Primary read
Rebounds diff (5g) (+3.7%)
Walk-forward backtests, calibrated probabilities, market benchmarking, and saved daily boards that can be audited after the fact. No touts — just logged predictions.
Primary read
Rebounds diff (5g) (+3.7%)
Updated May 20, 6:38 AM EDT
Headline accuracy follows the active evaluation artifact.
Through May 19, 2026
Headline and rolling chart read from reports/backtest_unified_playoff_gbm.csv.
Generated May 20, 6:38 AM EDT
Latest slate: 2026-05-20 · today
Walk-forward backtesting across the full evaluation window, never a random split.
Pregame-safe features only, with rolling windows excluding the current game.
Closing-line comparisons so performance is framed against a real market baseline.
20-game rolling accuracy from the same postseason-aware artifact as the headline metric.
1,214 games · Oct '25 – Apr '26 · today
| Date | Matchup | Home win % | Pick | Actual | Result |
|---|---|---|---|---|---|
| May 20, 2026 | SAS@OKC | 79% | OKC | OKC | Correct |
| May 19, 2026 | CLE@NYK | 84% | NYK | NYK | Correct |
| May 18, 2026 | SAS@OKC | 70% | OKC | SAS | Miss |
| May 17, 2026 | CLE@DET | 78% | DET | CLE | Miss |
| May 15, 2026 | DET@CLE | 74% | CLE | DET | Miss |
| May 15, 2026 | SAS@MIN | 44% | SAS | SAS | Correct |
| May 13, 2026 | CLE@DET | 75% | DET | CLE | Miss |
| May 12, 2026 | MIN@SAS | 78% | SAS | SAS | Correct |
| May 11, 2026 | DET@CLE | 40% | DET | CLE | Miss |
| May 11, 2026 | OKC@LAL | 31% | OKC | OKC | Correct |
| May 10, 2026 | SAS@MIN | 46% | SAS | MIN | Miss |
| May 10, 2026 | NYK@PHI | 40% | NYK | NYK | Correct |
| May 9, 2026 | DET@CLE | 45% | DET | CLE | Miss |
| May 9, 2026 | OKC@LAL | 36% | OKC | OKC | Correct |
| May 8, 2026 | NYK@PHI | 52% | PHI | NYK | Miss |
| May 8, 2026 | SAS@MIN | 47% | SAS | SAS | Correct |
| May 7, 2026 | CLE@DET | 77% | DET | DET | Correct |
| May 7, 2026 | LAL@OKC | 64% | OKC | OKC | Correct |
| May 6, 2026 | MIN@SAS | 84% | SAS | SAS | Correct |
| May 6, 2026 | PHI@NYK | 80% | NYK | NYK | Correct |