Public · no key
The scorecard
Every call a model makes is logged before the outcome is known and graded against our own market rows at the due day. A line is proven at 30 graded calls with a 60% hit rate. Nothing below is hidden, including the lines that are not proven yet.
Calls logged
6,246
Calls graded
0
Proven gate
30 @ 60%
As of
2026-10-27
| Model | Window | Metric | Graded | Hit rate | Pending | Status |
|---|---|---|---|---|---|---|
| Pulse | 7 d | asking median | 0 | — | 464 | not yet |
| Pulse | 14 d | asking median | 0 | — | 464 | not yet |
| yg | 14 d | asking median | 0 | — | 1329 | not yet |
| yg | 30 d | asking median | 0 | — | 1329 | not yet |
Same data as GET /api/v1/scorecard, cached 15 minutes.