'Nothing Ever Happens': This Bot Always Bets 'No' on Polymarket, And It Has a Point

Someone built a prediction marketplace bot with a doctrine borrowed from the astir cynical idiosyncratic astatine each meal table—the 1 who ever argues “nothing ever happens.” The bot, created by technologist Sterling Crispin, automatically buys "No" connected each non-sports marketplace connected Polymarket it finds and holds to resolution. The full trading strategy fits successful 1 sentence.

Turns out, that mightiness really beryllium capable to marque your bags grow—kinda.

The premise rests connected a fig Polymarket publishes openly: 73.3% of each resolved markets connected the level extremity successful "No." Crispin cited a astir identical figure—73.4%—when helium announced the bot connected X earlier this week, a station that collected 3.1 cardinal views.

Polymarket's ain accuracy page confirms the data, explaining that "there are usually much ways for thing not to hap than to hap successful 1 nonstop way."

The reflection cuts to thing existent astir however prediction markets are built. Most questions connected Polymarket are framed astir circumstantial events materializing by a deadline: Will a peculiar authoritative resign, volition a circumstantial measure pass, volition prices breach a circular number. The presumption quo has a structural advantage—it lone needs to hold, portion the Yes result needs 1 precise concatenation of events to implicit connected schedule. When the deadline passes and thing materializes, the bets connected “No” collect.

The effect is amplified successful longer-running markets. Our ain speedy investigation of implicit 2,300 closed Polymarket positions shows that markets unfastened for 90 to 180 days resoluteness “No” astatine a complaint of 73.5%, astir matching the platform's wide published figure. Short-duration markets—under a week—flip person to coin-toss territory, resolving "No" conscionable 52% of the clip (still a tiny advantage, but vantage nonetheless).

The longer a marketplace stays open, the much clip the satellite has to simply bash nothing.

The bot is not, arsenic immoderate dismissive replies connected Crispin's station assumed, conscionable spraying superior astatine random. The root codification reveals circumstantial filters: It lone targets non-sports markets, which person little “No” rates, and lone buys “No” erstwhile the champion inquire sits beneath $0.65. (All bets connected prediction markets similar Polymarket, Kalshi, oregon Myriad—operated by Decrypt’s genitor institution Dastan—resolve to $1, with users buying shares successful a “yes” oregon “no” presumption astatine little than $1 per stock and with the terms implying existent odds.)

That terms headdress does existent work. Buying “No” astatine $0.40 lone requires a 42% triumph complaint to interruption adjacent aft state fees connected Polygon; buying astatine $0.60 pushes the break-even threshold to 59%. By capping entries beneath 65 cents (those with implied likelihood supra 65%), the bot screens retired markets wherever the assemblage has already priced successful the apt “No” outcome, and wherever the borderline has evaporated.

The default 2% presumption sizing is simply a blimpish default, not a limit—the bot allows immoderate sizing up to 100%. At 10-20% sizing, yearly returns scope 16-33%, which is competitory with progressive trading strategies.

Prediction markets person grown into a $63.5 billion-a-year industry, according to a 2026 CertiK report, and Polymarket was valued astatine $9 billion aft NYSE genitor Intercontinental Exchange invested $2 cardinal successful October 2025. Markets of that standard don't permission exploitable inefficiencies sitting successful plain show for long.

The platform's apical earners don't stake “No” connected everything. Polymarket's all-time leaderboard shows the apical 10 traders person collectively made implicit $113 cardinal successful profit—and astir each of them are bettors with genuine domain expertise. Theo4, the starring wallet has made its ain luck connected Yes/No governmental markets by placing likelihood successful Yes oregon No depending connected the event.

Image: Polymarket screenshot

The statistical borderline Crispin identified is real. But turning it into a luck requires much than a Python publication acceptable to "buy No."

The bot's borderline (if any) comes from: Filtering for markets wherever “No” is priced beneath its existent probability (the 65 cent cap), avoiding markets wherever “Yes” is underpriced (sports, finance) and selecting longer-duration markets.

The bot's GitHub repository is public, licensed CC0, and includes a unrecorded dashboard, a Heroku deployment guide, and a publication called wallet_history.py for tracking realized performance. As of publication, Crispin has not shared his wallet code oregon existent nett figures from unrecorded trading.