The model went 68.8% (110/160) on the full card yesterday. Here's the tape.
๐ THE LEDGER
Every pick the model graded. Win or lose, it goes on the board.
| Window | Full Card | Batters | Pitchers |
|---|---|---|---|
| Yesterday | 68.8% (110/160) | 64.8% (81/125) | 82.9% (29/35) |
| This week | 69.1% (687/994) | 68.8% (528/768) | 70.4% (159/226) |
| This month | 70.1% (2559/3652) | 71.1% (2022/2843) | 66.4% (537/809) |
| This season | 69.9% (2917/4176) | 70.6% (2353/3331) | 66.7% (564/845) |
Batter markets (season): Hits 71.9% ยท Hits+Runs+RBI combo 72.4% ยท Runs 55.2% ยท Runs+RBI 63.2% ยท Total Bases 54.5% ยท Home Run 23.1% ยท RBI 50.0%
Pitcher markets (season): Strikeouts 70.8% ยท Walks 65.1% ยท Outs/Length 58.8%
๐๏ธ THE WRAP โ what we saw that the line didn't
๐ฏ Against the grain
Taj Bradley (Minnesota Twins) to work deep into the game โ and did. The matchup screamed coin-flip to the public, but the model had it flagged as a calculated dart. It cashed.
๐ง The honest miss
Matt Waldron (San Diego Padres) was top of the board to clear his strikeout number and came up empty. We don't bury these โ the model's edge is built over a full season, not any single night.
โก The matchup that didn't land
We leaned into a soft-matchup hitter cluster and it came up short โ 51% (41 bats). Some days the spot is right and the results aren't. We log it and move on.
Grades posted every morning. Today's plays drop as lineups lock โ watch for the Early / Mid / Late editions.