A player has scored three in a row. Is it really worth passing to him again?

Evidence of streaks after all. Picking only the shots that followed a run of successes is not a random sample, and with a perfectly fair coin the measure will show a rate below half. So the tool was biased against the claim it was testing. Corrected, across 29 NBA three-point contests, it found streaks in professional shooters.

For years it was taught that the hot hand is an illusion, and it became the classic case of intuition going wrong: a previous score does not raise the chance of the next.

Except that the measure itself hides a bias in selecting the streaks. Taking only the shots that came after a run of successes makes the sample non-random, and with a perfectly fair coin the measure will show a rate below half. The tool was biased against the claim it tested.

Researchers corrected the bias and reanalyzed twenty-nine years of three-point contests, an environment where the distance and the conditions are fixed and there is no defender and no decision about who to pass to.

There they found evidence of streaks in professional shooters. It is a rare case in which a statistical correction overturned a famous psychological conclusion, by reanalyzing existing data with no new data collected. A result from a controlled contest does not prove a hot hand in every game, for every player, or from every spot on the floor.