47.7% from ten markets
At the median, the ten largest resolved results supplied nearly half of all positive market-level PnL.
Conviction Lens Research · 10 August 2026
Just 10 markets produced nearly half of every winning dollar
We stress-tested 170 profitable prediction-market trader histories with at least 100 resolved markets and 60 days of activity. For the median wallet, its ten best markets generated 47.7% of gross winning PnL.
24 of 170 failed
After removing each wallet's ten best markets, 24 records moved from positive to zero or negative reconstructed net PnL.
85.3% at the 90th percentile
For the most concentrated tenth of the cohort, ten markets supplied at least 85.3% of gross winning PnL.
Why concentration matters
Two wallets can show positive PnL while carrying very different histories. A result spread across many markets is different from one that depends on a few outliers. Public PnL and win rate cannot reveal that difference by themselves.
Method
The snapshot was taken at 08:00 UTC on 10 August 2026. Every wallet had positive reconstructed net PnL, complete accounting, a completed full-history backfill, no truncated data, at least 100 resolved markets and at least 60 days of activity. The median history contained 140 resolved markets and spanned 300 days.
Important limits
Removing the best markets is a concentration stress test, not an estimate of skill or luck. The cohort covers profiles that passed Conviction Lens publication and data-quality gates, not every prediction-market wallet. Historical performance is not a recommendation.