The evaluation phase is not a contest to see who makes the most money. It’s a test to see who can stay inside the rules long enough to hit a modest target. Those are different games, and treating the first like the second is how skilled traders fail challenges they should have passed.
Why the evaluation is a survival test, not a returns contest
A challenge asks for a specific profit target while forbidding you from ever breaching a daily loss cap or a drawdown floor. The target is usually reachable with a boring, steady grind. The traps are the two floors — and those are absorbing barriers. Hit the profit target and you pass; hit a floor and you’re out, no matter how far ahead you were an hour earlier.
That asymmetry should dominate every decision. A blown target just delays you; a blown floor ends you. So the correct objective isn’t “maximise expected return” — it’s “maximise the probability of touching the target before touching a floor.” Everything below follows from that reframing.
How variance, not expectancy, drives breach probability
Two traders with the identical, positive expectancy can have wildly different pass rates. The difference is variance — the size of the equity swings around that expected value.
- A high-expectancy edge with huge swings can still clip the drawdown floor on a bad run before the target arrives.
- A modest edge with small, smooth swings can grind to the target while never coming near a floor.
During evaluation, variance is the enemy even when expectancy is your friend. You want the smallest swings that still reach the goal in time. It’s worth confirming your edge is genuinely positive first — feed your win rate and average win/loss into an expectancy calculator, because low variance on a negative edge just means you lose slowly and surely. If the concept feels fuzzy, what expectancy actually measures is the right place to anchor before you tune anything else.
Trimming size to shrink the equity swing during the phase
Position size is the single biggest lever on variance, and it’s fully under your control. Cutting size doesn’t change your win rate or your setups — it just narrows the distribution of outcomes, which is precisely what you want when a wide tail can end the account.
Concretely, run the challenge at a smaller risk-per-trade than you’d use on a live account you owned outright. Yes, it takes more trades to reach the target. That’s the trade you want: more time on the clock in exchange for a much lower chance of a floor breach. Size each entry deliberately with a position-size calculator so your R is constant — inconsistent sizing is a hidden variance source that ambushes traders who “feel out” each position.
The counter-intuitive part
Smaller size can raise your expected profit over the challenge, because it raises the probability you’re still trading when the good streak arrives. Dead accounts have zero expectancy. Staying alive is the whole edge.
Trading only A+ setups to raise win rate and cut drawdown
Fewer, better trades reduce variance on two fronts at once: a higher win rate tightens the swing, and fewer entries mean fewer chances to stumble into the floor.
Define your A+ setup narrowly and in writing before the phase starts — the exact conditions, session, and confluence you’d stake your account on. Then trade only those. The B and C setups that pad your trade count on a personal account are pure variance you can’t afford here.
This is where honest self-measurement matters more than usual. Automatic journaling captures every trade without you deciding what to log, so you can see afterward whether you actually held the line on A+ setups or quietly let mediocre ones creep in. Shibiki’s live edge health scores each strategy with a Wilson confidence interval, which is the right tool for a small sample — it tells you not just your observed win rate but how sure you can be of it given how few trades you’ve taken. During a short evaluation, that honesty about sample size stops you from over-trusting a lucky start.
Lower R targets that still clear the profit goal in time
Big R-multiple targets look efficient — fewer winners needed — but they lower your win rate and let trades round-trip from profit back to loss, both of which spike variance. During evaluation, a lower, more consistent R target usually pays better in survival terms.
| Target style | Win rate | Give-back risk | Fits evaluation? |
|---|---|---|---|
| High R, let it run | Lower | Higher | Only if the clock is generous |
| Modest R, bank it | Higher | Lower | Usually the safer pass |
Book profits at a level you actually hit consistently, then compute how many such trades reach the goal within the challenge window. A steady sequence of banked, modest winners produces a smoother equity curve than a few home-run attempts that spend days underwater.
Estimating your pass probability before you start
Don’t walk in blind. Before the phase, model it: your realistic win rate, average R, chosen risk-per-trade, the profit target, and the two floors. Even a rough estimate tells you whether your current plan has a comfortable margin or is a coin flip. A prop-firm challenge calculator turns those inputs into an expected path and a sense of your odds, so you can adjust size and targets before real money and a real clock are on the line.
Confirm the exact target and floor figures with your firm — they vary by account and change over time — then tune size and R until the estimated pass probability is one you’d take repeatedly. If the honest answer is “coin flip,” the fix is lower variance, not more courage.
Related: Trading expectancy explained · Expectancy calculator · Prop-firm challenge calculator