> The probability of passing a prop-firm evaluation before breaching it, from your own edge and the rules you type — across the range your sample supports. - Canonical: https://hadalinstruments.com/tools/prop-evaluation-survivor/ --- Tool # Prop evaluation survivor An evaluation has two exits: the profit target, and the rules that end the account first. Given your own win rate, payoff and stake, and the rules typed from your own rulebook, this simulates the account trade by trade and reports how often it passes, how often it breaches, and **which rule did the breaching**. And because a win rate is an estimate from a finite sample, it reports the odds across the range your sample supports rather than as one number. Pass probability — Pass probability, at your stated win rate — Range your sample supports — The same evaluation, across what your sample can and cannot rule out Win rate | Which reading | Expectancy / trade | Passes | Breaches | Expires | — | Pessimistic end of your sample | — | — | — | — | — | What you typed | — | — | — | — | — | Optimistic end of your sample | — | — | — | — | Which rule bound, across the breached paths at your stated rate — Consecutive losses from the start that breach the floor — Median trades to a decision, pass or breach — ## Why the rules are inputs, not assumptions Rulebooks differ between firms and between programmes at the same firm: the anchor a floor is measured from, whether it trails and whether it stops trailing, what a daily limit is assessed against and when it resets. Nothing about any firm is measured or assumed on this page. You transcribe your own rules into the fields above — the procedure for reading them is [what causes a prop firm drawdown breach](https://hadalinstruments.com/research/what-causes-a-prop-firm-drawdown-breach/) — and the page computes on those numbers and nothing else. Where this page and your rulebook disagree, your rulebook is the fact. The note's central observation is that an evaluation carries several rules that can end it, each moving on its own clock, and at any moment exactly one is nearest: the **binding barrier**. A breach is what happens when the trader is watching a different one. So this page does not only count breaches; on every breached path it records which rule bound — the floor or the daily limit, whichever was nearer before the trade that crossed it — and reports the split. If the daily limit binds far more often than the floor, the allowance you were watching was not the rule that was going to end you. ## What the simulation does Each simulated trade is a win with your stated probability and a loss otherwise. A loss costs the staked fraction; a win earns that fraction times your payoff ratio. After every trade the path is tested against the three exits in turn: the balance at or above the target passes; the balance at or below the floor breaches; the day's loss at or beyond the daily limit breaches. A static floor sits at the starting balance minus the allowance and stays there. A trailing floor follows the highest closed balance minus the allowance, so under it profit consumes room; with a lock at breakeven it stops rising at the starting balance. A path that reaches no exit inside the trades available — days times trades per day — expires, which is a real outcome and is counted rather than folded into either side. It is **deterministic**: the generator is seeded, so identical inputs always produce an identical answer. That is deliberate. A tool that returned a different figure on every keystroke would invite re-rolling until the answer felt acceptable, which is the habit this site exists to argue against. And where no simulated path passes, the panel says **under 1 in 4,000** rather than nought percent, because those are different statements and only the first was measured. ## What this does not do Trades here are closed results, so the trailing floor follows the highest *closed* balance. A rule whose floor follows equity including open positions ratchets on unrealised highs you never banked, which is harsher than anything modelled here — pass odds on this page are optimistic for such accounts, and the [architecture classifier](https://hadalinstruments.com/tools/trailing-vs-static-drawdown/) is where to establish which rule you actually have. Consistency rules, minimum trading days, scaling plans and the timezone at which a daily limit resets are not modelled; each can end or extend an evaluation that this arithmetic calls passed. The model assumes your trades are **independent and identically distributed**, and real trading is neither: losses cluster, and positions held at once in correlated instruments are one risk wearing several tickets, which pushes a real account toward its floor faster than these paths. Stops are instructions, not guarantees, so a real breach can land below the floor being defended. And it cannot tell you whether your win rate and payoff are honest; if they came from a backtest they are biased optimistic before anything else is wrong with them — the domain of [backtest overfitting](https://hadalinstruments.com/glossary/backtest-overfitting/). Every one of those effects moves the true pass probability below the figure here, so read it as a ceiling rather than an estimate. It publishes no pass rate for any firm, because none has been measured, and it does not tell you what to risk: what it gives you is the consequence of a stake against a rule set, across the range your own evidence supports. The one-exit version of this question — a run of losses against a single floor — is the [risk of ruin calculator](https://hadalinstruments.com/tools/risk-of-ruin-calculator/). The distance to a trailing floor from where an account stands today is the [trailing drawdown calculator](https://hadalinstruments.com/tools/trailing-drawdown-calculator/), and the rules themselves are defined at [trailing drawdown](https://hadalinstruments.com/glossary/trailing-drawdown/), [daily loss limit](https://hadalinstruments.com/glossary/daily-loss-limit/) and [hard breach](https://hadalinstruments.com/glossary/hard-breach/). Whether the win rate you typed is evidence yet is the [sample-size check](https://hadalinstruments.com/tools/sample-size-check/). The live version of this arithmetic — distance to every rule restated on every fill, with correlated positions folded into one exposure — is what the [Prop-Evaluee Risk Guardian](https://hadalinstruments.com/instruments/prop-evaluee-risk-guardian/) is being built to do. All the calculators are listed under [calculators](https://hadalinstruments.com/tools/). And when the inputs cannot support an answer, the panel above refuses and says why rather than rendering a zero — [why an instrument refuses to answer](https://hadalinstruments.com/docs/insufficient-n/) is that design, written down.