The PropRank Survivability Score: Our Formula, In Full

The PropRank Survivability Score is a 0-100, rules-based estimate of one thing: a competent trader's odds of getting from purchase to first payout on a firm's 50K product. Version 1.0 weighs five components: drawdown mechanics (30 points), cushion ratio (20), payout accessibility (20), rule complexity (15), and cost to recover (15). Every weight, threshold, and deduction is published below. As of August 18, 2026, scores run from Lucid Trading's 90 to Elite Trader Funding's 56.

Competitors publish rankings and refuse to publish the math. We think that is backwards. The methodology is the product, so here is all of it.

What does the Survivability Score measure?

It measures a competent trader's odds of surviving from the day they buy an evaluation to the day a first payout clears, using nothing but the firm's published rules. Not the odds of passing. Passing is the easier half: Topstep's own published 2025 numbers show 51.8% of Trading Combine participants advanced to the funded level, but only 33.3% of funded traders ever received a payout. The finish line that matters is the first check, so that is what the score models.

Every score is computed on the firm's 50K product, because it is the one tier all nine tracked firms sell in nearly identical shape as of August 18, 2026: a $3,000 profit target against a $2,000 maximum drawdown. Which is also the framing this whole site is built on: a $50K account with a $2,000 trailing drawdown is a $2,000 account with $50,000 of buying power. All nine 50K products scored here are $2,000 accounts, 4.0% of the advertised number. The real account size piece walks through the arithmetic.

One plain-words fact the score takes as given: these are simulated accounts. Payouts on sim-funded accounts are paid by the firm out of its own revenue, not by market counterparties. The score prices the rules of that arrangement; it does not pretend the arrangement is something else.

What the score deliberately excludes

Profit-split economics, trust signals, and track record are left out on purpose; they live in the verdict layer, not the score. A 90/10 split is worth exactly nothing to the majority of buyers who never reach payout one, so splits do not belong in a survival metric. Trustpilot ratings, years in operation, payout evidence, and sim-to-live paths matter, which is why they are handled where judgment belongs: in our written verdicts and the overall rankings. That separation is how Topstep can be our best overall firm at 86 while Lucid Trading holds the top score at 90. The score stays mechanical. The verdict layer does the weighing.

The full v1.0 formula

Five components sum to 100 points, and the thresholds below are the entire rulebook. There is no hidden adjustment and no sponsor thumb on the scale.

A. Drawdown mechanics: 30 points

The drawdown mechanic is the single largest determinant of survival, so it carries the most weight. Base points by type: EOD trailing that locks, 30; intraday trailing that locks, 18; intraday trailing that never locks, 8; static, 25 (a hard floor never chases your equity up, but it never locks in safety above the start either; no tracked flagship 50K is static today). Each stage is scored separately, then blended: funded stage weighted 60%, evaluation stage 40%, because the account deaths that cost real money happen after funding.

The logic: an EOD-calculated trail cannot be blown by an intraday spike you rode and gave back, and a lock ends the trailing threat permanently. Intraday trailing that counts unrealized gains is the account-killer; trailing drawdown explained covers why. Worked blend: My Funded Futures runs EOD trailing in the eval and intraday-with-lock in the funded stage, so A = 0.4(30) + 0.6(18) ≈ 23. One edge case: Take Profit Trader's funded PRO trail is intraday including unrealized gains, and whether it locks at the starting balance is unconfirmed, so its funded stage scored 10, between the locks tier (18) and the never-locks tier (8), giving A = 0.4(30) + 0.6(10) = 18.

B. Cushion ratio: 20 points

Cushion ratio is maximum drawdown divided by profit target: how much room the rules give you per dollar you must earn. Bands: 0.67 or higher scores 20; 0.5 or higher, 14; 0.4 or higher, 9; below that, 4. At the 50K reference tier every tracked firm currently sits at $2,000 ÷ $3,000 = 0.67, so B is flat across this table. It stays in the formula because it will not stay flat: thinner-cushion products at other tiers, and any future 50K that shrinks the drawdown, get caught here automatically.

C. Payout accessibility: 20 points

C scores how fast and how freely the first dollars actually leave. Base by the fastest published first-payout path: day one or 24 hours, 20 points; five days or fewer, 17; five to eight days, 14; eight to fourteen days, 10; longer than fourteen, 6. When the path demands a per-day profit bar of $200 or more, it scores one band slower, because those days take longer to bank: that is why Alpha's five $200 days and Apex's five $250 days sit at 14, not 17. Then the deductions: withdrawal buffer, minus 2; per-payout caps, minus 2; hard lifetime or payout-count caps, minus 4; slow processing, minus 1. The speed bands come from what firms publish, cross-checked in fastest payout prop firms.

D. Rule complexity: 15 points

Every firm starts at 15 and loses points for each rule that can trip a profitable trader. Soft eval-only consistency rule: minus 3. Tight soft percentage (30% or lower): minus 4. Funded-stage or payout-time consistency: minus 6, rising to minus 7 when the threshold is 35% or tighter. Other tripwires (forced flat times, MAE rules, minimum risk-reward rules): minus 2 each. The gradient is deliberate. An eval-only soft rule costs you a slower pass; a funded-stage or payout-time rule costs you money you already earned.

E. Cost to recover: 15 points

E prices what it costs to be wrong once and go again: effective entry price, reset fee, and activation fee. This is the least mechanical component and we say so plainly: there is no threshold table, just a banded judgment on those three inputs, scored on effective prices rather than list, because standing discount codes are the real price at most of these firms. PropRank may earn a commission if you use a code on this page. That never changes a score.

Every firm's component subscores, in full

Here are all 45 subscores, sorted by total. Ties at 86 are listed in our pillar-ranking order.

And the math behind each row, reproduced from our scoring database:

Worked example: how Apex Trader Funding scores 72/100

Apex is the cleanest demonstration that one weak component can sink an otherwise strong rulebook, so here is its score end to end.

A: 30/30. Apex's 50K runs an EOD trailing drawdown in both stages (the eval trail is calculated at 4:59:59pm ET), and the funded-stage trail locks at $50,100. That is the best drawdown architecture the model recognizes.

B: 20/20. $2,000 ÷ $3,000 = 0.67. Top band.

C: 5/20. This is where the score collapses. The first payout requires five qualifying days of $250+ profit, and the $250 bar scores it one band slower: base 14. Then every deduction in the book applies. Minus 2 for the Safety Net buffer ($52,600 minimum balance to request, with the $52,100 net held for the life of the PA). Minus 2 for per-payout schedule caps. Minus 4 for the hard cap of six payouts per account, after which the account closes. Minus 1 for 5-11 day processing. 14 minus 9 = 5. No other tracked firm takes all four deductions.

D: 5/15. Minus 6 for the 50% consistency rule checked at payout time (it blocks the payout, not the account), minus 2 for the MAE rule, minus 2 for the 5:1 risk-reward rule. Apex's March 2026 overhaul removed MAE and risk-reward language from its marketing, but denials citing supposedly removed rules were still being reported in 2026, so v1.0 keeps both deductions until the removal is unambiguous.

E: 12/15. Failure is cheap. Effective entry is $55 with the standing 90% code against the $550 one-time list price, there is no reset fee because an expired eval is simply repurchased, and funded-stage activation is roughly $99 (a third-party figure we could not verify directly).

Total: 30 + 20 + 5 + 5 + 12 = 72. Read it correctly: the trading rules are excellent and the getting-paid rules are the worst of the majors. Apex remains the right buy for cheap, discounted, multiple accounts. It is just a purchase you make knowing payout six is the last one.

What the score cannot tell you

The score reads published rulebooks and nothing else, which creates four blind spots we would rather name than hide.

Published rules only. Execution quality, fill behavior, data feeds, and platform risk are unscored. Alpha Futures losing NinjaTrader, Tradovate, and TradingView on July 12, 2026 is exactly the kind of event the formula cannot see; it appears in the firm's verdict, not its score.

No payout-denial behavior beyond published policy. Tradeify's "0% payout denials" and Topstep's 99.26% approval rate are self-reported and unaudited. The score neither rewards nor punishes such claims, and a firm that denies payouts in ways its rulebook does not disclose will score better than it deserves until the rulebook catches up.

Self-reported claims are flagged, not scored. Where a number is unverified or conflicting we flag it in the firm dossier, the way Lucid Trading's homepage still advertising a 4.8 Trustpilot against the real 4.5 (checked directly August 18, 2026) is flagged. Flags inform verdicts; they do not move the score.

It assumes a competent trader. The score models the hostility of the rules, not your edge. A 90 does not make an unprofitable trader survivable, and a 56 will not stop a disciplined one who genuinely understands what they signed.

Versioning: v1.0, dated, publicly logged

This formula is v1.0, dated August 18, 2026, and every change to it will be logged publicly. A reweighted component, a new deduction, or a re-banded threshold gets a version bump, a date, and a plain-language note on what moved and why. Firm scores can also change without the formula changing, whenever a firm edits its rulebook; those re-scores carry their own as-of dates. What we will not do is silently edit history. If Lucid's 90 ever becomes an 84, you will be able to see exactly when and exactly why.

The blunt recommendation

Use the score to shortlist and the verdict layer to choose. The score says Lucid Trading (90) sells the friendliest rulebook of the nine: EOD locks everywhere, $0 activation, $90 resets. The verdicts add what the score refuses to weigh: Lucid launched in 2025, while Topstep at 86 has fourteen years of operating history, published pass rates, and a genuine sim-to-live path, which is why Topstep is still our best overall firm. Below 80, buy only if you can name the deduction you are accepting. At Apex (72) it is the six-payout account closure; at Elite Trader Funding (56) it is intraday eval trailing plus the $25,000 lifetime sim cap. And whatever you buy, price it honestly: every 50K product on this page is a $2,000 account. Trade it like one.