Verified record
Every settled prediction, wins and losses, computed from the append-only ledger. Nothing is edited after publication; corrections create a new version.
83
settled predictions
49
wins
34
losses
59.0%
hit rate
2.03
avg. price at lock
0.502
Brier
10.0%
ROI, 1 unit each
—
CLV · not tracked yet
Every official WinPIQ pick is recorded. Wins and losses.
Stored before kickoff with a proof hash; locked at kickoff; corrections create a new version with a full audit trail. No cherry-picking is possible.
Probability bucket performance
Do predictions rated X% actually win X% of the time? Computed from the settled rows below.
rated 30–40%
0.0%
1 predictions · avg 39.4%
rated 40–50%
0.0%
3 predictions · avg 44.4%
rated 50–60%
40.0%
10 predictions · avg 53.9%
rated 60–70%
100.0%
2 predictions · avg 69.3%
When we say 60%, does it happen 60% of the time?
Backtest0 matches
0 matches
Live, settled83 matches
30–40%38% → 0% n=2
40–50%44% → 47% n=17
50–60%54% → 60% n=47
60–70%65% → 77% n=13
70–80%72% → 67% n=3
80–90%82% → 100% n=1
predicted observed
By confidence
| confidence | n | hit rate | Brier |
|---|---|---|---|
| <50 | 2 | 100.0% | 0.452 |
| 50-69 | 16 | 50.0% | 0.573 |
| 70-84 | 40 | 65.0% | 0.487 |
| 85+ | 25 | 52.0% | 0.486 |
Show me the receipts 3 shown
| published | match | pick | prob. | conf. | price at lock | model | result | proof |
|---|---|---|---|---|---|---|---|---|
| 29 Aug, 18:15 | Auxerre v AngersLigue 1 | 1 | Model probability 52% | 87 | 2.12 | ens_v0.1 | LOSS | MQ-000112.v2.v3.v4.v5 |
| 29 Aug, 18:15 | Brest v ToulouseLigue 1 | 1 | Model probability 41% | 87 | 2.51 | ens_v0.1 | LOSS | MQ-000113.v2.v3.v4 |
| 29 Aug, 18:15 | Lyon v Le HavreLigue 1 | 1 | Model probability 57% | 85 | 1.45 | ens_v0.1 | LOSS | MQ-000109.v2.v3.v4 |