the Scoop trading stats
The conviction book: trades filtered to positions of $100+ and excluding sports markets. This strips out the small, for-fun bets and isolates the edge-driven, research-backed trades.
Average trade outcome
Winner vs. loser expectancy (conviction book)
Edge summary
What the conviction numbers mean in plain language.
- Brier score
- 0.1257 contract-weighted · 0.2261 unweighted, scored on stance-side entry VWAP against outcome. Lower is better and 0.25 is a coin flip, so the unweighted number is only just ahead of chance while the weighted one is far clear of it. That gap is sizing discipline: conviction went where the calls were right.
- Profit factor
- 3.24× means gross gains are over three times gross losses on the conviction book, against 1.94× for the full book. This page used to show 7.12×. That was not deterioration, it was fourteen more months of markets diluting a small flattering sample.
- Win rate context
- A 70.5% hit rate with average wins of $62.66 against average losses of -$21.02 produces strong positive expectancy on the conviction book.
- Blended ROI
- +65.7% return on capital deployed, roughly double the full book's +32.9%.
Full book vs. directional
Side-by-side comparison. Filtering for conviction trades lifts win rate, profit factor, and ROI across the board.
| Metric | Full book | Conviction ($100+, no sports) |
|---|---|---|
| Trade rows | 2,883 | 791 |
| Unique markets | 121 | 45 |
| Total contracts | 182,301 | 113,918 |
| Capital deployed | $57,909 | $41,327 |
| Total fees | $965.94 | $633.81 |
| Realized PnL | $17,171 | $20,531 |
| Gross wins | $35,364 | $29,708 |
| Gross losses | $18,193 | $9,177 |
| Profit factor | 1.94× | 3.24× |
| Trade win rate | 51.0% | 61.9% |
| Avg winning trade | $24.06 | $60.63 |
| Avg losing trade | -$12.88 | -$30.49 |
| Trade win/loss ratio | 1.87× | 1.99× |
| Market win rate | 57.0% | 71.1% |
| Avg winning market | $375.20 | $756.63 |
| Avg losing market | -$167.66 | -$283.19 |
| Market win/loss ratio | 2.24× | 2.67× |
| Swing PnL | $10,023 | $9,706 |
| Blended ROI | +29.7% | +49.7% |
Volume-weighted implied probability calibration curve
Conviction book only. Comparing market-implied odds to actual resolution rates by bin.
Click any bar or table row to see the underlying markets.
R = Resolved (held to expiry at 0¢ or 100¢). S = Swing (exited before expiry).
| Implied probability bin | N (markets) | R/S | Contracts | Avg implied | Actual rate | Edge |
|---|---|---|---|---|---|---|
0-10% | 7 | 3R/4S | 29,362 | 5.8% | 0.0% | -5.8% |
10-20% | 1 | 1R/0S | 3,186 | 19.3% | 0.0% | -19.3% |
20-30% | 8 | 3R/5S | 11,497 | 23.3% | 47.6% | +24.3% |
30-40% | 6 | 4R/2S | 3,902 | 36.5% | 56.0% | +19.5% |
40-50% | 6 | 2R/4S | 5,420 | 43.2% | 95.0% | +51.8% |
50-60% | 5 | 2R/3S | 10,145 | 52.2% | 88.0% | +35.8% |
60-70% | 4 | 3R/1S | 3,429 | 62.2% | 25.7% | -36.5% |
70-80% | 5 | 4R/1S | 13,891 | 75.2% | 97.5% | +22.2% |
80-90% | 3 | 1R/2S | 9,102 | 80.3% | 100.0% | +19.7% |
| Total | 45 | 23R/22S | 89,934 |
Footnote on the 0–20% bins: those are dominated by deliberate cheap-tail hedges (DHS Apr 15 lottery, DHS May 15 disaster insurance, the TRUEV tail ladder). Losing them is the designed cost of the hedge structure, not a calibration miss. Read the left end of the curve with that in mind.
All positions are on Kalshi. All returns are real. Track live stats at kalshi.com/ideas/profiles/fz2026.