MLB Edge Board

Opening line = first snapshot this browser recorded for the game (stored locally, persists across days). Flags: TOTAL ±1.0+, ML ±20¢+. Reverse line movement needs bet-percentage data no free API provides — paste your own in Settings → bet % JSON and RLM rows will light up.

Implied runs derived from the consensus total + moneyline (see Settings → math notes). Season & L15 = actual runs/game. vs-hand column scales season runs/game by the team's OPS split against tonight's starter's handedness.

Model: player HR/PA × pitcher HR/9 factor × park HR factor × wind-out-to-CF factor, over ~4.1 PA. Paste a Baseball Savant leaderboard CSV in Settings to add barrel%, HR/FB and pull% multipliers. Edge = model probability − de-vigged prop probability. Load props with the button above (uses Odds API credits).

Each row is a starter; the last four columns model the OPPOSING team hitting ≥1 HR in innings 1–5 off him. λ = opp HR/game (scaled by handedness split) × 5/9 × starter factor (season HR/9 blended 70/30 with last-5, dampened ^0.75, 85% of F5 innings) × park × wind; P = 1 − e^−λ. The Odds API doesn't carry F5 team-HR markets, so the fair price is shown for shopping against your book's "home run in first 5" props manually. L5 from game logs; ≥5 IP = share of starts completing five innings.

"Model" here = de-vigged consensus probability across all books (median). Edge = model − implied prob of the single best price. This is a market-relative edge (are you beating the closing consensus), not a fundamental projection.