Writing a strategy you can actually test

"Buy pullbacks in an uptrend" is an idea, not a strategy. A strategy is a set of rules specific enough that two people would take the same trades.

The template

1. Market and timeframe — what you trade and on which chart.
2. Context (long / short conditions) — the conditions that must be true before you look for entries. Example: daily structure is bullish (last break was a BOS up).
3. Entry rule — the exact trigger. Example: 15m change of character up inside a daily demand zone.
4. Stop-loss rule — where invalidation is. Example: 0.2 × ATR below the swing low that formed the trigger.
5. Take-profit rule — how you exit winners. Example: 50% at 2R, remainder at the next daily swing high.
6. Risk rule — size and limits. Example: 0.75% per trade, max 2 open positions, stop after −3R in a day.

Make every rule observable

If a rule includes words like strong, clean or obvious, replace them with something measurable.

Then measure it

Create the strategy in the Strategy Lab, link each journal trade to it, and review its expectancy, drawdown and R distribution as the sample grows. Change one rule at a time, and consider creating a new version of the strategy rather than editing the old one, so results stay comparable.

A note on backtesting

Historical testing is valuable, but it needs clean data and honest assumptions about costs and fills. Forward-journaling your rules is slower, but it measures exactly what you actually did.