The Kelly Criterion calculator estimates what percentage of a betting bankroll could be allocated when your assessed probability is higher than the break-even probability represented by the odds.
To use the calculator:
- Enter the bookmaker’s decimal odds.
- Enter your independently estimated win probability.
- Enter the bankroll reserved exclusively for betting.
- Choose full, half, quarter, or custom Kelly.
- Review the calculated edge, bankroll percentage and stake.
If the result is zero or negative, the entered values do not indicate a positive mathematical edge. The calculator should return a suggested stake of ₦0 rather than forcing a positive recommendation.
What Is the Kelly Criterion?
The Kelly Criterion is a mathematical staking model that connects three values:
- the available odds;
- your estimated probability of winning;
- your current bankroll.
It calculates a theoretical fraction of the bankroll based on the size of the estimated advantage. A larger calculated edge produces a higher Kelly percentage, while a small or negative edge produces a smaller or zero allocation.
The formula was developed for maximising long-term logarithmic growth under specific mathematical assumptions. It does not predict match results, determine whether your probability is accurate, or eliminate losing runs.
Kelly Criterion Formula
For decimal odds:
Kelly fraction = (b × p − q) ÷ b
Where:
- b = decimal odds minus 1;
- p = your estimated probability of winning;
- q = probability of losing, calculated as 1 − p.
The same formula can be written as:
Kelly fraction = (decimal odds × win probability − 1) ÷ (decimal odds − 1)
The theoretical stake is:
Kelly stake = bankroll × Kelly fraction
For fractional Kelly:
Adjusted stake = full-Kelly stake × selected fraction
If the result is negative, the practical calculator output should be zero.
Kelly Betting Example in Naira
Assume the following values:
- Decimal odds: 2.20
- Estimated win probability: 50%
- Betting bankroll: ₦100,000
Step 1: Calculate the Break-Even Probability
Break-even probability = 1 ÷ 2.20 = 45.45%
Your estimate is 50%, which is 4.55 percentage points above the break-even probability.
Step 2: Apply the Kelly Formula
b = 2.20 − 1 = 1.20
p = 0.50
q = 1 − 0.50 = 0.50
Kelly fraction = (1.20 × 0.50 − 0.50) ÷ 1.20
Kelly fraction = 0.0833, or 8.33%
Calculated Stakes
|
Strategy
|
Bankroll allocation
|
Calculated stake
|
|
Full Kelly
|
8.33%
|
₦8,333.33
|
|
Half Kelly
|
4.17%
|
₦4,166.67
|
|
Quarter Kelly
|
2.08%
|
₦2,083.33
|
These amounts are mathematical outputs, not personalised betting recommendations. If the 50% probability estimate is too optimistic, the calculated stakes will also be too high.
Full Kelly vs Fractional Kelly
Full Kelly
Full Kelly uses 100% of the calculated fraction. It assumes that the entered probability is accurate and that the mathematical conditions closely match reality.
Because football probabilities are estimates rather than known values, full Kelly can result in aggressive stakes and substantial bankroll volatility.
Half Kelly
Half Kelly uses 50% of the full calculation. In the example above, an 8.33% full-Kelly result becomes approximately 4.17%.
Quarter Kelly
Quarter Kelly uses 25% of the full result. It reduces exposure further and may be more suitable when there is considerable uncertainty in the probability estimate.
Fractional Kelly reduces the calculated stake, but it cannot turn an inaccurate probability into a reliable one.
Why Your Probability Estimate Matters?
The calculator does not create a win probability. It only processes the number you enter.
A user who enters 70% instead of a more defensible 52% may receive a dangerously large suggested stake. For this reason, probability estimates should be based on structured analysis rather than confidence, recent headlines or loyalty to a team.
Potential inputs include:
- recent home and away performance;
- opponent strength;
- expected line-ups and absences;
- rest and travel schedules;
- underlying attacking and defensive data;
- market conditions;
- model uncertainty;
- comparable historical fixtures.
Kelly Criterion and Implied Probability
Implied probability comes from the bookmaker’s price:
Implied probability = 1 ÷ decimal odds
Your estimated probability comes from your independent analysis.
The Kelly calculator compares the two. A positive Kelly result appears when your estimated probability is sufficiently higher than the break-even probability.
Use the implied probability calculator to examine an individual price. If you want to remove the bookmaker’s margin from a complete market, use the no-vig calculator.
When the Kelly Result Should Be Zero?
Suppose the odds are 2.00 and your estimated win probability is 48%.
The break-even probability at 2.00 is 50%. Your estimate is lower than the probability required to break even.
The formula produces a negative Kelly fraction, so the calculator should display:
- Positive edge: No
- Full-Kelly percentage: 0%
- Suggested stake: ₦0
A negative result is useful information. It means the entered price and probability do not support a theoretical allocation under the Kelly formula.
Common Kelly Calculator Mistakes
Using the Bookmaker’s Probability as Your Own Estimate
Entering the raw implied probability from the same odds normally leaves no independent advantage to calculate. Your probability should come from separate analysis.
Using an Unrealistic Bankroll
Your betting bankroll should contain only money intentionally set aside for betting. It should not include rent, food, savings, borrowed funds or money required for essential expenses.
Treating the Result as a Required Stake
The Kelly output is a theoretical fraction. You are not required to stake that amount, and personal risk limits should always take priority.
Ignoring Bookmaker Margin
Raw odds include a bookmaker margin. When using market prices as a reference for your probability, consider examining the complete market with the no-vig tool first.
Using Full Kelly With Uncertain Estimates
Small errors in probability estimation can have a large effect on the result. Fractional Kelly and a fixed maximum stake cap can reduce exposure.
Combining Unrelated Bets
The simple Kelly formula assumes one defined opportunity. Multiple correlated selections require more advanced portfolio modelling and should not be treated as independent bets.
Kelly Criterion Limitations
The standard calculation assumes that:
- the probability estimate is accurate;
- the quoted odds remain available;
- the bet can be placed at the calculated stake;
- the bankroll value is current;
- the outcome is correctly defined;
- there are no omitted fees or deductions;
- repeated opportunities behave consistently with the model.
Real sports betting does not satisfy these assumptions perfectly. Probabilities are uncertain, odds move and results can be correlated. The calculated stake should therefore be treated as an educational reference rather than an instruction.
NaijaScore9 should also allow users to apply:
- a fractional-Kelly setting;
- a maximum bankroll-percentage cap;
- a maximum stake amount;
- a zero-stake rule for negative results;
- clear warnings for unusually high allocations.
