Kelly sizing tries to maximize long-run bankroll growth when you have an edge. In a prediction market, that edge comes from the difference between the market-implied probability and your estimate of the event actually occurring.

• Full Kelly is the theoretical maximum-growth fraction.
• Fractional Kelly reduces the damage from bad estimates and variance.
• A hard cap protects the book from one oversized idea.

Start with your best fair probability estimate and current contract price, then decide what fraction of full Kelly you trust. If your edge estimate is model-driven but still noisy, half-Kelly or quarter-Kelly is often more defensible than full Kelly.

1. Confirm the trade is positive EV first.
2. Choose a confidence fraction based on estimate quality.
3. Set a hard cap that matches your overall risk policy.
4. Check whether existing positions already create hidden overlap.

Imagine a contract trades at $0.41, you estimate the win probability at 55%, your strategy bankroll is $5,000, and you prefer half-Kelly with an 8% cap. The calculator first computes the theoretical full Kelly fraction and then applies your softer risk settings.

• A good edge can still result in a small stake if the bankroll is small.
• A large theoretical Kelly number does not mean you should ignore concentration risk.
• Recommended contracts help convert the fraction into executable size.

The most dangerous Kelly mistake is feeding it an inflated probability estimate. If the edge is wrong, the size will be wrong too, and Kelly magnifies that error instead of protecting you from it.

• Using full Kelly when your model quality does not justify it
• Ignoring correlation with other open positions
• Treating bankroll as total wealth instead of strategy capital
• Skipping hard caps because the formula feels authoritative

Professional default: When in doubt, reduce the confidence fraction. Smaller size keeps you alive long enough to discover whether your edge estimate is genuinely repeatable.