Skip to content
BetSim

Betting-decision simulator

Simulate your decisions, not the match.

Betting Optimization Simulator puts thousands of virtual bettors through your staking rules, bet slips and habits in parallel worlds — and shows you the range of outcomes, the risk of ruin and the confidence intervals before you risk a cent.

18+ · Education and risk management — no tips, no guarantees.

Bankroll after 300 bets — 1,000 simulated bettors
  • 5–95% of bettors
  • 25–75% of bettors
  • Median bettor
  • €1,000

Illustrative example with synthetic data: a small edge and quarter-Kelly staking. Bands show where 50% and 90% of bettors finish; the line is the median.

How it works

Four steps from a vague idea ("I'll stake 5% on value bets") to measured risk.

  1. 1

    Describe the environment

    Odds range, bookmaker margin and how good your estimates really are — including the noise in them.

  2. 2

    Set your rules

    Value filter, staking method, maximum stake, ruin threshold. Add tilt if you want to be honest about behaviour.

  3. 3

    Run thousands of bettors

    Micro-agents follow your rules on identical random sequences, so strategies are compared fairly.

  4. 4

    Read the risk, not a prediction

    Median outcome, 5–95% bands, probability of profit and ruin with confidence intervals, drawdowns and losing streaks.

Four tools, one question: how should I decide?

Every result is backed by established mathematics, explained in plain language.

Game theory

We remove the bookmaker margin with the Shin model and treat staking as a game against an uncertain nature: the maximin (Nash) allocation and CVaR-robust Kelly protect you when your probabilities are wrong.

Dynamic programming

Bellman's equation finds the stake that is optimal at every bankroll level — for long-run growth, or for reaching a target before hitting a ruin line.

Markov chains

Behaviour has states. A calm/tilt chain models how losses push you into bigger, emotional stakes, and how long losing streaks really last.

Monte Carlo & micro-agents

Thousands of independent virtual bettors in parallel worlds with common random numbers give distributions and confidence intervals, not single numbers.

How the method works

What you get

Strategy simulator

Compare flat, proportional, Kelly variants, robust Kelly, DP-target and martingale on the same random worlds.

Bet-slip analyser

Enter up to 15 selections. See probability ranges, EV with confidence intervals, fair odds and which accumulators actually add value.

Training academy

Short lessons with interactive calculators and quizzes: margins, EV, Kelly, variance, accumulators and tilt.

Confidence intervals everywhere

Every probability comes with its uncertainty. No false precision, no single "magic" number.

Responsible-gaming tools

Self-exclusion, daily limits and a model that tells you plainly when not betting is the best decision.

Start free, upgrade when you need more

The Free plan includes daily simulations and the core academy. Pro raises the limits, unlocks advanced courses and larger agent populations.

Compare plans

Frequently asked questions

An honest disclaimer

Betting Optimization Simulator is an educational and analytical tool. It does not accept bets, is not a bookmaker and is not affiliated with any. Results are simulations based on the assumptions you enter; they are not forecasts and not financial advice. Betting can lead to the loss of money you cannot afford to lose. Only adults (18+) may use the service.

Will Betting Optimization Simulator tell me what to bet on?

No. Betting Optimization Simulator does not predict sporting results and gives no tips. It simulates the consequences of your own decisions — stake sizes, slips, habits — so you can see the risk they carry.

Can it guarantee profit?

No tool can. Bookmakers build a margin into every price, and most bettors lose over time. What we show is the realistic range of outcomes, including the probability of losing money and of ruin.

What is a micro-agent simulation?

We create thousands of virtual bettors who all follow the same rules but experience different random sequences of results. The spread of their final bankrolls shows what could happen to you.

Why confidence intervals?

A probability estimated from a simulation is itself uncertain. We report a range (for example 52%–61%) so you never treat a rough estimate as a fact.

Who can use the platform?

Adults aged 18 or over only. We provide self-exclusion and limits, and we encourage anyone who feels betting is becoming a problem to seek help.

Test your rules before your money does.

Create a free account and run your first simulation in under a minute.

Create free account
Betting Optimization Simulator — betting-decision simulator