Automated Polymarket Accounts Placed 81% of Orders and Made $247M

The October 1 study reviewed Polymarket’s on-chain trading history and separated 125,429 high-frequency accounts from roughly 2.9 million retail accounts. Galaxy used more than 50 orders per active day as the cutoff.

That does not mean every account in the group was necessarily a bot. The threshold is a proxy for likely automation. Still, the difference in activity was large: those accounts represented only 4.1% of the sample but placed 80.8% of all orders and generated 41% of notional volume.

By 15M’s calculation, their share of orders was almost 20 times their share of accounts.

Retail Accounts Lost $338.9M

The retail side finished with a combined loss of $338.9 million. Galaxy found that 69.2% of those accounts ended below break-even.

Most individual losses were far smaller than the total suggests. The median result was about minus $3, while half of the accounts finished between a $36.64 loss and a $0.40 gain.

There was also a noticeable change in activity after losing trades. Stork, which supplied the indexed historical data used in the research, said 15.2% of accounts made no new trade during the following 30 days after a loss.

After a winning trade, that figure was 6.1%.

Sports Traders Had the Lowest Profit Rate

Galaxy also looked at users who concentrated at least 60% of their activity in one topic. Around 44.1% of accounts met that definition.

Sports specialists came out at the bottom, with only 25.1% finishing in profit. Tech and science traders were at the other end of the table at 41.2%.

That leaves a 16.1-percentage-point gap between the two groups. Generalist accounts, meanwhile, had a 30.4% profit rate.

Across the full sample, the aggregate results of the likely automated and retail groups were $585.7 million apart.