How can I compare two trading periods without changing the question?
By Chart Reporter Reviewed 2 min read
The short answer
Use equally sized calendar windows in the same timezone and preserve account, currency and filter scope. Compare counts and distributions as well as headline metrics. A better-looking second period does not by itself establish that a process change caused the difference.
Define the comparison before viewing results
Write down the question: perhaps whether a documented review process was followed more consistently, or whether a strategy's net outcome distribution changed. Choose the periods and filters before searching for a favourable answer. Retain the definition with the result.
Calendar windows use the selected timezone, so matching dates in Local and Broker modes can represent different instants. Keep the same mode on both sides. If an explicit named session is selected, Chart Reporter's previous-period comparison is disabled until that session is cleared.
Inspect access and eligibility on both sides
History access can restrict one window more than the other. All history is also not a finite window with an equally sized preceding period. Select a finite range and read any incomplete-sample or access notices before using the comparison.
The same issue applies to specialised analytics. If one period lacks initial stops or captured candles, its eligible R or excursion sample can be smaller. Compare the coverage and counts, not just the values. Avoid treating a missing result as a zero and thereby making the complete period look artificially better.
Describe change without overstating causality
Different trade counts, instruments, position sizes or market conditions can explain differences between periods. Money-based outcomes also need compatible currencies. Inspect concentration so a single large result does not stand in for a broad change.
State what changed in the observed sample and what other inputs changed at the same time. A process intervention can be a useful hypothesis even when the data cannot isolate its effect. Keep the next check specific and repeatable. That produces a more useful review than repeatedly shifting the window until the chart supports the conclusion you wanted.
Sources and further reading
Related guides
What should a repeatable weekly trade review include?
Build a review around a complete sample, record checks, one question and a documented follow-up.
How can I review a losing trade without rewriting the original decision?
Separate the information available at entry from later outcomes when writing trade-review notes.