Crunching this week's bets
Running the numbers on every player, defense, line and forecast. This can take a few seconds.
For information and entertainment only, not betting advice. Chances come from a statistical model built on this season's and last season's box scores; they are estimates, not locks. Must be 21+ where legal. If gambling is a problem, call 1-800-GAMBLER.
The SportsProof Score rates every player 0–100 on how good a betting target they are this week, from real stats only: how much they're used, how efficient they are, how steady their numbers are, and this week's matchup. 50 is a typical player at the position.
No one fits here yet. A score needs 2 games with a role this season, or 4 last season for the same team.
Each bet type (catches, receiving yards, rushing yards, touchdowns, passing yards and so on) gets its own score from up to five parts, each measured against a typical player at the position last season:
Role — targets, carries or dropbacks per game, recent games counting most and pulled toward last season. Efficiency — yards per target or carry, catch rate, touchdowns per touch, heavily pulled toward average so one big game can't fake it. Reliability — how tightly the stat clusters week to week. Matchup — the defense, the Vegas game total and spread, weather, quarterback changes and injuries, the same adjustments our bet chances use. Trend — last 3 games against the player's usual.
The weights were fitted on the 2025 season using only what was known before each week, then checked on later weeks. Role carries most of every score. Recent trend barely predicted the next game once role was known, so it is shown (the ▲ / ▼ arrow) but counts little or nothing. The overall score blends a player's main bet types.
Hot is 80+, Warm 65–79, Neutral 35–64, Cold under 35. Questionable players lose about 15 points; players ruled out or without a game this week aren't scored. “Provisional” means fewer than 4 games behind the score, so it is pulled toward 50.
The SP Score ranks and explains. It is not a chance to hit and not a price edge: sportsbooks already price role and matchup. In testing, higher scores did hit more often, but they didn't beat our own bet chances, so they don't change which bets we suggest. Use the chance and the price for that.