Unit 3 · Level 3 · Building Your System
Backtesting: interrogate the past
Backtesting means replaying history and checking: 'if I had traded my rules exactly, what would have happened?' Scroll a chart back, walk forward candle by candle, log every signal your rules produce, wins AND losses. Fifty historical reps cost nothing and reveal what a year of live guessing can't.
Free to play. No ads, no token, no account needed to start.
What you get asked
The one requirement that makes a backtest honest?
No adjusting them per trade to 'catch' winners. If the rules bend during the test, you're not testing the strategy anymore. You're fitting a story to the past. Rigid rules, honest count.
Match the backtesting sin to its damage
A strategy tuned until the past looks perfect usually shatters on contact with the future. Rough rules that survive many conditions beat perfect rules that survived one.
A backtest that only covers a bull market tells you nothing about the strategy in a ___.
Test across regimes: trend, chop, crash. A support-bounce play behaves VERY differently in each. Better to learn that in the archive than in your account.
Order the honest manual backtest
Archive → paper → small live. Each gate filters; your account only meets strategies that passed twice.
Your backtest shows +0.4R expectancy over 60 signals across two regimes. What did you actually gain?
Backtests don't promise the future. They buy you CONVICTION with evidence. Conviction is what survives drawdowns. Untested traders quit at loss #4; tested ones check the plan and continue. 🐜
The rest of this unit
Playbooks, backtesting, the paper-to-live bridge, and the art of not trading.