Setting a Fair Backtest Window

Objective

Learn to avoid test windows that are too short, too convenient, or too different from the market you plan to trade.

Workflow

• Start with the symbol, timeframe, and date range that match the intended bot configuration.
• Include at least one calm period and one volatile period when history allows it.
• Keep the baseline profile unchanged before testing variants.
• Record the reason for the selected window so the test can be repeated later.

Operator Checklist

• The backtest has enough trades to avoid reading too much into a tiny sample.
• The timeframe matches the planned execution rhythm.
• The market regime is noted in the result summary.
• Fees, spread assumptions, and slippage assumptions are understood.

Common Mistakes To Avoid

• Choosing the best-looking historical window after seeing the results.
• Comparing one-hour settings against five-minute settings without noting the difference.
• Ignoring periods where the strategy would have struggled.

Practice

Select two windows for the same symbol: one trending and one choppy. Explain what each window is meant to test.