Luck or skill on Polymarket? Take away the best market and look again
A large realised profit can be one market that went right. The simplest test of a Polymarket record is to remove the single most profitable market and see what is left — and most records do not survive it.
The test
Take a trader's realised profit over a window and subtract the profit from their single best market. If what remains is still positive, they made money across many markets. If it is zero or negative, the whole record rests on one outcome — and following them means waiting for the next one.
How often records fail it
In PMWallets' 30-day window published on 2 October 2026, there were 74,100 entities with at least ten markets won or lost. For 77.8% of them, the remaining settled markets, without the single best one, added up to no net profit. Of the 23,663 that were profitable overall, 7,219 — about 30% — were no longer profitable once their best market was removed.
Where PMWallets shows it
Every entity carries this figure (pnlExTop in the API) beside its realised profit, and it decides what an address costs to unlock: a wallet that made nothing without its best market is free to unlock, because there is little to learn from it; the rest are priced from $1 to $5 in bands by how much they made without it.
Read it with the win rate
Profit without the best market says whether the money came from more than one place; the lower bound of the win rate's confidence interval says whether the wins did. A trader who passes both — positive without the best market, a lower bound that stays high over many decided markets — has a record worth studying. One who passes neither has a story.
Questions
How do I know if a Polymarket trader is lucky?
Remove their single most profitable market from the window and see whether the rest is still profitable, and read their win rate by the lower bound of its confidence interval over many decided markets. One lucky market fails the first test; a few lucky wins fail the second.
What is pnlExTop?
The PMWallets API field for realised profit without the entity's single most profitable settled market in the window.
How many Polymarket traders depend on one market?
On PMWallets' 30-day window of 2 October 2026, 77.8% of the 74,100 entities with at least ten markets won or lost had no net profit left without their best market.