All-time record
- Wins / losses
- 0 / 0
- Net units
- +0.00u
- ROI
- —
- Hit rate
- —
0 picks · 0 push · 0 void · 0 pending
Last 30 days
- Wins / losses
- 0 / 0
- Net units
- +0.00u
- ROI
- —
- Hit rate
- —
0 picks · 0 push · 0 void · 0 pending
Every graded pick
No graded picks yet.
Calibration
Observed wins among decided picks, grouped by the original model probability.
| Forecast range | Decided picks | Mean forecast | Actual hit rate |
|---|---|---|---|
| 0–10% | 0 | — | — |
| 10–20% | 0 | — | — |
| 20–30% | 0 | — | — |
| 30–40% | 0 | — | — |
| 40–50% | 0 | — | — |
| 50–60% | 0 | — | — |
| 60–70% | 0 | — | — |
| 70–80% | 0 | — | — |
| 80–90% | 0 | — | — |
| 90–100% | 0 | — | — |
Methodology
A negative-binomial model blends recent and season shot rates with the previous season, then adjusts for ice time, opponent shots allowed and venue. Early-season projections shrink toward the measured position average.
The feed compares each side with the median de-vigged probability for that line and uses the best quoted price. Current cutoffs are 4 percentage points of edge and 3% expected value.
Each pick risks a flat 1u at its recorded price. A final NHL boxscore settles the pick. Ties push and players who did not play are void. ROI divides net units by settled stakes including pushes. Hit rate and calibration use wins and losses. Both time windows use the original slate date.