Verdict grade
GOOD
GOODWhat it means
A GOOD verdict means the gap between the model's probability and the best price on the board is large enough to survive its own uncertainty. The model prices the bet, the vig is stripped from cross-book consensus to get the market's fair number, and the difference is the edge. GOOD is reserved for the top band of that gap at reasonable confidence — it is not a prediction that the bet wins, only that the price is better than the outcome deserves. Over a large enough sample of GOOD bets you should expect to win more often than the price implies. Any individual one is still a coin weighted slightly in your favour, not a lock.
What to do with it
Bet it at the listed price or better. Below that price, re-check the edge.
How it is produced
Every grade comes out of the same four-step sequence: sample the results behind the market, fit a model probability, strip the vig from cross-book consensus to find the fair price, and publish the gap. The full walkthrough, with numbers →
The size of that gap and the confidence behind it together decide which grade a bet receives.
The rest of the scale
Positive edge, but modest — directional, not a hammer.
Market and model agree. No bet.
Slightly overpriced — the other side is closer to fair.
Clearly overpriced against the model's number.
Not enough sample to trust the number. Informational only.
Displayed but not graded — no model covers this market yet.