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 5 Oct.
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 +1.41%. At one unit a bet instead, they returned −4.74%.
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 5 Oct | Los Angeles Lakers at Sacramento KingsNBA | Sacramento Kings +3.5Point Spread | −110 | lost | |
| Mon 5 Oct | Atlanta Falcons at New Orleans SaintsNFL | Bijan Robinson Bijan Robinson70+ Rushing Yards | −272 | won | |
| Mon 5 Oct | Atlanta Falcons at New Orleans SaintsNFL | Chris Olave Chris Olave70+ Receiving Yards | −227 | won | |
| Mon 5 Oct | at NHL | Jeremy Swayman U 24.5Player Saves | −110 | won | |
| Mon 5 Oct | at NHL | Brandon Hagel U 0.5Player Power Play Points | −175 | won | |
| Mon 5 Oct | at NHL | Jamie Drysdale U 1.5Player Shots on Goal | −190 | won | |
| Mon 5 Oct | at NHL | Travis Sanheim U 1.5Player Blocked Shots | +132 | lost | |
| Mon 5 Oct | Memphis Grizzlies at Atlanta HawksNBA | Atlanta HawksMoneyline | −164 | lost | |
| Mon 5 Oct | Chicago White Sox at Cleveland GuardiansMLB | Anthony Kay O 1.5Pitcher Total Hits Allowed | −135 | lost | |
| Sun 4 Oct | at NHL | Kevin Lankinen U 27.5Player Saves | −115 | won | |
| Sun 4 Oct | Detroit Lions at Carolina PanthersNFL | Amon-Ra St. Brown Amon-Ra St. Brown60+ Receiving Yards | −250 | won | |
| Sun 4 Oct | at NHL | Lars Eller U 1.5Player Shots on Goal | −185 | lost | |
| Sun 4 Oct | Atlanta Braves at Los Angeles DodgersMLB | Shohei Ohtani U 0.5Batter Home Runs | −280 | won | |
| Sun 4 Oct | Atlanta Braves at Los Angeles DodgersMLB | Max Muncy U 0.5Batter Singles | −247 | won | |
| Sun 4 Oct | at NHL | J.T. Miller U 0.5Player Power Play Points | −200 | won | |
| Sun 4 Oct | at NHL | Alexis Lafreniere U 0.5Player Power Play Points | −300 | won | |
| Sun 4 Oct | at NHL | Nick Schmaltz U 0.5Player Power Play Points | −275 | lost | |
| Sun 4 Oct | at NHL | Logan Cooley U 0.5Player Power Play Points | −250 | won | |
| Sun 4 Oct | Kansas City Chiefs at Las Vegas RaidersNFL | Tre Tucker U 0.5Total Rushing Attempts | −105 | won | |
| Sun 4 Oct | Kansas City Chiefs at Las Vegas RaidersNFL | Malik Benson U 4.5Total Rushing Attempts | −135 | won | |
| Sun 4 Oct | Kansas City Chiefs at Las Vegas RaidersNFL | Malik Benson U 16.5Total Rushing Yards | −110 | won | |
| Sun 4 Oct | Los Angeles Chargers at Seattle SeahawksNFL | Jaxon Smith-Njigba Jaxon Smith-Njigba75+ Receiving Yards | −295 | won | |
| Sun 4 Oct | Denver Broncos at San Francisco 49ersNFL | Marvin Mims U 1.5Total Receptions | −200 | won | |
| Sun 4 Oct | Denver Broncos at San Francisco 49ersNFL | Christian McCaffrey Christian McCaffrey50+ Rushing Yards | −267 | won | |
| Sun 4 Oct | Denver Broncos at San Francisco 49ersNFL | Mike Evans Mike Evans30+ Receiving Yards | −260 | won |
What the model claimed, and what happened
Tue 6 Oct · 253,142 settled predictionsClose to honest on average — 32.4% claimed against 33.6% 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. 15 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: 13 of 70 markets. Nothing is removed and nothing is sorted flatteringly.
18 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.