METHODOLOGY · HOW WE RANK, AND WHY

How PropRank ranks prop firms

Prop firms are unregulated counterparties selling rulebooks, and the rulebook — not your skill alone — decides whether money ever comes back. This page is the whole model in one read: the odds you are really buying, the three layers we score, why the rules look the way they do, and how it all adds up. At the bottom, the model itself runs in your browser: set any rule and watch the score move.

01 · THE QUESTION

Out of 100 people who buy an evaluation, 7 ever see money.

About 14 in 100 evaluations pass; about 45% of funded traders are ever paid. . You are buying a rulebook from an unregulated counterparty, and the rulebook — not your skill alone — decides most of those odds. So that is what we score.

7 get paid · 7 pass but never get paid · 86 never pass

Every figure below runs one through one , using the same code that scores every published board.

02 · WHAT YOU ARE BUYING

ROI is the wrong question.

It only matters if you get through. On the friendliest version of that account — one phase, a floor that never moves, no gates — our trader looks fine:

Passes the evaluation86%60 trading days
Reaches a first payout98%five winning days
Trading days to cash70≈ 3 months
Expected net per ticket$283retries priced in
Chance of passing, by program structure
  • One phase, static floor86%
  • One phase, intraday trail54%
  • Two phases, intraday trail30%
  • Three phases, intraday trail16%

Same trader, same $2,500 drawdown — only the shape of the product changes.

03 · LAYER 1 — THE FIRM

A perfect rulebook is worthless at a firm that will not pay.

Eleven signals ask about the counterparty, not the trading: named leadership, jurisdiction, , slowed payouts, , and whether there is a at all. This is the only layer that multiplies.

Firm A

Named CEO, US-registered, itemised payout reports, narrow denial clause, a live path.

Layer 1
89
Multiplier
×0.93
Score
79.0

Firm B

Anonymous team, mailbox jurisdiction, nothing published, "at our sole discretion", simulated forever.

Layer 1
56
Multiplier
×0.74
Score
54.1

Identical rulebooks, 25 points apart, because Firm B keeps only 74% of what its rules earn.

04 · LAYER 2 — THE RULEBOOK

Every account has a kill floor. The only question is whether it moves.

A static floor never moves, so everything you earn is cushion. An . Same $2,500, same trader, same trades:

Chance of passing, by drawdown floor
  • Static — never moves86%
  • End-of-day trail83%
  • Balance-based trail77%
  • Intraday trail, locks at start67%
  • Intraday trail, never locks54%

A never-locking trail leaves 38% of the advertised drawdown usable.

Chance of passing, by daily loss limit
  • No daily limit86%
  • $500 — ends the day86%
  • $500 — fails the account0%
  • $1,000 — fails the account86%

A limit that fails the account is a second kill floor — and its size decides everything.

Conduct rules sit here too: a firm that forbids something your style needs — — is capped at 30 for you, however good the rest looks.

05 · LAYER 3 — THE PAYOUT

Passing is not getting paid.

A do not take your profit; they keep you exposed to the floor for longer, which is the same thing.

Trading days from passing to a first payout
  • No gates10 days
  • 50% best-day cap31 days
  • 30% best-day cap52 days
  • 20% best-day cap78 days
  • 5 winning days over $25032 days

The same profit, metered out in time.

Expected spend$107ticket plus expected resets
First payout, your share$45990% split
Expected net$283probability-weighted
Return on the ticket265%on the friendliest plan

Then : the split, the cap, the buffer, and whether a withdrawal lowers your floor.

06 · WHY THE RULES LOOK LIKE THIS

A firm with no live path only earns when you lose.

The rules across phase one, phase two, funded and payout look arbitrary. They are one rule — protect the prize pot — and how hard a firm must protect it comes down to . A firm that splits real wins earns when you do; a firm that only sells simulations earns when you fail.

07 · HOW IT ADDS UP

Twenty-four signals, three layers, one risk curve, hard caps.

Each signal scores 0–100. A weights them into a base score, layer 1 multiplies it, and an unresolved flag caps it. It is — and : compensation is never an input.

Layer 1 · The firm11 solvency & conduct signalsLayer 2 · The rulebook8 rule-family signalsLayer 3 · The money5 survival-model outputs× weightsBase scoreweighted mean, 0–100Published scoreper lens, per market× risk curve: 1 − strength × (1 − layer 1/100), flooredcaps: retroactive change 20 · critical flag 25 · major flag 45
The ember path is why a firm that rewrote the deal sinks on every board, whatever lens you pick.

08 · THE LIVE MODEL

Change the rules. Watch the score move.

This is the scoring module itself, running in your browser on a fictional firm — the same code that produces every published board. Set any rule, pick the trader, pick the lens.

Show me as a
Board lens
Layer 1 · The firm
Layer 2 · The rulebook
Layer 3 · The payout
PropRank · typical trader67.2base 77.5 × risk curve 0.87
  • The firm78
  • The rulebook84
  • The money69
Chance to pass
83%
Chance of a first payout
87%
Trading days to cash
91
Usable cushion
91%
Expected spend
$111
Expected net
$903
  • No path to live capital: the firm only earns when traders fail.
All 24 signals for this firm
  • 100No retroactive change on record.
  • 100No approved rule changes in the trailing year.
  • 70No regulator named; no action on record.
  • 100No distress flags.
  • 100Headquartered in the United States.
  • 100Named: A. Founder.
  • 555 year(s) operating.
  • 20Every stated plan is simulated; no live path.
  • 80Strongest tier: firm_published.
  • ?0 published review(s); needs 3.
  • 55Narrow, quotable denial language.
  • 8750K: end_of_day_trailing/locks_at_start_balance, 91% usable cushion → 86.8% reach first payout.
  • 6720 full-stop losses of room at $125/trade.
  • 100No daily loss limit.
  • 8450% best-day cap at payout.
  • 695 qualifying day(s) per payout, first payout after 0 days.
  • 92Split 90%, cadence 7d, buffer 0% of drawdown.
  • 62Allocation $300000, 5 accounts.
  • 100News allowed, overnight allowed, automation allowed.
  • 60$40 per $1k of drawdown at the cheapest plan.
  • 89$110.69 expected spend on 50K (1.2 expected attempts).
  • 7283.5% pass × 86.8% first payout on 50K.
  • 991.1 trading days to first payout on 50K.
  • 100ROI 816% (expected net $903.29 on $110.69) on 50K.

A fictional firm. Published scores come only from reviewed evidence and a human publish; this panel invents nothing about any real firm.

Six questions at sign-up — what you trade, how long you hold, how often, your win rate and reward-to-risk, what you risk per trade, and what you are optimising for — and every published firm is re-scored for the trader you described.

METHODOLOGY FAQ

Questions about the model

What does a PropRank score measure?

The likelihood a real trader actually gets real money out of a firm: whether the firm can and will pay (layer 1), how hard its rules work against the trader (layer 2), and what the path to a first payout costs in money, time and odds (layer 3). It is 0–100, deterministic, and computed only from reviewed evidence.

Why does an intraday trailing drawdown score lower than a static one?

Because it chases your best moment, including profit you never banked. For the same advertised $2,500 a marginal trader can use about half of it under an intraday trail and under 40% if the trail never locks, and the odds of passing fall with the usable cushion. That is not an opinion about the rule; it is the survival model's arithmetic, and the sandbox on this page lets you run it.

Why is a firm with cheap evaluations not automatically ranked higher?

Because price is only one of 24 signals, and the universal layer multiplies every score. A $29 ticket at a firm with no live path, an anonymous team, a mailbox jurisdiction and broad denial language scores about 55 on the universal layer, so it keeps roughly 64% of its composite on the Best-for-traders board and lands in the mid-40s despite a strong rulebook. Retries are where the money goes, and the model prices them.

Does an affiliate offer change a firm's score?

No. Compensation, sponsorship, and partner codes are excluded from every score and ranking input.

Are simulated funded accounts the same as live brokerage accounts?

No. PropRank labels simulated and live stages separately and does not describe simulated buying power as deposited trader capital.

Methodology changes require a new algorithm version (this page describes proprank.brain.v1 and extraction schema proprank.rules.v4). Historical boards retain their original version and as-of date so a later formula cannot rewrite an earlier ranking. The sandbox above scores a fictional firm in your browser and never touches a published score. Recommendations are research, not financial advice.