Decision review
- What was the original thesis?
- What did the agent see?
- Did the trade match the written rules?
- Was the invalidation respected?
Use this review after a paper trade closes to turn the result into better rules, better sizing, or a decision to keep collecting evidence.
Choose one output: keep collecting paper data, reduce risk, tighten the rule, retire the rule, or write a new test. Avoid changing multiple things from one small sample.
A post-trade review should turn a closed paper trade into a decision about process. The review should not simply label the trade good or bad. It should explain whether the original thesis was followed, whether the risk was respected, whether the exit matched the plan, and whether the next sample needs a rule change.
Start with the pre-trade record. If the thesis, invalidation, and size were written clearly, the review has a stable reference point. If they were missing, that is the first finding. Trading Boy keeps the template connected to the pre-trade checklist, paper trading journal template, and the AI trading journal so a trader can compare intent with behavior instead of rewriting the story after the result.
Next, classify the paper trade. A clean loss may be acceptable if the setup was valid and the invalidation was respected. A profitable paper trade may still be a process problem if the agent ignored position size, chased late, or failed to record a clear exit reason. This distinction is important because paper trading is most useful when it improves the workflow, not when it creates a shallow win-loss diary.
The final step is to choose one next action. Keep collecting evidence, reduce size, narrow the setup, improve the prompt, add a skip rule, or retire the test. The best reviews are specific enough that the next group of paper trades can confirm whether the change helped.
Original plan: The paper agent was supposed to enter only after a confirmed reclaim and use a fixed invalidation level. The pre-trade checklist shows that the setup was documented before the simulated entry.
Outcome: The trade closed at a paper profit, but the journal shows the agent entered before confirmation and relied on later momentum. The result was positive, but the process did not match the rule.
Review decision: The trader marks the trade as a rule-fit failure, not a success. The next action is to tighten the confirmation instruction and run another sample before changing anything else.
A good review produces a specific tag and one next action: keep testing, reduce size, tighten entry, change exit logic, or retire the setup.
A weak review says the agent was right or wrong without explaining which rule was followed, broken, missing, or too vague to inspect.
Mistake tags make post-trade review searchable, but they only help if the tags describe behavior precisely. Tags like bad trade or poor market do not explain what should change. Tags like early entry, missing invalidation, oversized paper position, correlated exposure, or unclear exit reason give the next review something concrete to inspect.
Keep the tag list short enough that it remains usable. If every paper trade receives a unique label, patterns will stay hidden. If every trade receives the same broad label, the review will not produce specific action. A practical template uses a few recurring tags and adds a plain-language note when the situation needs more context.
After several trades, sort by tag before changing rules. Repeated early entries may require a confirmation filter. Repeated oversized positions may require a hard sizing rule. Repeated unclear exits may require a better exit checklist. That pattern-first approach keeps the review from overreacting to one result.
When the review has no clear tag, mark the trade as unclassified instead of forcing a conclusion. That keeps the record honest and tells you the pre-trade plan may need clearer fields before the next sample.
Include the original thesis, entry context, exit reason, rule fit, paper result, risk behavior, mistake tag, and next action.
Change rules when the review finds repeated behavior or a clear process flaw. Avoid changing several variables because of one isolated trade.
The journal records what happened. The review interprets what happened and decides what the workflow should do next.