Sometimes the best strategy is the one nobody thinks will work. This week, our backtests threw us a curveball: two nearly identical "longshot YES" strategies that both crushed it by playing the odds on ultra-cheap contracts.
We're talking about contracts trading between 5–15¢—the kind of bets where you're looking for something unlikely to happen, and you're getting paid handsomely if it does. The kicker? Both strategies had a 17% win rate and turned $100 into $327–$339 in simulated profit. That's not a typo.
This is where probability gets fun. At 5–15¢, these contracts are pricing in roughly 5–15% implied probability. If the market is even slightly miscalibrated, and the true probability is 17–18%, you're getting +EV bets. Miss 83% of the time, but when you hit, you double or triple your money. The math works.
Over 300 trades, that edge compounds. You don't need to be right often—you just need to be right more often than the market thinks, and you need enough scale to smooth out variance.
These strategies are nearly twins. The only meaningful difference: closing timeframe.
That $12 difference might sound tiny, but it hints at something: shorter-duration bets in this price range might be hitting the sweet spot between "enough time for movement" and "less time for unexpected news to blow up your thesis." Plus, faster closing means faster recycling of capital—you can take another 17% bet next hour.
The zero drawdown on both is also notable. No single bad streak forced you underwater. Consistent small winners bleeding into periodic home runs.
The backtest ran across seven categories: weather, sports, crypto, financials, politics, economics, and culture. That breadth suggests this wasn't a one-category phenomenon. Markets were generating enough mispriced longshots across the board to feed both strategies.
Translation: There was genuine opportunity in the chaos, not just luck in one pocket of the market.
This is simulated, not real money. These backtests are run against actual Kalshi order book data from last week, but they're not accounting for slippage, fills at worse prices, or the psychological weight of watching a 5¢ position swing 40% in minutes. Real trading is messier. Your actual P&L will be different.
Also: past performance doesn't predict future returns. Just because longshots paid off this week doesn't mean they will next week.
So here's the question we're sitting with: Was this a one-week anomaly, or did we stumble onto a structural quirk in how Kalshi prices low-probability events? Keep an eye out next week.
Want to test these strategies yourself? Backtest and analyze any of them in chat — 250,000 free credits, no credit card required.
Try Prediction PilotEvery week we test strategies against real Kalshi data and share the profitable ones.