this way. When Miller and Sanjurjo reanalyzed the original data correcting for this bias, they found genuine evidence of a hot hand effect the original methodology had been structurally incapable of detecting. Subsequent peer-reviewed work, including a study finding a strong hot-hand effect in baseball hitting, added further support. It’s worth being honest that this reversal isn’t unanimous even now — Gilovich himself has remained skeptical that the correction meaningfully changes his original conclusions, and researchers continue debating how large the real effect is and how much of it might be masked in raw data by opposing defenses adjusting their coverage against a player perceived to be hot.
Cross-Domain Connection
Financial momentum has an entirely different evidentiary history. Narasimhan Jegadeesh and Sheridan Titman’s 1993 paper found that stocks with the strongest returns over the previous three to twelve months continued to outperform stocks with the weakest returns over the following several months — and unlike the hot hand, this finding was not overturned or seriously disputed afterward. It has instead been described as perhaps the most pervasive contradiction of the efficient market hypothesis, replicated across international markets, multiple asset classes including bonds, currencies, and commodities, and more than thirty years of subsequent data. The existence of momentum, in other words, was never really the open question in finance the way the hot hand’s existence was in basketball.
What Remains Undemonstrated
The genuinely open dispute in finance is a different one entirely: not whether momentum is real, but why it exists, and whether it’s fading. Two broad camps remain unresolved after decades of research — behavioral explanations, holding that investors underreact to new information and prices adjust only gradually as that underreaction corrects itself, versus risk-based explanations, holding that momentum profits compensate investors for bearing some real, systematic risk that standard asset-pricing models fail to capture. Layered on top of that is a documented complication with no clean basketball equivalent: momentum’s measured profitability appears to have shrunk in some markets since the effect became widely known and traded on, with recent research finding U.S. momentum profits averaging a modest 0.31 percent monthly over 2000-2020 compared to 2.21 percent monthly in the same period for New Zealand’s smaller, less arbitraged equity market — a pattern consistent with a real anomaly getting partly competed away once professional investors start exploiting it, a dynamic basketball defenses don’t quite have an equivalent of, though the defensive-adjustment confound in basketball gestures at something structurally similar.
So the two debates aren’t the same shape at all. Basketball’s was, for three decades, a does-this-exist dispute, substantially though not universally resolved by uncovering a specific technical error in the original null finding. Finance’s was essentially never a does-this-exist dispute after 1993 — it’s a why-does-this-exist-and-is-it-shrinking dispute, a more mature and differently structured kind of open question. There is, however, a genuine technical bridge worth naming precisely, beyond the thematic resemblance: the exact statistical bias Miller and Sanjurjo identified in the hot-hand literature — the systematic underestimation that comes from naively computing “probability of continuation given a recent streak” from a finite sequence — is a general mathematical fact about analyzing runs in any finite binary data. It isn’t specific to basketball. A financial researcher naively computing conditional streak probabilities in stock-return data using the same flawed method could, in principle, run into the identical bias, understating real momentum or overstating its absence in exactly the way the original 76ers analysis did. That’s a real, transferable methodological lesson connecting the two fields, distinct from and more precise than the loose observation that both involve “streaks.”
Why It Matters
Keeping these two debates properly separated matters for how much confidence either comparison should lend the other. It would be a mistake to treat financial momentum’s rock-solid empirical status as somehow validating the hot hand’s more contested one, or to treat the hot hand’s messy thirty-year reversal as evidence that momentum’s well-established existence might also be shakier than it looks. They’re different claims, backed by different quantities and qualities of evidence, and the specific statistical trap that tripped up basketball researchers for three decades is the one genuinely portable lesson worth carrying from one field into the other — a reminder to check your method for exactly that bias before trusting any finite-sequence streak analysis, in a box score or a stock chart alike.
Human Dimension
There’s something worth sitting with in how differently these two “streak” questions actually resolved, despite getting lumped together in casual conversation for decades. One field spent thirty years confidently telling people their eyes were lying to them about basketball, based on an analysis that turned out to have a real, fixable mathematical flaw. The other field never seriously doubted its own streak effect existed at all, and instead spent thirty years arguing about what, exactly, was causing it. Both are legitimate scientific stories. They just aren’t the same story, and treating them as interchangeable examples of “people see patterns that aren’t there” flattens a distinction that, on closer inspection, is really the whole point.
Sources:
1. Wikipedia — “Hot hand” — https://en.wikipedia.org/wiki/Hot_hand
2. Nautilus — “The ‘Hot Hand’ Is Not a Myth” — https://nautil.us/the-hot-hand-is-not-a-myth-643539
3. Skeptical Inquirer — “A Closer Look at the Gambler’s Fallacy and the Hot Hand” — https://skepticalinquirer.org/exclusive/a-closer-look-at-the-gamblers-fallacy-and-the-hot-hand/
4. Statistical Modeling, Causal Inference, and Social Science (Andrew Gelman) — “Explaining to Gilovich about the hot hand” — https://statmodeling.stat.columbia.edu/2015/10/18/explaining-to-gilovich-about-the-hot-hand/
5. Stanford Graduate School of Business — “Why the ‘Hot Hand’ May Be Real After All” — https://www.gsb.stanford.edu/insights/jeffrey-zwiebel-why-hot-hand-may-be-real-after-all
6. Scientific American — “Momentum Isn’t Magic—Vindicating the Hot Hand with the Mathematics of Streaks” — https://www.scientificamerican.com/article/momentum-isnt-magic-vindicating-the-hot-hand-with-the-mathematics-of-streaks/
7. arXiv — “Do Steph Curry and Klay Thompson Have Hot Hands?” — https://arxiv.org/pdf/1706.03442
8. Springer Nature Link, Financial Markets and Portfolio Management — Wiest, T., “Momentum: what do we know 30 years after Jegadeesh and Titman’s seminal paper?” — https://link.springer.com/article/10.1007/s11408-022-00417-8
9. SSRN — Jegadeesh, N. & Titman, S., “Momentum: Evidence and Insights 30 Years Late” — https://papers.ssrn.com/sol3/Delivery.cfm/SSRN_ID4602426_code16600.pdf?abstractid=4602426&mirid=1
10. ScienceDirect — “Momentum: Evidence and insights 30 years later” — https://www.sciencedirect.com/science/article/abs/pii/S0927538X23002731
Idea originated at artificialideas.org. Article researched and written by Claude Sonnet 5. Published at artificialideas.org.
Artificial Ideas 