August 18, 2026 · 5 min
Backtesting without lying to yourself: four traps that inflate your results
A perfect backtest is almost always a wrong backtest. Here are the four biases that turn an average strategy into a dream machine.
You tested your strategy on six months of charts: 68% win rate, 2-to-1 reward. Live, it loses. Your strategy didn't change. Your backtest was lying.
Trap 1: knowing what comes next
Scrolling through a chart, you see the next candle before deciding. Even unintentionally, you "skip" the setups that failed right after. The only cure: a candle-by-candle replay where the future stays hidden until you decide.
Trap 2: adjusting rules along the way
"I wouldn't have taken that one, there was a wick." Every exception added during the test improves the result and makes the strategy impossible to follow live. Write your rules before testing, and don't touch them until the end.
Trap 3: forgetting costs
Spread, commissions, slippage. On a scalping strategy with 5-pip stops, a one-and-a-half-pip spread eats a huge share of every win. A backtest without costs tests a strategy that doesn't exist.
Trap 4: too few trades
Thirty trades prove nothing. A lucky streak is enough to produce 70% wins on a small sample. Aim for at least a hundred trades, across several market conditions: trend, range, high volatility.
And the trap that remains live
Even an honest backtest only tests your method. It doesn't test your ability to follow it after three losses in a row. That's why the gap between backtest and live is measured in your journal.
TradeDiscipline Premium includes a backtest with candle-by-candle replay on 19 instruments and 4 years of history, using your own rules. Then your journal shows whether you stick to them live. See the plans.
Turn theory into practice
Free · No credit card · Ready in 2 minutes