HERON LAUNCH APP
SCOREBOARD

Was the screen right?

Meteora publishes no history, so none of this is backfilled. Every call the screen makes — the entries and the declines — is written down when it is made and settled twelve hours later against what the pool actually paid. Until a call settles there is no number here, and a blank is the honest reading.

— ENTRY WIN RATE

reading…

— DECLINE ACCURACY

reading…

— EDGE, POINTS

reading…

— MEDIAN CAPTURE

reading…

The ledger, filling up

settled — re-read and scored due — settling on the next scan open — still inside its 12 hours
WHAT COUNTS AS A WIN

The prediction

At call time the screen sees a fee rate of predicted %/hour — the 1h window, divided by its one hour.

The settlement

Twelve hours later the pool's 12h window covers exactly the period the call was open. Divided by twelve, that is realized %/hour — what an LP would actually have collected.

The score

capture = realized ÷ predicted. Held at 0.50 or above: you kept at least half the rate you were shown. Below, the pool faded.

Why declines are scored too

Scoring only entries measures nothing — a screen that admitted everything would look identical. Edge is the entry win rate minus the rate the declines would have won at. If it is not positive, the screen is not sorting anything.

Every call, in order

PAIR VERDICT CALLED PREDICTED %/H REALIZED %/H CAPTURE RESULT
reading the ledger…

Your positions

Nothing is signed and nothing is ever moved — connecting only reads your open Meteora DLMM positions from the public portfolio endpoint.

Today's screen

  1. reading…