Our Record, Graded in Public
Monday, November 16 •
Founder’s Picks
The founder’s own picks across every sport, vetted against our projections and settled on an open ledger — every win and every loss, at the price actually paid. Members get each new card the moment it posts.
See the full ledger →NRFI Record
Record by Window
| Window | Edge calls vs market | Directional accuracy |
|---|---|---|
| 2026 season | 416-41350.2% | 798-75251.5% |
| Last 30 days | 2-166.7% | 4-266.7% |
| Last 7 days | — | — |
| Yesterday | — | — |
Engine History
We upgraded the simulation engine mid-season. V1 results are frozen exactly as they finished, not rewritten, so the two eras can be compared honestly.
| Engine | Dates | Edge calls vs market | Directional accuracy |
|---|---|---|---|
| V1 engine (frozen) | Thursday, March 26 – Monday, April 20 | 20-2643.5% | 120-10154.3% |
| V2 engine (live) | Tuesday, April 21 – Wednesday, October 28 | 396-38750.6% | 678-65151.0% |
The Rest of the Boards
HR Power Score
Top-rated bats homered 3.8x more often than bottom-rated over the final 30 days of the 2026 regular season — graded daily from official scoring, top 10 vs bottom 10 of every slate. Playoff slates are excluded: with half the league gone, top-10-of-the-day stops meaning anything, so the number freezes where the regular season ended.
Strikeouts, Hits & Bases, and Full Game Projections publish descriptive distributions rather than graded picks; their calibration tracking lives inside each board. NFL player props are projections-only on the public board — any pick we publish as a pick lands in the Founder’s ledger above, where it is graded at the price paid. Nothing gets promoted to a public record until it has survived that grading in the open.
How NRFI Grading Works
Edge calls vs market: a game counts when lineups are confirmed and our scoreless-first-inning probability beats the market-implied probability by more than 0.1%. A win means the first inning finished scoreless; a loss means a run scored.
Directional accuracy: a game counts whenever our probability is above 50.1%. With 10,000 simulations per game the margin of error is about 0.5%, so 51%+ is a statistically meaningful call. At or below 50% we have no opinion and the game is excluded (892 such games this season).
Predictions lock at first pitch and results are graded automatically from official MLB scoring. Every qualifying game counts — there is no hand-picking which results make the record. Read how the projections are built for the full methodology.
The 2026 MLB season is wrapped
Daily NRFI boards return for spring training. Until then, every graded result from the season stays up in the archive.
Browse the results archive →See the full day-by-day results archive →
For entertainment and informational purposes only. These are data-driven projections, not guarantees. 21+. If you or someone you know has a gambling problem, call 1-800-GAMBLER. Please bet responsibly.
