How the engine thinks

Product questions

Who built this?

PropSurvival is built and maintained independently by a systematic trader — one person, no prop firm money, no broker affiliations. That independence is the product's spine: a firm-sponsored calculator can't tell you your expected cost is 2.7 attempts, because that's marketing against itself. The full methodology is on this page and the rule dataset is versioned on the public changelog — if the math is wrong somewhere, say so, and it gets fixed with credit.

What exactly do I get for free?

The full Monte Carlo strategy engine, every firm rule preset, CSV import with confidence diagnostics, the GO/REDUCE/NO-GO verdict, the Survival Score, the safe risk band, expected attempts and expected fee spend for your selected firm, a prescriptive action plan for your dominant failure cause, your single best-fit firm from the Firm Fit ranking, and a shareable verdict card (PNG and Discord-ready text). Free means free — no signup, no trial clock.

What does Pro ($39, one-time) add?

The complete Firm Fit ranking across all firms — funded probability (both phases composed for 2-step programs), expected attempts, expected fee spend, and modeled first-year economics — plus its annex in the institutional PDF report, Deep simulation mode (report-grade 10,000-path prop verdicts and 2,000-path risk-band scans), firm-rule updates for the life of the product, and use on 2 devices. A single up-front payment, not a subscription; nothing to cancel; 14-day refund.

How does license activation work?

After checkout, Lemon Squeezy emails you a license key. In the engine, choose Upgrade → Already purchased? Enter your license key, paste it, and activate. The key is verified directly against Lemon Squeezy's license API and the result is stored in your browser. Once activated, Pro keeps working offline indefinitely — the engine only revalidates in the background when it can reach the internet, and a check it cannot complete never downgrades you. The single thing that turns Pro off is an explicit refund or revocation, which by definition requires you to be online. Two devices per license; if you replace a machine, contact support and we'll move your activation.

Does my data really never leave my device?

Correct — and it's testable, not a promise. Load the engine, switch off your connection, and everything except license activation still works: CSV import, all simulations, charts, reports. There is no application server in the architecture at all.

What happens to my access if you ever stop maintaining it?

You keep the tool you paid for. Because the engine is a self-contained page that runs on your device — not access to a server we could switch off — the version you have keeps working even if updates stop. Your Pro license is built for this too: once activated it isn't downgraded just because a validation server is unreachable, so a company or payment-provider outage can't lock you out of what you bought. And if we ever permanently shut the product down, our intention is to publish a final build with the license check removed entirely, so nothing depends on a server that no longer exists. What a discontinuation ends is the release of new firm-rule verifications and features, logged on the changelog — not your access to the simulations, reports, or Pro features already on your machine. We don't guarantee the product will be offered or updated forever, and we've written our Terms to say that honestly rather than promise something we might not be able to keep.

Methodology

The core simulation

The engine performs path-level Monte Carlo simulation. For each of typically 600–10,000 paths (you choose the depth), it draws a sequence of trades from your statistics — win rate, average win (R), average loss (R), trading cost (R), trades per day (Poisson-distributed around your average) — and compounds equity at your chosen risk-per-trade. Each path is checked against the challenge's rules trade by trade, day by day, and terminates as pass, daily-drawdown fail, total-drawdown fail, consistency fail, or timeout. The aggregate frequencies are the probabilities you see.

Firm rule modeling

Mechanic	How it's modeled
Static drawdown	Fixed equity floor: initial balance minus the drawdown allowance. Never moves.
Trailing — end of day (EOD)	The floor ratchets upward with your highest end-of-day equity. Intraday spikes don't move it.
Trailing — intraday	The floor ratchets with your highest intraday equity peak, including unrealized gains at trade granularity — the harshest variant (Apex-style).
Freeze at initial	Once the trailing floor climbs to the initial balance, it locks and stops rising (Apex/Topstep behavior).
Daily loss limit	Percentage measured from each day's starting equity. Firms without a daily limit are modeled with the limit off.
Consistency rule	At target-hit, if your best single day exceeds the allowed percentage of total profit, the pass is blocked and the path keeps trading to dilute the best day; paths that never satisfy it inside the window count as consistency failures.
Time windows	Minimum trading days enforced before a pass; "unlimited" evaluation periods are modeled as 60 trading days.

Each firm preset carries a last verified date and modeling notes that disclose every simplification. Firms change rules without notice; their documentation is the final authority.

Survival Score

A 0–100 composite for one challenge configuration: 0.62 × pass probability + 0.28 × (1 − drawdown-failure probability) + 0.10 × risk-band discipline, where risk-band discipline measures how far your risk-per-trade sits above the computed safe band. It exists to compress "can I survive this" into one comparable number; the components are always shown alongside it.

Firm Fit economics (Pro)

For each firm, the Firm Fit ranking computes:

1. Funded probability — full rule-aware simulation at your stats and account size. For 2-step programs (FTMO, FundedNext, Funding Pips, The5ers), Phase 2 is simulated with its own target and day window under the same drawdown rules, and the funded probability is Phase 1 × Phase 2 — so 2-step FX firms and 1-step futures firms compare honestly.
2. Expected attempts — 1 ÷ funded probability (capped for near-zero probabilities). A Phase-2 failure is modeled as a full paid restart — the conservative reading of most firms' terms.
3. Expected fee spend — attempts × the firm's launch-window list fee.
4. Modeled first funded year — your per-trade expectancy × risk × account size × trades/day × ~220 trading days × the firm's profit split, then discounted 40% as an explicit haircut for funded-stage rules, payout gates, and real-world friction.
5. Modeled Net · Yr 1 — funded probability × modeled funded year, minus expected fee spend. Shown in dollars because a ratio on a $99 denominator produces absurd-looking percentages.

Known limitations, stated plainly

• Independence assumption. Trades are sampled i.i.d. from your statistics. Real returns show autocorrelation, regime shifts, and volatility clustering; treat outputs as a baseline, not gospel.
• Two-point outcome distribution. Wins and losses use your average R values (with costs), not your full R distribution. Fat-tailed strategies (rare huge winners) are approximated conservatively.
• Sample-size sensitivity. A win rate from 50 trades has wide error bars; the engine surfaces confidence labels rather than hiding them.
• No psychology model. Revenge trading, hesitation after losses, and rule-breaking are outside the model — and are, empirically, a major failure cause the simulation cannot see.

Buying & support

Refund policy

14 days, no questions asked. Reply to your Lemon Squeezy purchase email or write to support@propsurvival.com.

Will the price change?

$39 is the launch price for the first 50 licenses; after that it moves to $49. The commitment is public because honest urgency beats a fake countdown timer. Early buyers keep their price and continue to receive product updates at no extra charge for the life of the product.

I found a rule that changed. What do I do?

Email support with a link to the firm's documentation. Rule corrections ship as preset updates and are logged, dated, on the public rule changelog — that's the maintained-dataset part of the product, and exactly what the verification stamps are for.