The five numbers
Tally from the journal: total R for the week; number of trades; rule-adherence percentage, the share of trades graded fully compliant; R from compliant trades versus R from violations, the split that locates the real problem; and best-session concentration, where the profitable trades clustered. Five numbers, five minutes, and the week's shape is on paper instead of in feelings.
The three questions
One: did losses come from variance or violations, the diagnosis that decides whether anything needs fixing at all, per the streak protocol. Two: which single behaviour cost the most this week, named specifically, chased a lull-hour breakout, widened a stop, not traded badly. Three: what did the market reward this week, and is my playbook positioned for more of it? Fifteen minutes, honest sentences.
The one change
Choose exactly one process adjustment for next week, written as a rule: no entries in the midday lull; alerts only at H4 levels; size recalculated Sunday. One change is measurable; five are noise. Log it, and next week's review begins by grading it. That closing loop, change, measure, keep or revert, is the entire engine of improvement, and it fits in the template's last five minutes.
FAQ
When should the weekly review happen?
Any consistent slot after Friday's close and before Sunday's reopen, while the week is fresh and the next one is plannable.
What if the week had very few trades?
Review it anyway: adherence, declined setups and the market-reward question work at any sample size. Quiet weeks often teach selectivity best.
How long before reviews show results?
Behavioural metrics, adherence, violation costs, improve within weeks; expectancy follows as the corrections accumulate.
ⓘ See these ideas on real price: open the free XAUUSD live chart.