Omniwer

Kelly Criterion

A mathematical model for determining the portion of capital to allocate to a bet based on the estimated win probability and payout odds.

f* = (bp − q) / b

How It Works

01 / PROBABILITY

Outcome Estimate

The model uses the estimated probability of winning and compares it with the payout odds.

02 / EDGE

Finding the Edge

If the calculated expected value is positive, the formula determines the portion of the bankroll to allocate to the bet.

03 / SIZE

Position Size

The resulting fraction is multiplied by the current bankroll, after which the bet can be reduced using the selected Kelly fraction.

Core Idea

Full Kelly uses the bankroll fraction calculated by the formula. Half Kelly and Quarter Kelly use only a portion of that fraction, reducing the size of each position.

Important Limitation

The Kelly Criterion depends on the accuracy of the estimated probability. An inaccurate probability estimate can result in an excessively large bet size.

Model Parameters

Set the bankroll, outcome probability, odds, and the fraction of the calculated Kelly allocation to use.

Handling Negative EV
Under the classic Kelly Criterion, a negative EV results in no bet.

Calculation Result

The current expected value and recommended position size.

Expected Value Status
Positive Expected Value
Edge
+2.00%
Full Kelly
2.00%
Calculated bankroll fraction
Effective Fraction
1.00%
After applying the selected Kelly fraction
Next Bet
$10.00
After rounding to the selected step
Formula
f* = (b × p − q) / b
where b is the net profit per unit bet, p is the probability of winning, and q is the probability of losing.

Interactive Trainer

Record outcomes and observe how the bankroll and next position size change.

Current Bankroll
$1000.00
Next Bet
$10.00
The bet size is recalculated after each change in bankroll. Therefore, the next position may change after a win or loss even when the model parameters remain unchanged.

Strategy Simulation

Simulate a series of random outcomes using the current Kelly Criterion parameters.

Kelly Criterion Simulator (Monte Carlo Simulation)

Analysis of 30 independent simulation sessions with 100 bets

$1.3k$1.0k$0#0#25#50#75#100
Liquidation frequency (30 sessions)0.0% (0 of 30)
Median final capital$1013.00
Peak capital (Max Peak)$1163.00
ℹ️

What This Section Shows

This section shows how the estimated probability and odds are translated into a position size. The Kelly Criterion does not predict the outcome of the next event; it provides a mathematical approach to managing bet size under the specified assumptions.