Model vs market: biggest stored disagreements
Where upcoming stored model probabilities differ most from the raw implied bookmaker price—and why disagreement alone is not a bet.
Engine evidence checked .
This comparison uses the latest stored rows available at 2026-08-03T05:35:21.741Z. Raw implied probabilities are calculated as one divided by decimal odds and do not remove bookmaker margin.
Marta Kostyuk vs Katherine Sebov — Marta Kostyuk: model 57%, raw market 95%, gap -38%. Scheduled 2026-08-05T19:00:00.000Z. R. Colby vs S. Johnson — S. Johnson: model 54%, raw market 88%, gap -34%. Scheduled 2026-08-03T16:00:00.000Z.
D. Sakellaridis vs A. Dougaz — A. Dougaz: model 51%, raw market 85%, gap -34%. Scheduled 2026-08-03T12:30:00.000Z. D. Recek vs M. Topo — M. Topo: model 52%, raw market 83%, gap -31%. Scheduled 2026-08-03T11:00:00.000Z. D. De Jonge vs V. Gonzalez-Galino — D. De Jonge: model 68%, raw market 97%, gap -29%. Scheduled 2026-08-04T10:00:00.000Z. Aryna Sabalenka vs Moyuka Uchijima — Aryna Sabalenka: model 65%, raw market 93%, gap -28%. Scheduled 2026-08-05T19:00:00.000Z.
These are upcoming stored fixtures, not necessarily today's matches. A large disagreement can reflect missing data or model uncertainty, so the normal value and confidence gates still apply.