Engineer’s bot buys ‘No’ on low-priced non-sports Polymarket

Engineer Sterling Crispin built and deployed a trading bot on Polymarket that automatically buys ‘No’ on non-sports markets when the best ask for ‘No’ is below $0.65 and holds positions to resolution. Crispin announced the bot on X and cited a 73.4% rate of resolved markets ending in ‘No.’

Polymarket frames many questions as whether a specific event will occur by a deadline. Polymarket’s accuracy page states “there are usually more ways for something not to happen than to happen in one exact way.” An analysis of more than 2,300 closed Polymarket positions shows markets open 90 to 180 days resolved ‘No’ about 73.5% of the time, while markets open less than a week resolved ‘No’ roughly 52% of the time.

The bot’s source code limits activity to non-sports markets and only places buys when the best ask for ‘No’ is under $0.65. The code was published under a CC0 license and the repository includes a live dashboard, a Heroku deployment guide and a script named wallet_history.py to track realized performance.

Because contract payouts are $1 for a correct side, the entry price affects the break-even win rate. Buying ‘No’ at $0.40 requires about a 42% win rate to break even after transaction costs on Polygon; buying at $0.60 requires about a 59% win rate. The repository’s default 65-cent cap screens out markets where the crowd has largely priced in a ‘No’ outcome.

The bot’s default position size is 2% of capital, and the code allows a configurable maximum up to 100% per position. The repository includes back-of-envelope examples that show higher projected annualized returns with larger position sizes, but Crispin has not published verifiable live profit figures or provided an identified trading wallet.

A 2026 CertiK report estimated the prediction-market sector at $63.5 billion a year, while Bernstein projected prediction markets could hit $1 trillion by 2030. Polymarket was valued at about $9 billion after an Intercontinental Exchange investment in October 2025. Polymarket’s all-time leaderboard shows the top 10 wallets have realized more than $113 million in profit, primarily from event-specific bets rather than blanket strategies.

The bot’s code and documentation make the trading approach reproducible. Outcomes will depend on market selection, chosen price thresholds, transaction costs and competition. Without an identified trading wallet, there is no public record of the bot’s performance at scale.