TL;DR: Many failed prop firm challenges fail for predictable reasons: one bad day of correlated losses, a losing position carried past the daily reset, too much leverage, or a pass rate that was in-sample all along. In our replays of Propr's challenges, the 3% daily limit ended 36% of starts on the fastest Classic 1-Step plan, against 4% for total drawdown. Below are twelve mistakes, the evidence, and the fix for each.
Which mistakes happen before you buy the challenge?
1. Picking the challenge type without doing the math
Challenges differ in more than price. With no edge, the chance of reaching the target before the drawdown floor is about drawdown ÷ (target + drawdown), a gambler's ruin result. On Propr's rules, that is about 37.5% for the Classic 1-Step (10% target, 6% drawdown), 25% for the Turbo 1-Step (9%, 3%), about 29.4% for the Pro 1-Step (12%, 5%) and, for the Classic 2-Step (10%, 8% trailing), about 32% (the simple formula's 44% ignores the trailing floor).
Fix: compare cost per pass, not fee alone. As planning estimates, a $10k Classic 1-Step costs about $390 per pass, and a Turbo 1-Step played for a pass within 1–2 days costs about $200–220 per pass ($50 fee at roughly 23–25% odds). See Classic, Turbo, Pro or 2-Step.
2. Trusting one backtest
A single backtest from one start date shows one path. A challenge can start on any day, and a strategy that passes from January 1 may fail from January 2.
Fix: replay the challenge from every start date and look at the share that pass and how the rest fail.
3. Believing in-sample pass rates
Our house plans show 58–100% in backtests. Those numbers are in-sample: the plans were chosen with the 2025 data visible, and overlapping daily starts share most of their path, so 798 starts are not 798 independent tries. When our team picked candidates on 2020–2023 starts and scored them on 2025–26, picks that looked like 40–80% in-sample plateaued at about 25–34% (median around 30%). A published community bundle claiming about 94% passed 0–44% under the real rules and never passed a 2025 start within a year.
Fix: plan on out-of-sample numbers. Our realistic planning figure for the slowest, smallest-size tier is 40–55%. Read about overfitting, selection bias and out-of-sample testing.
4. Skipping the forward test
A backtest cannot show you execution bugs, missed signals or costs it did not model. Paying a challenge fee to find them is expensive.
Fix: run the same strategies on a paper agent first. On Trigr, paper agents fill at live Hyperliquid prices with trading and builder fees, though without slippage or funding.
Which mistakes break the loss limits?
5. Ignoring where the daily loss is measured from
Many traders assume the daily limit is measured from their initial balance. On Propr it is measured from the balance the day began with, the 00:00 UTC realised-balance snapshot. After a losing week, the dollar allowance is smaller than you think.
Fix: recompute the day's floor in dollars every day: opening realised balance × 0.97 on a 3% limit.
6. Holding a loser through the daily reset
The snapshot is realised balance, but breaches are judged on equity. A position down $200 at midnight on a $10,000 account sets a floor of $9,700 while equity is already $9,800: the new day has $100 of room, not $300.
Fix: close or cut losing positions before 00:00 UTC, or count any open loss at midnight as already spent. The daily loss limit guide has the full example.
7. Stacking correlated positions
Three altcoin positions at 2x, each with a third of a $10,000 account and a 2% stop, lose about $434 together (4.3%) if they all stop out on the same day, including costs. No single trade looks dangerous; the three together breach a 3% limit.
Fix: assume correlated crypto positions all lose together on bad days, and size the sum, not each slot.
8. Using maximum leverage
At 10x, a round trip of about 17 bps in costs uses about 1.7% of the account per trade. Two full-size round trips cost more than a 3% daily limit before the market moves. Our best results came at 1–3x.
Fix: keep leverage low and check the venue caps. Propr allows 10x on BTC, ETH and SOL and 2x on other crypto. The position sizing guide shows the cost table.
9. Ignoring fees
Every round trip is a small loss you pay whatever the outcome, so fees push you below the zero-edge ceiling. A high-frequency strategy with a thin edge per trade can have negative expectancy after costs.
Fix: backtest with fees and slippage. Our research for these numbers applied trading fees, 3 bps of slippage per side and funding. See trading fees in perp backtests.
10. Misreading trailing drawdown
On Propr's Classic 2-Step, the floor is the high-water mark minus 8% of the starting balance ($800 on $10,000), and it stops rising at the starting balance. It is not 8% of the peak. And while it trails, profits you give back cost room: a trader up 1% after a 4% rally can have less room than on day one.
Fix: track the floor in dollars after every new high. Our trailing vs static drawdown guide walks through a $10,000 example.
Which mistakes happen during and after the attempt?
11. Chasing after a loss
The classic manual failure: a bad day, then bigger size to win it back, then a breach. Raising size after losses moves you closer to the floor at exactly the moment room is smallest.
Fix: fix size before the challenge and do not change it after a loss. Automating the strategy removes the temptation, though not the need for sensible size.
12. Passing without a plan for the funded account
The funded account keeps the same loss limits, with no target. A strategy with no edge that passed by luck keeps taking the same odds, and can breach the funded account just as easily.
Fix: decide before you pass what size you will run funded, what drawdown makes you stop, and what evidence would make you switch strategies.
What do these mistakes have in common?
Almost all of them come from treating a challenge like a normal account. A normal account forgives a bad day; a challenge does not. The fix in every case is the same idea: design for the worst moment, measured exactly the way the firm measures it, and judge your odds on data the strategy was not chosen on.
How Trigr fits in
Trigr is a strategy builder and backtester whose agents trade on Hyperliquid and on Propr. Several of the fixes above are built into the Propr flow. When you deploy your own agent on Propr (house plans show their measured record instead), a rules check replays the combined backtest with a fresh challenge on every UTC day against the account's real daily-loss and drawdown rules, and Optimize sizing can shrink positions until it fits with a 0.5% buffer. Discover › Propr shows house plans per challenge with their failure split and a live record, which started on 2026-09-30 and, as of 2026-10-05, has six challenges running per plan with none breached or finished. Setup is in the trading agents docs. If you decide to take a challenge, you can open one through our Propr referral link. No tool removes the need for an edge.