Advanced

The Kelly Criterion for Prediction Markets: Bet Size That Maximizes Long-Term Gains

The Kelly Criterion for Prediction Markets: Bet Size That Maximizes Long-Term Gains

The Kelly Criterion for Prediction Markets: Bet Size That Maximizes Long-Term Gains

A trader spots a Polymarket event trading at 50 cents that he's certain has a 60% chance of happening. He's got $10,000 in his account and smells easy money. So he bets it all — and loses everything on a single unlucky outcome. This happens more than you'd think. Having an edge isn't enough. The Kelly Criterion tells you exactly how much to bet when you find a mispriced prediction market — the mathematical sweet spot between growing your bankroll and avoiding ruin.

What Is the Kelly Criterion?

The Kelly Criterion is a formula that calculates the optimal bet size when you have an advantage. John Kelly developed it at Bell Labs in 1956, but traders have been using it ever since to answer one crucial question: how much of your bankroll should you risk on any single bet? The magic of Kelly is that it maximizes your long-term growth rate while ensuring you never go broke. It's not about maximizing your next bet's payout — it's about making sure you're still trading a year from now.
Kelly Criterion Formula: Bet Size = (Edge) ÷ (Odds - 1)
Where Edge = (Your probability × payout) - 1
The formula automatically scales your bet size based on how confident you are and how mispriced the market is. Big edge, bigger bet. Small edge, smaller bet. No edge, no bet.

Kelly Formula for Prediction Markets

In prediction markets, the Kelly formula becomes beautifully simple. If you think an event has a higher probability than the market price suggests, Kelly tells you what percentage of your bankroll to bet. Here's the practical version: Bet Size = (Your Probability - Market Price) ÷ (1 - Market Price) Let's break this down. Your probability is what you think the real chance is. Market price is what the prediction market is currently trading at. The denominator accounts for the payout structure. The key insight? Kelly automatically reduces your bet size as the market gets closer to your estimated probability. If you think something has a 55% chance but it's trading at 54 cents, Kelly tells you to bet tiny. If it's trading at 30 cents, Kelly tells you to bet much larger.

Real Polymarket Example

You're browsing Polymarket and see "Will it rain in New York tomorrow?" trading at 40 cents. You check the weather forecast and decide there's actually a 60% chance of rain.
Ready to apply Kelly sizing to your prediction market trades? EdgedUp's probability models help you identify mispriced markets across Polymarket and Kalshi, giving you the edge estimates you need for optimal Kelly betting.
Using the Kelly formula: - Your probability: 60% (0.6) - Market price: 40 cents (0.4) - Kelly bet size: (0.6 - 0.4) ÷ (1 - 0.4) = 0.2 ÷ 0.6 = 33.3% Kelly says bet 33% of your bankroll. If you have $1,000, that's a $333 bet. Here's what happens if you're right: you win $333 × 1.5 = $500 profit. If you're wrong, you lose $333 but still have $667 left to trade another day. Kelly ensures that even a string of bad luck won't wipe you out.

Why Most Traders Use Fractional Kelly

Full Kelly is mathematically optimal but practically aggressive. Most experienced prediction market traders use "fractional Kelly" — betting only 25% to 50% of the Kelly recommendation. The reason? Real life is messier than math. You might be wrong about probabilities more often than you think. Market liquidity might prevent you from getting your exact price. Fractional Kelly trades some growth potential for much better downside protection. Professional trader wisdom: start with quarter-Kelly (25% of the Kelly recommendation) until you've proven your edge over dozens of trades. If our rain example suggested a 33% Kelly bet, quarter-Kelly would be about 8% of your bankroll. The biggest Kelly mistake isn't underbetting — it's overbetting. Bet twice the Kelly amount and your risk of ruin skyrockets, even with a genuine edge. I've seen traders blow up accounts with 70% win rates because they bet too big on each trade.

Kelly's Hidden Superpower

The most counterintuitive thing about Kelly? It automatically tells you when not to bet. If the market price is higher than your estimated probability, Kelly gives you a negative number. That's the formula's way of saying "short this market" or "don't bet at all." Kelly also scales with uncertainty. The more confident you are in your probability estimate, the more Kelly tells you to bet. Less confident? Kelly naturally reduces your position size. This is why Kelly works so well for prediction markets. Markets are constantly mispricing events based on emotions, biases, or incomplete information. Kelly gives you a systematic way to capitalize on those mispricings without gambling your future on any single outcome. The Kelly Criterion doesn't just help you bet smarter — it helps you survive long enough to compound your edge over hundreds of trades.

Find your edge → Try EdgedUp

Find your edge → Try EdgedUp →
← Back to Blog Get Early Access →