Betting · player props
Prop Edges
Every player prop where our backtested model, blended with the market's own price, beats the sportsbook's break-even by at least three points, sorted by market and by game, with the reasons. Every one we post is tracked and graded, wins and losses alike.
The record
Every pick, graded →How this works
Each player's projection is built only from games already played: his rate in every situation, weighted toward recent games; his recent minutes, re-shared across tonight's actual lineup so a teammate's absence moves minutes to the players who dress; the opponent; and the game total. Goalie saves come from the opponent's shot volume, his own team's shots allowed and his save percentage.
Our number is never used alone. It is blended with the sportsbooks' own price, with weights fitted on the 2025-26 season. That also corrects the market's lean toward overs (the public bets overs, and the prices follow), which is why most of our plays are unders. Tested on February to April 2026, games the model never saw, plays with an edge of three points or more returned about 5% at the median price across books.
A play qualifies when the blended chance beats the price's break-even by at least 3 points. Games are priced about two and a half hours before puck drop, after line rushes, and every play that clears the bar is posted, one per player (his biggest edge). Once posted, a play stays posted.
Every posted play is recorded once, at the price when it was posted, and graded after the game on the full box score, empty-net goals included. A player who doesn't play, or a goalie who doesn't start, is a void, as the books settle it. Beat the closing line is how often the price moved our way before puck drop: over a season the best sign the edges are real. Past results don't guarantee future ones; bet what you can afford to lose.
Want to price a line yourself? Use the props calculator.