advanced · TRADING LEARNING GUIDE
Validation and journaling
What is this concept?
Validation tests frozen rules on unseen data with realistic execution. A journal records what was known, what action was taken and why. Live signals and historical hit rates are different outputs.
Step by step
Version rules; log timestamp, source, entry assumptions, exits, missed fills and costs. Separate training from evaluation and define ambiguous intrabar outcomes in advance.
From observation to decision
Record the instrument, venue, session and timeframe. Classify the observation using the definition, then write at least one alternative interpretation. Decide what evidence would invalidate your idea before entry. Mark missing information explicitly. Examples are hypothetical. Collect paper observations, including failures. Include fees, spread, slippage and timestamp alignment when measuring outcomes. Recognizing a chart shape and establishing a profitable trading rule are separate tasks.
Worked example
Ten signals: three TP1, two stops, five expired. Counting expiry as no TP1 gives 3/10=30%; using only five filled signals gives 3/5=60%. These measure different populations.
Common mistakes and limits
Avoid hiding expiry, selecting the best backtest as proof or omitting repainting and revised data.
Practice
Complete a journal for paper trades. On each rule change, start a new version and keep earlier results separate.
Resources and further reading
These sources explain definitions and underlying mechanics. Worked examples are original illustrations; a source does not establish the profitability of a custom pattern.
Further reading / foundational context. Worked examples on this page are original illustrations, not source performance claims.