When this trader buys at $0.70, they imply 70% probability. Perfect calibration = the event happens 70% of the time.
| Bucket | Bets | Expected | Actual | Error |
|---|---|---|---|---|
| 0.30-0.40 | 1 | 37% | 0% | 36.7% |
| 0.40-0.50 | 8 | 45% | 0% | 44.8% |
| 0.50-0.60 | 6 | 53% | 0% | 53.3% |
| 0.60-0.70 | 4 | 64% | 0% | 64.4% |
On-chain verification: wallet age 71 days, 2000 txs, provenance grade C. Bot score: 0/100, wash trading score: 0/100.
Polymarket on-chain coverage: $1,054 in / $0 out across 189 withdrawal tx since 2026-02-13.
3100 total trades across 19 markets.
19 bets on resolved markets available for calibration scoring.
Calibration error: 51.2% — needs improvement.
Skill: 0/100 (calibration quality). Variance: 8/100 (higher = more volatile returns).
Brier Skill Score: -7.7% vs naive baseline (>0% = better than always predicting base rate).
Brier decomposition: REL=0.2687 RES=0.0000 UNC=0.0000.
Log loss: 0.7340 (skill: -9182.1% vs naive). Lower log loss = better calibration on rare events.
Timeliness: avg entry 0.6 days before resolution, 0% early mover.
Below average. The data shows poor calibration, thin evidence, or both. When this trader expresses high confidence, events don't happen at the rate they imply.
Confidence: D/42 ± 25 (very_low confidence, 19 resolved bets). Low confidence — take this score with a grain of salt until more markets resolve.
Methodology: Brier Score Decomposition (Murphy 1973), Log Loss, On-Chain USDC Verification. Same approach used by IARPA to identify superforecasters.