Multiple Comparisons Problem

mechanism

An outcome can be rare per trial yet unsurprising across many trials. Judge its rarity only after counting all the opportunities for it to occur.

Rohit Sharma losing 15 tosses looks astonishing: treated as one isolated streak, the odds appear to be roughly 1 in 30,000. But international cricket has supplied India and other countries with match after match in which some remarkable streak could emerge.

E1

Rare per attempt, ordinary across attempts

The apparent paradox comes from changing the denominator. The eye notices one dramatic result and asks, “What were the chances of this exact thing happening here?” The relevant question is broader: “Across all comparable players, teams, matches, and possible starting points, how many chances were there for a streak like this to appear?” A small probability multiplied across a large field of opportunities can produce an event that remains striking but no longer demands an extraordinary explanation.

E1

Where it shows up

The toss streak

Sharma’s run feels almost impossible when he is treated as the only candidate and his streak as the only trial. Once the history of matches played by India and other countries enters the frame, the number of opportunities expands—and so does the chance that someone, somewhere, eventually records an extreme run.

E1

More chances do not explain everything

Counting opportunities can show that a streak is less surprising than its headline odds suggest; it does not prove that every unusual pattern is meaningless or identify the true cause. The conclusion depends on choosing an honest comparison set rather than enlarging it merely to dismiss the anomaly.

E1

Write down the hidden denominator

Before repeating odds attached to a remarkable streak, list who else could have produced it, how many trials each had, and how many different outcomes would have seemed equally remarkable. Reassess the claim using that total opportunity count.

E1

Episodes that teach this