Skip to content

Studio · Lab

Which of your ideas actually beat the book?

Every idea is placed by its win rate against the line you choose. Below the line, the idea beats the line you chose on the sample shown, and deeper is better; above the line it does not. Bigger fish have more graded games. Ideas that pick the same sides most of the time are grouped into one school, recomputed each night.

Reading the research ledger…

Reading the research ledger
Loading

Floor and ceiling, not a p-value

A 61–49 record is 55.5%, but ten more games could move it four points either way. The card gives the floor and ceiling behind that record. An idea found by testing many at once has to clear a higher bar, and losing is called sooner than winning, on purpose.

Why a school matters

Two ideas that take the same side most of the time are one idea with two names. Schools are cut each night from pick agreement on this season's graded games. One idea per school is a portfolio; three from one school is one bet, three times.

Breeding

In the Hatchery two or three parents make a child by one of three rules: Agree, Vote or Stack. A child is graded only on games no parent was trained on; nothing borrows a parent's record.

Unlock more Studio tools

Keep exploring for free, or unlock the complete Studio toolkit with Pro.

$1 first week, then $2.99/week.

See Pro

Research verdicts and angles keep their research record and a checksum of the results they were graded on, so each number can be audited; model ideas are graded on their own recorded picks.