Track record
Every bet the model rated 4 or 5, settled and unedited. Nothing is removed after the fact.
Bets rated 4 or 5, one per side, at the Consensus price — the average across the books we cover. Sun 30 Aug – Mon 12 Oct. Pools every rating rule run so far; "Current rule" shows the one in use today.
How these are priced and staked. Each bet is graded at the last Consensus price before the game — close to the closing number, not the price on the board when the bet appeared — and staked at the size the board recommended: full Kelly on a 10-unit bankroll. At the first price a reader could have taken, the same stakes returned +1.42%. At one unit a bet instead, they returned −4.60%.
The settled bets
The 25 most recently graded, at the Consensus price. Every graded bet in this scope is in the download. Download this scope as CSV
| Rating | Day | Match | Bet | Price | Result |
|---|---|---|---|---|---|
| Mon 12 Oct | Levante v SevillaSoccer | U 2.5Total Goals (Over/Under) | −133 | lost | |
| Mon 12 Oct | Levante v SevillaSoccer | Sevilla PKPoint Spread | −135 | push | |
| Mon 12 Oct | Coventry v NewcastleSoccer | Harvey Barnes Harvey BarnesPlayer Shots | +100 | lost | |
| Mon 12 Oct | Coventry v NewcastleSoccer | Yoane Wissa Yoane WissaPlayer Shots | +100 | won | |
| Mon 12 Oct | Coventry v NewcastleSoccer | Lukas Hornicek Lukas HornicekGoalkeeper Saves | +100 | won | |
| Mon 12 Oct | Coventry v NewcastleSoccer | Carl Rushworth Carl RushworthGoalkeeper Saves | +100 | lost | |
| Mon 12 Oct | Florida Panthers at Buffalo SabresNHL | Olen Zellweger U 0.5Player Total Points | +100 | lost | |
| Mon 12 Oct | Florida Panthers at Buffalo SabresNHL | Peyton Krebs U 1.5Player Shots on Goal | −110 | won | |
| Mon 12 Oct | Florida Panthers at Buffalo SabresNHL | Sandis Vilmanis U 1.5Player Shots on Goal | −105 | won | |
| Mon 12 Oct | Florida Panthers at Buffalo SabresNHL | Ukko-Pekka Luukkonen U 25.5Player Saves | −120 | won | |
| Mon 12 Oct | Atalanta v VeneziaSoccer | Marco Carnesecchi Marco CarnesecchiGoalkeeper Saves | +100 | lost | |
| Mon 12 Oct | Atalanta v VeneziaSoccer | Jonathan Rowe Jonathan Rowe1+ Shots on Target | −125 | won | |
| Sun 11 Oct | Pachuca v NecaxaSoccer | Oussama Idrissi Oussama Idrissi1+ Shots on Target | −132 | won | |
| Sun 11 Oct | Carolina Hurricanes at Philadelphia FlyersNHL | David Jiricek U 0.5Player Total Points | −250 | won | |
| Sun 11 Oct | Carolina Hurricanes at Philadelphia FlyersNHL | Denver Barkey U 1.5Player Shots on Goal | −200 | won | |
| Sun 11 Oct | Carolina Hurricanes at Philadelphia FlyersNHL | William Carrier U 0.5Player Total Points | −211 | lost | |
| Sun 11 Oct | Carolina Hurricanes at Philadelphia FlyersNHL | Jamie Drysdale U 0.5Player Total Points | −203 | won | |
| Sun 11 Oct | Carolina Hurricanes at Philadelphia FlyersNHL | Travis Sanheim U 1.5Player Shots on Goal | −200 | won | |
| Sun 11 Oct | Carolina Hurricanes at Philadelphia FlyersNHL | Mark Jankowski U 1.5Player Shots on Goal | −157 | lost | |
| Sun 11 Oct | Carolina Hurricanes at Philadelphia FlyersNHL | Joseph Woll O 28.5Player Saves | −115 | lost | |
| Sun 11 Oct | Atlético San Luis v Santos LagunaSoccer | Fran Villalba Fran Villalba1+ Shots on Target | +129 | won | |
| Sun 11 Oct | Vancouver Canucks at New York RangersNHL | Jonathan Lekkerimäki U 1.5Player Shots on Goal | −175 | lost | |
| Sun 11 Oct | Vancouver Canucks at New York RangersNHL | Jonathan Lekkerimäki U 0.5Player Total Points | −300 | lost | |
| Sun 11 Oct | Seattle Kraken at Washington CapitalsNHL | Timothy Liljegren U 0.5Player Total Points | −278 | won | |
| Sun 11 Oct | Seattle Kraken at Washington CapitalsNHL | Ilya Protas U 0.5Player Total Points | −138 | lost |
What the model claimed, and what happened
Mon 12 Oct · 223,260 settled predictionsClose to honest on average — 32.6% claimed against 33.9% observed. That average hides the finding below it.
The further we depart from the market, the more we overstate
Every prediction, grouped by the rating the model gave it. Low ratings track the price closely and are calibrated; the strongest depart furthest, and that is where the claim and the outcome come apart.
Why a rating is capped rather than trusted at the top of its range, and why this site publishes no return by rating: the ordering of these buckets is stable, their levels are not.
Calibration, market by market
Claimed probability along the bottom, the fitted correction up the side; the faint diagonal is perfect calibration. 18 markets carry a fitted curve — the rest are measured but not yet fitted, and are in the table below rather than dropped.
Where we are worst
Every market we publish a board for, sorted by skill — the Brier score against a constant guess at the same base rate. Below zero the model lost to that guess: 15 of 73 markets. Nothing is removed and nothing is sorted flatteringly.
23 sit on fewer than 500 settled predictions. A skill score swings hard on a small sample, so those rows follow the rest rather than leading them — reported, not ranked.
When the model changed
Every day the blend’s source weights were rewritten. A record is only meaningful against a model that was not quietly swapped underneath it.