Beta
    Strategy

    "Nothing Ever Happens": What a Viral Prediction Market Bot Teaches About Base Rates

    A viral bot that bets No on every Polymarket question reveals a key truth: 73% of prediction markets resolve No. Here's what that means for your trading strategy.

    By PredictionMarkets.usSunday, April 19, 20269 min read

    A developer built a bot that automatically bets "No" on every non-sports prediction market. The internet lost its mind — and the math is more interesting than the meme.

    The Bot That Broke the Internet

    On April 12, 2026, software engineer Sterling Crispin posted a single tweet that collected 3.1 million views. The premise was simple: he built a Polymarket bot that automatically buys "No" on every non-sports market and holds to resolution.

    "Why predict the future when 73.4% of all Polymarkets resolve as No?" Crispin wrote. "Stop over thinking it. Nothing Ever Happens."

    The repo — titled "Nothing Ever Happens" — attracted over 800 stars on GitHub within a week. It hit the front page of Hacker News with 468 points and 273 comments. It spawned articles across Yahoo Finance and other outlets. Polymarket itself now hosts a series of meta-markets literally called "Nothing Ever Happens," packaging bundles of unlikely events and letting traders bet that none of them occur.

    But beneath the meme is a genuinely useful insight about how prediction markets work — and why most traders get it wrong.

    The 73% Reality

    Polymarket publishes this data openly on its accuracy page: 73.3% of all resolved markets on the platform end in "No." Only 26.7% resolve "Yes."

    This isn't a bug. It's a structural feature of how prediction market questions are designed.

    Most markets are framed around specific events materializing by a deadline: Will a particular official resign by June? Will this bill pass before August? Will a specific price target get hit by December? The status quo has a built-in advantage — it only needs to hold. The "Yes" outcome requires a precise chain of events to complete on schedule.

    When the deadline passes and nothing materializes, "No" wins.

    This effect gets stronger over time. Data scientist Alex McCullough's analysis of Polymarket data found that markets open for longer periods have higher "No" resolution rates. Short-duration markets (under a week) resolve "No" about 52% of the time — barely above a coin flip. But markets open for 90 to 180 days resolve "No" at a rate of approximately 73.5%.

    The longer a market stays open, the more time the world has to simply do nothing.

    Why Knowing This Doesn't Make You Rich

    Here's where most people's intuition breaks down. If 73% of markets resolve "No," shouldn't you just buy "No" on everything and print money?

    No. And the reason is that the market already knows this.

    If a market resolves "No" 73% of the time, the "No" shares should be priced around $0.73. Buying at that price means you're paying 73 cents to win $1.00 about 73% of the time — roughly break-even before fees. There's no edge.

    The prediction market structure makes this intuitive: every "No" share you buy at 73 cents earns you 27 cents when it wins (73% of the time) but loses you 73 cents when it doesn't (27% of the time). The math:

    • Expected gain: 0.73 x $0.27 = $0.197
    • Expected loss: 0.27 x $0.73 = $0.197
    • Net expected profit: ~$0.00

    This is how efficient markets are supposed to work. The price already reflects the probability.

    What the Bot Actually Does

    Crispin's bot is smarter than the meme suggests. Reading the source code on GitHub reveals several specific filters:

    1. Price ceiling of $0.65. The bot only buys "No" when shares are priced at or below 65 cents — meaning the market thinks there's at least a 35% chance the event will happen. At this price, the break-even win rate is about 65%, well below the 73% historical base rate. That gap is where any potential edge lives.

    2. Non-sports markets only. Sports markets have lower "No" resolution rates because games actually happen and outcomes are more evenly distributed. The bot avoids them entirely.

    3. Paper trading by default. The bot ships in simulation mode. You have to explicitly set three separate environment variables to enable live trading with real money — a safety mechanism most crypto trading bots don't bother with.

    Crispin acknowledged this on Hacker News: "This has to buy below $0.73 long term. The bot has a configurable ceiling set at $0.65 and checks for new markets buying closer to $0.50."

    At those prices, the math shifts meaningfully. Buying "No" at $0.50 only requires a 50% win rate to break even — and if the true resolution rate is 73%, that's a 23-percentage-point cushion. Buying at $0.40 pushes the break-even threshold down to just 42%.

    The Uncomfortable Truth: Most Traders Lose

    The bot's thesis gains more weight when you look at how actual Polymarket traders perform. Independent on-chain analyst Andrey Sergeenkov analyzed 2.5 million wallet addresses using Dune Analytics and Polygon blockchain data, publishing his findings on April 6, 2026.

    The numbers are stark:

    • 84.1% of all Polymarket wallets are in the red
    • Only 2% have ever earned more than $1,000 total
    • Only 0.033% (about 840 wallets) have crossed $100,000
    • Less than 1% achieve $5,000 profit in any single month
    • Among those who do hit $5,000/month, only 2.6% sustain it for more than a year

    A separate December 2025 academic study by researchers Felix Reichenbach and Martin Walther — analyzing 124 million trades on Polymarket — found that only 30% of traders earn positive profits. Their data also documented a systematic tendency among traders to overtrade the "Yes" option.

    In other words: most traders are betting that things will happen in a world where they usually don't. The simplest possible strategy — to bet on "No" and walk away — would outperform the majority of the platform's users, even if it doesn't necessarily turn a profit.

    Polymarket's Meta-Markets: Betting on Nothing

    Perhaps the most fascinating development is that Polymarket itself has leaned into the premise. The platform now hosts a "Nothing Ever Happens: 2026" parlay market that bundles together a set of unlikely events — and lets traders bet that none of them occur.

    As of mid-April 2026, "Nothing" was priced at approximately 59 cents, with $489,000 in total volume traded. That price implies the market gives about a 59% chance that nothing on the list happens, and a 41% chance that at least one does.

    Previous editions haven't been kind to the "nothing ever happens" crowd. The Jerome Powell Edition resolved "No" — meaning something did happen. The US Strike Edition also resolved against "Nothing" after US military action met its trigger conditions. In a world of Iran ceasefires, trade wars, and CFTC rulemaking, the status quo is less stable than the meme implies.

    The Real Lesson: Think in Base Rates

    The lasting value of the "Nothing Ever Happens" bot isn't as a trading strategy. It's as a cognitive tool.

    Base-rate thinking is the practice of starting with the historical frequency of an outcome before layering on specific analysis. It's one of the most well-documented improvements you can make to your forecasting accuracy, and it's the one most people skip.

    When you see a market asking "Will X happen by Y?" and the price is 45 cents, your first question should be: What's the base rate for markets like this resolving "Yes"? If it's around 27%, then 45 cents looks expensive — the market might be overpricing the event's likelihood.

    This isn't theoretical. Polymarket's own accuracy data shows the platform achieves 94% accuracy a full month before resolution. The markets are generally well-calibrated. But calibration depends on prices reflecting true probabilities — and when traders systematically overweight "Yes" outcomes, "No" shares get mispriced.

    For prediction market newcomers, the base-rate takeaway is practical:

    • Most "Will X happen?" markets resolve No. That doesn't mean buy "No" blindly — it means don't overpay for "Yes."
    • Time is on the side of inertia. The longer a market runs, the more likely nothing changes. Deadline-driven markets are structurally biased toward "No."
    • Check the price, not the narrative. A compelling story about why something will happen doesn't matter if the shares already price in that story.
    • If you can't articulate why a market is mispriced, you probably don't have an edge. And that's fine — prediction markets are still valuable as information tools even if you never trade them.

    Important: US Access Restrictions

    Note that the "Nothing Ever Happens" parlay markets and most non-sports markets discussed in this article are on global Polymarket (polymarket.com), which is not accessible to US users. US users can access prediction markets through QCX LLC (d/b/a Polymarket US) for sports markets only, or through CFTC-regulated platforms like Kalshi for event contracts across multiple categories. For a full comparison, see our platform guide.

    FAQ

    Does the "Nothing Ever Happens" bot actually make money?

    As of publication, the bot's creator Sterling Crispin has not shared wallet addresses or live profit/loss data. The bot ships in paper-trading mode by default and carries a prominent disclaimer: "For entertainment only. Use at your own risk." The statistical edge from base rates is real, but exploiting it profitably requires buying below the correct price floor — which the bot's 65-cent ceiling attempts to do.

    Why do 73% of prediction markets resolve "No"?

    It's a function of question design. Most prediction market questions ask whether a specific event will happen by a specific deadline. There are usually more ways for something to not happen than to happen in one exact way within a defined timeframe. Polymarket's accuracy page explicitly notes this framing effect.

    Is buying "No" on everything a good strategy?

    Not without price discipline. If markets are correctly priced, buying "No" everywhere breaks even at best. The potential edge only exists when "No" shares are cheaper than the true probability — which means you need to identify markets where the crowd is systematically overpricing "Yes." The bot's configurable price ceiling (default $0.65) is designed for exactly this filter.

    What platforms are available to US users?

    US-based traders can access CFTC-regulated prediction markets including Kalshi (events, sports, and financial contracts), QCX LLC / Polymarket US (sports only), Coinbase (via Kalshi), Robinhood (via Kalshi), and Interactive Brokers / ForecastEx. See our full platform comparison.

    Can this analysis apply to Kalshi markets too?

    The 73% "No" resolution rate is specific to Polymarket. Kalshi's market design and category mix (including weather, economic data, and sports) may produce different resolution distributions. However, the base-rate principle — that most "Will X happen by Y?" questions resolve in the negative — applies broadly across prediction market platforms. A 2022 study of Kalshi data by Russ Clay found similar positive Brier Skill Scores, confirming that markets outperform naive base rates but still reflect the overall tendency toward "No" resolutions.

    Sources & Verification

    Back to Articles

    Related Articles

    Strategy

    Best Prediction Market Tools for Serious Traders: The 2026 Guide

    The definitive 2026 guide to prediction market analytics tools, terminals, and software for Kalshi and Polymarket traders. Ranked by what actually gives you edge.

    10 minApr 1, 2026
    Strategy

    UFC Prediction Markets: The Complete Guide to Betting on MMA (2026)

    Learn how to trade UFC fight outcomes on Kalshi and Polymarket. Real examples from UFC 326 Holloway vs. Oliveira, current market data, and a step-by-step setup guide.

    9 minMar 21, 2026
    Strategy

    Geopolitical Prediction Markets: How Traders Are Betting on Wars, Crises, and Global Events in 2026

    How geopolitical prediction markets work, which platforms offer them (Kalshi, Polymarket, ForecastEx), how to read live data, and how to trade Iran, Hormuz, and global crisis markets in 2026.

    9 minMar 15, 2026