Understanding Expected Value in Prediction Markets

Expected value (EV) is the single most important concept in prediction market trading. This guide explains what it is, how to calculate it, and how to find +EV opportunities systematically.

What is expected value?

Expected value is the average outcome you'd get if you made the same bet thousands of times. It's calculated as:

EV = (Probability of Winning × Payout) − (Probability of Losing × Cost)

If EV is positive (+EV), the bet is profitable over time. If negative (−EV), you lose money in the long run — regardless of any individual outcome.

EV in prediction markets

Prediction markets like Kalshi and Polymarket price contracts between 0¢ and 100¢, where the price represents the market's implied probability. If you believe the true probability differs from the market price, there's an edge.

Example: A contract trades at 40¢ on Kalshi (implying 40% probability). Your model says the true probability is 55%. The EV calculation:

• EV = (0.55 × $0.60) − (0.45 × $0.40) = $0.33 − $0.18 = +$0.15 per contract

That's +37.5% expected return. Over many such bets, this edge compounds.

Cross-platform edges

The easiest +EV opportunities come from price discrepancies between platforms. When the same event trades at different prices on Kalshi, Polymarket, and sportsbooks, at least one price must be wrong.

EVSignals scans 500+ sources continuously to find these gaps. The scanner flags opportunities where the cross-platform spread exceeds a configurable threshold.

How to get started

1. Use the scanner to find markets with large cross-platform spreads 2. Open a notebook to analyze the underlying data and build conviction 3. Backtest your thesis against historical settlement data 4. Size positions according to Kelly criterion or your preferred risk model

Understanding EV is the foundation. The tools help you find and act on it efficiently.