The Gap #10: the book and the crowd disagree on Miller, Ty Cole vs Goff, Billy

Polymarket's UFC crowd and crypto bookmaker lines are not always telling the same story this week, and in three fights the gap is wide enough to matter.

The fights

The clearest separation sits in the Miller–Goff prelim. Ty Cole Miller opened at 1.30 on BC.Game, implying a 72.9% win probability, while the Polymarket crowd prices him at 75.5% — a 2.6-percentage-point gap. In money terms, BC.Game is laying you 1.30 on a fighter the crowd thinks should be 1.32. That is a small discrepancy in absolute decimal terms, but at short prices a two-and-a-half-point gap in implied probability is meaningful; the book is effectively underpricing Miller relative to where sharp crowd money has settled. The fight goes off on 9 August, so there is still time for the line to move toward consensus.

Billy Quarantillo against Diego Ferreira is the second market worth watching. Stake's vig-stripped number on Quarantillo sits at 37.2%, whereas Polymarket has him at 39.6%, producing a 2.4pp gap at a book price of 2.55. The crowd fair price comes out at 2.53, so you are being offered 2.55 on a selection the market thinks is worth 2.53 — marginal in decimal terms but the direction is right. The gap clears the 2pp threshold, which means it qualifies rather than gets skipped, though only just.

Delvalle over Elkins rounds out the live edges. Stake's implied number is 83.1% for Yadier Delvalle, and Polymarket sits 2.0pp higher at 85.1%. The book price is 1.14 against a crowd fair price of 1.17. At these short odds even a small probability gap translates into a meaningful pricing inefficiency relative to expected value, but position sizing at 1.14 demands discipline — you are staking a lot to win a little, and any late-breaking injury news will collapse the market instantly.

The skips

Several fights look tempting on the surface but do not clear the bar honestly. Alexia Thainara against Amanda Lemos shows a 1.8pp gap, just below threshold — no edge. Mackenzie Dern at 1.43 is similarly only 1.4pp wide versus Robertson — no edge. The main event angle, Ian Machado Garry to beat Islam Makhachev at 3.65 on Cloudbet, carries less than a 1pp gap against the Polymarket crowd — no edge, regardless of how attractive the upset price looks in isolation. Gamrot and Asplund are both cases where the book's implied probability is actually higher than the crowd's, meaning the book is the sharper number in those spots; fading the crowd there carries no structural argument.

As always, odds move between the moment we capture data and the moment you place a bet, and the price printed at the book when you click confirm is the only price that counts.