Sample Size

Glossary · Quant & Stats

Sample size is the number of independent observations behind a claim — trades behind a win rate, launches behind an odds ratio. Small samples produce large, confident-looking numbers by chance: five wins in six trades is an 83% win rate and also a coin flip's normal streak. Below a few dozen, a percentage is an anecdote.

"Six trades is a story. Sixty is a hint. Six hundred is a record. Anyone selling you a win rate without the count is selling the story."

A rough rule: the standard error of a proportion is about half over the square root of n, so a 60% win rate on 25 trades carries a margin of roughly ±10 points — anywhere from 50% to 70% is consistent with the data. On 100 trades that halves to ±5; on 400, ±2.5. This is why our own record reports every slice with its n, and calls anything under thirty directional evidence only.

The trap is not just small samples but selected ones: a strategy tuned on the trades it was tested on will look good on exactly those trades. Hold out data the rule never saw, count the trades honestly, and treat any edge that disappears when n grows as what it was — noise that fit a story.

Pattern backtests, with their sample sizes

Related jargon

Questions

What is Sample Size in crypto trading?

Sample size is the number of independent observations behind a claim — trades behind a win rate, launches behind an odds ratio. Small samples produce large, confident-looking numbers by chance: five wins in six trades is an 83% win rate and also a coin flip's normal streak. Below a few dozen, a percentage is an anecdote.

How would a trader say it?

Six trades is a story. Sixty is a hint. Six hundred is a record. Anyone selling you a win rate without the count is selling the story.

Part of the Crypto Jargon glossary — every term, decoded the way a trader would say it.

Crypto Jargon

Live numbers load with JavaScript and refresh every few minutes; the text above is the page at build time.