plan Trader

Ranking: which patterns survive the control, with their sample

First the patterns that survive the multiple-comparisons correction; then, by how far they sit from what happens on any day. Every row with its n.

No card required. Analysis tool, not investment advice.

A ranking, not a flat list

Instead of opening report after report, you first see the patterns that survive the false-discovery control, ordered by how far they sit from the any-day rate in that market.

A small sample does not pass on its own

A high percentage on few cases rarely survives the correction. By default only the patterns that pass it are listed, each with its n.

Full history or recent window

You can rank using the full history or a recent window on instruments with in-session data.

What data it uses and what its limits are

  • !A high place in the ranking is context for prioritizing, not a buy or sell signal.
  • !The recent window works on ES, NQ, YM, NKD and RTY; the other markets use the full history.

How we calculate every figure: methodology.

Frequently asked questions

How is the ranking ordered?
First come the patterns that survive the multiple-comparisons correction. Within each group, by how far the percentage sits from its own base: what happens on any day in that market. There is no single score.

Other features

Perfiltrade is an analysis and decision-support tool. It does not issue signals or investment recommendations. Past results do not guarantee future results; trading futures involves risk of loss. Risk disclosure.