The ratio is the difficulty knob
An evaluation’s profit target and its max drawdown are usually sold as two separate numbers. For the probability of passing they are one number: their ratio. Hold the trader, the risk-per-trade, the max drawdown, and the horizon fixed, and raise only the target — the model’s probability of passing falls in a smooth, measurable curve.
Under an intraday trailing floor, pass probability ran from 61.9% (±0.34) at ratio 0.5 (target 3% on a 6% drawdown) down to 14.6% at ratio 2.5 — a span of 47.3 points across the same trader.
The common industry design of an 8% target on a 6% drawdown is ratio 1.33. At that design, with risk fixed at 1.0% of balance, the trailing archetype passed on 34.2% of attempts and the static archetype on 44.4% — a floor-mechanic gap of 10.2 points.
| Risk/trade | 1.0% |
| Max DD | 6% |
| Window | 20 days |
| Daily cap | none |
| Edge | +0.05R |
Only the profit target moves. The x-axis is target ÷ max DD.
| Win rate | 50% |
| Avg win | 1.2R |
| Avg loss | 1.0R |
| Cost/trade | 0.05R |
| Trades/day | 4 |
Same modest edge used in the sizing study, so the two pages are comparable.
Pass probability against the target-to-drawdown ratio for one illustrative trader. Each point is 20,000 seeded paths. Wine marks the common 8% target / 6% drawdown design (ratio 1.33).
Synthetic archetype, not any named firm. Reproduce from the linked dataset with the published seed.
The inconvenient half of “easy” rules
A ratio of 0.5 — target equal to half the max drawdown — is the “easy” end of this sweep. It is not free. Under the trailing floor, 38.1% of attempts still failed, and 37.9% of all paths ended against the drawdown rather than the clock (timeout share 0.3%).
Raising the ratio from 1.0 to 2.0 cost 23.4 pass-probability points under trailing — more than the entire static-vs-trailing floor gap at the common 8/6 design (10.2 points). The ratio knob is larger than the floor-mechanic label.
Under the trailing floor, drawdown-share exceeded timeout-share in every cell; under the static floor it did too, except at the hardest ratio (2.5), where the clock overtook the floor (timeout 42.4% vs drawdown 37.6%). Outside that corner the floor, not the clock, is the binding constraint — so “more days” would not rescue a high-ratio design the way a lower target would.
Ratio 0.5 still failed 38.1% of attempts under a modest edge.
1.0 → 2.0 ratio cost 23.4 pts; 8/6 floor gap 10.2 pts.
Failure mix against the ratio (trailing archetype). Wine is the drawdown-breach share; graphite is timeout. The floor stays dominant at every ratio in this sweep.
Synthetic archetype, not any named firm. Reproduce from the linked dataset with the published seed.
Static and trailing move together — the gap is smaller than the span
Swapping only the floor mechanic (static vs intraday trailing) shifts the whole curve, but it does not erase the ratio effect. At the easiest ratio the two floors are close; at the common 8/6 design the trailing floor costs about 10.2 points relative to static; at the hardest ratio both floors sit in the mid-teens.
The trailing span is 47.3 points from easiest to hardest ratio; the static span is 48.8 points. Both floors tell the same story about difficulty: raise the target relative to the drawdown and the model’s pass probability falls, whether or not the floor trails.
| Easy 0.5 | T 61.9% / S 68.8% |
| 8/6 1.33 | T 34.2% / S 44.4% |
| Hard 2.5 | T 14.6% / S 20% |
Pass probability at three ratios under static (graphite) and trailing (wine) floors. Same trader, same risk, same max drawdown.
Synthetic archetype, not any named firm. Reproduce from the linked dataset with the published seed.
What this is not
Model-derived and illustrative. The archetype is synthetic and does not represent any specific proprietary-trading firm. A real evaluation may add rules (a daily loss limit, a consistency rule, higher minimum trading days) that only tighten these results. Your own edge, cost, risk-per-trade and trade frequency change every number; this study shows the SHAPE of the target-to-drawdown relationship, not a design to copy.
No figure here is a pass rate for a named firm. The subject is a synthetic archetype chosen to isolate the ratio. Real rule sets add daily loss limits, consistency caps, and payout rules that only tighten these results. Position size is held fixed at 1.0% so this page does not re-answer how much to risk per trade — that study sweeps size; this one sweeps design difficulty.
Questions the numbers answer
What is the target-to-drawdown ratio?
Profit target (%) divided by max drawdown (%). An 8% target on a 6% drawdown is ratio 1.33. The same max drawdown with a 3% target is ratio 0.5; with a 15% target it is ratio 2.5.
Does a lower ratio always mean an easier evaluation?
In this model, yes for the shape: pass probability falls as the ratio rises. It does not mean a low-ratio design is free — at ratio 0.5, 38.1% of attempts still failed under a modest edge.
Is the floor type (static vs trailing) more important than the ratio?
Not in this sweep. Moving from ratio 1.0 to 2.0 cost 23.4 points under trailing; the static-vs-trailing gap at the common 8/6 design was 10.2 points.
Can these numbers be reproduced?
Yes. Seed base 20260816, 20,000 paths per cell, generator tools/build-target-to-drawdown-ratio.js. The CC BY 4.0 dataset is the machine-readable source of every figure on this page.
Is this advice on which firm to buy?
No. PropSurvival models published rules; it does not rank firms, predict markets, or tell anyone which evaluation to purchase.
What this is. A model-derived result with disclosed assumptions — not an empirical fact about the real world, and not a claim about any named firm.
Engine. src/lib/state-conditional-mc.js simulateFromState — the same Monte Carlo kernel the PropSurvival app runs — driven through src/lib/case-file-sections.js safeRisk. Each of 10 target levels is an independent estimate of 20,000 simulated attempts, seeded from base 20260816 plus a fixed per-level stride.
Isolation. Risk-per-trade, trader profile, max drawdown (%), horizon, and daily-loss (disabled) are held constant. Only the profit target (%) moves; the x-axis is target/maxDD.
Trade model. Each day: Poisson(mean 4) trades. Each trade wins with probability 0.5 for +1.2R or loses for −1R, minus 0.05R cost. Equity compounds; one R = the risk-per-trade percentage of current balance.
Validation. The engine passes its directional invariants (tests/engine.invariants.test.js: a trailing floor is never easier than a static one, regimes differ by a measurable margin, results are seed-deterministic and bounded) and reconciles against hand-derived worked examples (tests/rules-reconciliation.test.js).
Cite. PropSurvival. (2026). Target-to-drawdown ratio: how the evaluation difficulty knob changes pass probability [Dataset]. https://propsurvival.com/articles/target-to-drawdown-ratio
Reproduce. Dataset (CC BY 4.0): target-to-drawdown-ratio-data.json. Command: node tools/build-target-to-drawdown-ratio.js.
- Dataset (CC BY 4.0): target-to-drawdown-ratio-data.json — accessed 2026-08-16.
- Related: How much to risk per trade · Static vs trailing · Drawdown landscape census · Research methodology.
- Run your own numbers: the free path — 61.9%→14.6% · easy fails 38.1%.
Model-derived with disclosed assumptions. Synthetic subject.
20260816 + per-level stride; 20,000 paths/cell; xoshiro128** (PSBoot rngFor).