Reading an autopsy report

Every report carries the same sections in the same order.

​

What’s in a report

One scrolling page, top to bottom: Three panels sit down the right-hand side alongside all of that: The rest of this page takes them in order, and says what each one measures.

​

Header and verdict

The top line is the ticker, the instrument, and the date, followed by the result and a grade. The grade is two letters, not one. Process grades the quality of the decisions, Outcome grades what the market paid you. Hover the grade for Tradion’s note on the gap between them.

​

What this actually means

Four combinations, and the two in the middle carry the information. When the two grades differ by two letters or more, the attribution section explains why.

​

Timeline reconstruction

Five to eight events between entry and exit, in order, each with a time and a label. Entry and exit get a filled dot; warnings and decision points sit between them. You already remember the entry and the exit. The middle is what this section is for: it names the moments where the trade stopped being the one you planned. The first event that was not part of your plan is usually earlier than the point where the loss showed up.

​

Risk metrics

Four tiles under the scores: Entry and Exit prices, Duration, and R-Multiple. R-Multiple is what you made or lost measured in units of what you risked. Losing exactly what your stop allowed is -1.0R; making three times that is +3.0R. Because it’s a ratio rather than an amount, a small trade and a large one compare directly. Losses that regularly come in worse than -1.0R mean the stop isn’t being held. That’s an exit discipline problem, and it’s the most common one this tool finds.

​

The scorecard

Three scores, 0–100, on every report. Above 70 shows green, 50–70 neutral, below 50 red. Each score comes with a why (collected into the analysis section) and a fix (collected into Coaching Recommendations). The three are not interchangeable, and the pairs diagnose better than any single number: Do this: rather than reading any single number, watch which of the three sits lowest across your last ten reports. Your running averages are on the Overview tab and in Profile.

​

The attribution breakdown

Headed Mistake Classification on a loss and Edge Attribution on a win. Directly above it sits Root Cause Analysis (Trade Analysis on a win) which is the written reasoning behind each score. The list splits the result into contributing factors that add up to 100%, each with a share and an impact label: critical, high, medium, or low on a loss; primary, significant, supporting, or minor on a win. What it separates is the part that was your decision from the part that was the market. If the largest share names something you control, it belongs in your playbook. If it names market conditions, there is nothing here to fix.

​

Counterfactuals

Headed What-If Analysis, usually two or three of them. A counterfactual is a version of events that didn’t happen, the same trade run again with one thing changed. Each names the change, states its impact, and explains what would have followed. Typical shapes: the stop held, the trade carried to your original target, half the size, or the trade not taken.

​

What this actually means

A counterfactual measures the size of one habit, and it is not hindsight scolding you. If “exited at my stated target instead of closing early” turns this trade around, and the next four as well, you know in percentage terms what that habit is worth. A single counterfactual is thin evidence. The one that recurs across reports is the one worth a rule.

​

The lesson

Every report ends with one rule, in quotes, plus a short pre-trade checklist and a trigger, the specific situation in which the rule applies. One rule per report, deliberately. A list of fifteen things to do better is a list nobody follows.

​

Data coverage

Bottom right, under Data Confidence: a percentage, and counts of how many fields were Known, Inferred, Assumed, and Missing. This is the report grading its own evidence. Known means you told Tradion. Inferred means it worked the value out from something you did say. Assumed means it filled a gap with a default and is saying so. A report built mostly on assumptions is a hypothesis rather than a finding. If the number is low and the conclusion matters, log the trade again with more detail. A logged trade cannot be edited in place, so that means deleting it and re-running.

​

Turning a finding into a playbook rule

Save to Playbook sits under the lesson. The generated rule is a starting point: open Playbook and sharpen it into something you could tick a box against, then give it a trigger. A rule without a trigger never fires, because you never remember to look for it.