How to test a trading strategy before you pay for an eval

Write the strategy as rules someone else could follow, test it on enough trades to mean something (dozens, ideally more), include commissions and slippage, and then run it under the exact drawdown and daily loss limit of the account you plan to buy. A strategy that makes money but breaks the eval's rules still fails the eval.

1. Write it down as rules

If two traders read your rules and take different trades, they aren't rules yet. Each needs:

  • Market and time: which contract, which session hours.
  • Entry: the exact condition, not "when it looks strong".
  • Invalidation: where the idea is wrong, which is your stop.
  • Exit: target, trailing stop or time-based exit.
  • Size: fixed risk in dollars per trade.
Example trading strategy written as five rules: market and time, entry, invalidation, exit and size, for an opening range breakout on MES
Five rules, no judgment calls. If another trader can't take the same trades from this card, it isn't finished.

2. Get enough trades

Ten trades tell you almost nothing. With a few dozen you start to see the shape; with a hundred or more you can start trusting the win rate and expectancy. Results from one week of a trending market won't hold in a range.

Chart: the measured win rate of a 50 percent strategy can land anywhere from 19 to 81 percent after 10 trades and narrows to 40 to 60 percent after 100 trades
Same strategy, same edge: after 10 trades your measured win rate could be anything from 19% to 81%. After 100 it settles to 40% to 60%.

3. Include the costs

Subtract commissions on every round turn and add a tick of slippage to stop fills. A strategy that makes two ticks per trade on MES can disappear once fees and slippage come out.

Per-trade example on MES: $2.50 gross profit minus about $1.00 commission and $1.25 of slippage leaves $0.25
Example on MES, where one tick is $1.25: a 2-tick edge turns into 25 cents a trade once costs come out.

4. Run it under eval rules

This is the step backtests usually skip. A strategy can be profitable and still fail an eval, because:

  • its worst day is bigger than the daily loss limit,
  • its losing streaks are deeper than the trailing drawdown,
  • its best day is too large a share of the profit for a consistency rule.

Take your trade list and play it through the account rules day by day. The simulator does this for Apex and Topstep accounts: log the trades and see whether the account survives.

Chart of a 50K eval: the strategy finishes 12 days up $3,400, but on day 7 the close of $50,350 is under the trailing floor of $50,400, so the account fails
The dashed line is what the backtest reports. The eval ended on day 7, when a $2,050 pullback over three days closed under the trailing floor.

5. Change one thing at a time

If you tweak the stop, the target and the entry at once and the result improves, you don't know why. Worse, you may have fitted the rules to this sample. Change one rule, test it on new trades, and keep a note of what you changed.

Test log comparing version 1 and version 2 of a strategy where only the stop changed, with trades tested and the worst day for each
Keep a log like this: one change per version, tested on new trades, with the worst day next to the daily loss limit you'll trade under.

What this won't tell you

No test tells you what the next hundred trades will do. It tells you whether the idea held up on the trades you had, under the rules you'll face. That's the bar to clear before paying, not a promise beyond it.

Blow it here, not on a paid eval

Run the same rules on your own trades in the simulator. Free, no card, and a blown account costs you nothing but pride.

Start practicing