Kelly Criterion Explained: How to Stake Like a Quant
The Kelly Criterion tells you the mathematically optimal bet size for any edge. Here is how to apply it without blowing your bankroll.
Kelly Criterion Explained: How to Stake Like a Quant
Most bettors size stakes by feel. The Kelly Criterion replaces feel with math — and over a long enough horizon, math wins.
What Kelly actually solves
Given a bet with a known edge, what fraction of your bankroll maximizes long-term growth? Bet too little and you under-compound. Bet too much and one cold streak wipes you out. Kelly finds the optimum.
The formula
For a binary bet with decimal odds:
f* = (bp - q) / b
Where:
f*= fraction of bankroll to stakeb= decimal odds minus 1 (the profit per unit staked)p= your estimated probability of winningq= 1 - p
A worked example
You estimate Manchester City has a 60% chance of beating a mid-table side. The bookmaker offers 2.00 (50% implied).
b = 1.0p = 0.60q = 0.40f* = (1.0 × 0.60 - 0.40) / 1.0 = 0.20
Full Kelly tells you to stake 20% of your bankroll. On a £10,000 bankroll, that is £2,000. On one bet.
Why nobody bets full Kelly
Full Kelly assumes your edge estimate is perfectly accurate, you can rebalance continuously, and you have infinite emotional tolerance. None of these hold in real betting.
The standard professional adjustment is fractional Kelly:
| Fraction | Stake on the example above | Risk profile |
|---|---|---|
| 1.00x (full) | £2,000 | Reckless |
| 0.50x (half) | £1,000 | Aggressive |
| 0.25x (quarter) | £500 | Standard professional |
| 0.10x (tenth) | £200 | Conservative |
Quarter Kelly is the industry default. It captures ~75% of the long-term growth of full Kelly with a fraction of the drawdown.
Common Kelly mistakes
- Re-staking after a loss. Kelly uses your current bankroll, not the starting figure.
- Compounding too quickly. Update bankroll daily, not after every bet.
- Ignoring correlated bets. Three accumulator legs on the same match are not independent — Kelly underestimates the risk dramatically.
- Using bookmaker odds as your edge. Your edge requires your own probability estimate.
Kelly in MoBet
The MoBet betslip surfaces recommended Kelly stakes once you record at least 20 settled picks — enough data to estimate your historical edge per market. Conservative bettors should stick to ¼ Kelly.
The takeaway
Kelly is not a magic formula — it is the correct answer to the question "how much should I bet?" Pair it with a calibrated edge estimate and a fractional adjustment for real-world uncertainty, and you will outlast 99% of punters who size by gut.
0 comments
Sign in or create a free account to join the discussion. Comments are open to registered users only.
- Be the first to comment.