Where the model disagrees with the market
9 fixtures — every row opens the full prediction.
| Match | League | Pick | Edge | Ours | Market | Conf. | Best odds | Bet |
|---|---|---|---|---|---|---|---|---|
| Bournemouth vs LeedsPremier LeagueA | Premier League | BournemouthOur prob 53%Best odds 2.18 | +8% | 53% | 45% | A | 2.18 | Bet Now |
| Union Berlin vs FC CologneBundesligaA | Bundesliga | FC CologneOur prob 43%Best odds 2.72 | +7% | 43% | 36% | A | 2.72 | Bet Now |
| Villarreal vs EspanyolLa LigaA | La Liga | VillarrealOur prob 70%Best odds 1.50 | +4% | 70% | 66% | A | 1.50 | Bet Now |
| Hull City vs IpswichPremier LeagueA | Premier League | Hull CityOur prob 40%Best odds 2.73 | +3% | 40% | 36% | A | 2.73 | Bet Now |
| Udinese vs RomaSerie AA | Serie A | RomaOur prob 64%Best odds 1.63 | +3% | 64% | 61% | A | 1.63 | Bet Now |
| Deportivo La Coruna vs OsasunaLa LigaA | La Liga | Deportivo La CorunaOur prob 47%Best odds 2.28 | +3% | 47% | 44% | A | 2.28 | Bet Now |
| Deportivo La Coruna vs OsasunaLa LigaA | La Liga | Deportivo La CorunaOur prob 47%Best odds 2.03 | +3% | 47% | 44% | A | 2.03 | Bet Now |
| Elche vs ValenciaLa LigaA | La Liga | ElcheOur prob 42%Best odds 2.55 | +3% | 42% | 39% | A | 2.55 | Bet Now |
| Rayo Vallecano vs Celta VigoLa LigaA | La Liga | Rayo VallecanoOur prob 43%Best odds 2.44 | +3% | 43% | 40% | A | 2.44 | Bet Now |
Our prob = the probability we publish for that pick (the Ours column) — our model blended with the de-vigged market price where one is recorded; Market = that de-vigged market probability. Edge = Ours minus Market; grade A/B = the pick is backed by both/one team's xG record. Best odds = the top bookmaker price on our pick, taking each book at its own last update.
We only record bookmaker prices for the 1X2 market, so every row here is a 1X2 selection — an edge on any other market would be invented rather than measured. xG-backed picks (grade A or B) rank first, then by edge — so a grade C row can sit below smaller edges, because a number the model produced without an xG record behind either club is a weaker claim than the same number with one. Prices older than 24 hours are dropped rather than shown stale; the newest snapshot behind this page was taken at 08:14 UTC on 28 October 2026.
What an edge is, and what it is not
Edge is the gap between two numbers we both publish: our model’s probability for a selection, and the same probability implied by the best recorded price once the bookmaker’s margin is stripped out. A 2-point gap is the smallest we treat as signal rather than rounding.
An edge is a statement about price, not about the result. A pick with a large edge still loses most of the time when its probability is low — the claim is that the price is wrong, not that the outcome is certain. The grade next to each match says how much of that claim rests on expected goals: A means both clubs have an xG-backed record behind their rating, B one of them, C neither.