Risk of Ruin Calculator

A positive edge tells you the average bet is profitable. It tells you nothing about the order the results arrive in — and the order is what decides whether you are still betting when the edge finally shows up. This simulates a thousand runs of a staking plan and reports how they ended.

What the simulation reports#

Each run plays out the same number of bets at your chosen strike rate and price, staking a fixed percentage of the current bankroll each time. Four numbers summarise the thousand runs:

• Fell to 20% — the share of runs that dropped to a fifth of the starting bankroll. This is the ruin figure.
• Worst fall — the median run's deepest peak-to-trough drop. Half of all runs fell further than this at some point.
• Finished ahead — the share that ended above where they started.
• Median finish — the middle run's final bankroll, as a multiple of the starting one.

The simulation is seeded, so the same inputs always give the same answer. A calculator that reshuffles on every keystroke is an invitation to press it until you get the number you wanted.

Why "ruin" is a threshold, not zero#

Staking a percentage of your current bankroll means the stake shrinks as the bankroll does. Lose half and you are betting half as much; lose 90% and you are betting a tenth. Mathematically, you can approach zero forever and never arrive.

That makes literal bankruptcy the wrong thing to measure. What actually ends a betting plan is the point at which it stops being worth continuing — where the stakes are trivial, the recovery would take years, and you quit. The default here puts that line at a fifth of the starting bankroll. Move the number in your head if you like; the shape of the results does not depend on where it sits.

Stake size, not selection, decides survival#

This is the point the tool exists to make. Hold everything else fixed — a 53% strike rate at 2.00, which is a +6% edge per bet (0.53 × 2.00 − 1) — over 500 bets, and change only how much of the bankroll goes on each one:

Read the last two columns together. Going from 1% to 5% roughly doubles the median finish, which looks like a good deal. Going from 5% to 20% turns a 2.40× median into a 0.19× one while ruin goes from one run in thirty to nine in ten. Past a certain point, betting more per bet does not buy more money — it buys a worse distribution of the same money.

Push the stake slider in the calculator and watch it happen. Ruin and drawdown rise steeply and immediately; the median finish rises slowly, peaks, and then collapses.

Where the turning point sits#

It is not a coincidence that the median finish tops out near 5–6%. For these inputs, the Kelly criterion says the growth-optimal stake is exactly 6% of the bankroll — the edge divided by the net odds, (0.53 − 0.47) ÷ 1. Above that, you are staking more than the edge can support and the compounding works against you.

Almost nobody should bet full Kelly anyway, because it assumes your probability estimate is exactly right. It never is. A half or a third of the Kelly stake keeps most of the growth and cuts the drawdowns sharply, which is why conventional staking plans land around 1–2% flat. Work out the number for your own inputs with the Kelly calculator.

What the model assumes#

Be clear about the ways this is a simplification, all of which flatter the results:

• Every bet is independent. Real portfolios are correlated — same league, same day, same weather, same reasoning — and correlated bets lose together.
• The odds and the strike rate never change. Real prices drift, and the edge you think you have is an estimate with error bars around it.
• You stake to plan every time. Chasing, tilting and skipping bets are not modelled, and they only make it worse.
• The edge is assumed, not proven. Set the strike rate to a number your results actually support, not one you hope for. Establishing that a 53% strike rate is real rather than luck takes far more bets than most people ever place — see sample size.

That last one deserves emphasis. Tofiko has not demonstrated an edge over closing prices, and says so publicly. If you enter a winning strike rate here, you are asking a conditional question — "given this edge, what would survival look like?" — not measuring one you have.

Set the edge to zero and the picture is bleaker still: the median run loses money and a bigger stake only loses it faster. No staking scheme repairs a price that is against you, which is the whole argument against progression systems like Martingale. Everything on this page is about surviving long enough for an edge to matter, and the bankroll management guide covers how to set the starting number in the first place.