Eighty-five million dollars is the figure people now quote for one French trader's profit on the 2024 US election. The first count, drawn from four accounts, came to forty-eight million.
The trading did not change between those two numbers. The number of accounts being counted did.
After that trade, prediction-market leaderboards started getting read as shopping lists. Polymarket runs one, and it lets you rank by profit over a day, a week, a month, or all time.
The case for reading the daily version is real. A large one-day move usually means a large position, and size is where information shows up first, if it shows up at all.
Size and record are two different things, though. A wallet can top a single day's table on one lucky resolution with nothing behind it, and a wallet with years of patient work can miss that table entirely because it never has a big enough day.
So the record axis gets its own list. A wallet joins it with ten closed markets behind it, a win rate above 55 percent, a lifetime result in the black, and an average buy price low enough to rule out farming.
Between yesterday's daily snapshot and today's, the twenty largest one-day P&L moves we track included exactly two wallets from that second list. Eighteen carried no tag at all.
Across the past week that daily count has stayed at six or lower, which leaves at least fourteen of any day's twenty sitting outside the filter.
The filter is narrow on purpose, and missing it is not a verdict. A wallet can be millions of dollars up over its life and still fall out, because the filter also asks about win rate and about where that wallet tends to buy.
Every one of these lists prints the money. Which window it came from is the part the reader has to supply.
The second list is what the leaderboard shows by default: a record filter rather than a day's scoreboard. Its four thresholds are printed above the first row. orcalayer.com/leaderboard
