This is the scoreboard. Every number the model publishes is locked before the outcome is known, graded when its 90-day window closes, and never edited afterward. The misses stay up next to the hits. If you want to know whether to trust this site, this page is the answer, and it will say so either way.
Fourteen of 28 reached the majors. The model ranked all 28 in the correct order, and all six top-band players debuted. The probabilities ran low, which produced a 0.269 raw Brier, corrected to 0.148 by the v0.5 recalibration. Both numbers are on the log, with a named reason for every miss. Full results ›
The first cohort scored on the recalibrated v0.5 table, so September 9 is the model’s first true out-of-sample test. Thomas White leads at 62 percent and is on the injured list, a likely miss we flagged in July, before it resolves. The fourteen ›
Fifteen featured predictions plus a 50-player sweep of the healthy Top 100, scored specifically to put real outcomes behind the low-probability bands. When it grades, the v1.0 training set locks at 107 outcomes. The sixty-five ›
The PIX board keeps its own record. When a tracked player debuts, his card is frozen as a receipt: the live number at debut, the number 30 days prior, and the sourced events that moved it. Recent receipts, including Luis Lara, the call-up the daily scan missed and owned in public, live on the board, and every point move ever applied sits in the append-only event ledger.
September 9: Sub-Batch B resolves, fourteen outcomes, the first clean test of v0.5. October 8: Sub-Batch C resolves, sixty-five outcomes, and the 107-outcome training set for v1.0 locks. Both checkpoints publish the same way June 18 did: original numbers, Brier score, named misses.
The deeper material lives one click away: the calibration log holds every graded prediction, the June 18 technical report shows the recalibration math, and Inside v0.5 documents the whole model, including what we know is still wrong with it.