Every closed trade gets a letter grade, A through F.
The letter judges the decision you made with the information you had at the time. That is why a trade that made money can still come back an F.
Your wallet already tells you whether a trade made money. That is the only review most traders get, and it teaches the wrong lesson. The reckless entry that happened to run gets filed as skill. The disciplined trade that lost gets filed as a mistake.
Dossier grades the trade instead. Every closed position gets a letter, A through F, on the decisions you made with the information available at the time. Buys, sells and partial exits are grouped into one position, so scaling in and out produces one grade rather than six unrelated rows.
New trades land in seconds. The graded post-mortem typically follows in under a minute, and close-to-grade is under 45 seconds at p95.
Why a letter and not a score out of 100
A 0-100 score claims a precision that does not exist. It asserts that the gap between a 71 and a 74 is real and defensible. It is not. What you need to know about a trade is coarse: was that fine, was that sloppy, or was that the kind of trade that ends accounts.
A letter is a bucket, and buckets are what you act on. You have read letter grades your whole life, so a C lands without a legend.
It also has to survive volume. Four hundred trades produce four hundred grades, and the only thing you can do with that many is sort and count them. Letters sort. A two-decimal index just gives you something to argue with.
The grade judges the decision, not the outcome
This is the part people do not like. A profitable trade can grade badly, and it should. You chase a vertical candle, size it well above your normal, hold with nothing defined as wrong, and get out green because the token kept running. The money was luck. The process was a disaster. Grade it on the profit and you have taught yourself to repeat it until the day it does not run.
The reverse holds. A trade with a stated entry condition, size in line with the rest of your book, and an exit taken for a reason can lose money and still grade well. Losing trades are not the problem. Unrepeatable ones are.
Grading on PnL would be free and useless. You already have that number.
Exits are judged only on what you could see before the sell
The reconstruction runs to 60 minutes past your exit so you can see what happened next. The grade does not use it. Exit quality is judged only on data that existed before the sell went through.
Otherwise the feedback rots. Every exit before the top becomes EXIT_TOO_EARLY, every hold that survived becomes patience, and every panic sell that happened to dodge a rug turns into a good read. That is a grader that rewards whatever worked and marks you down for information you never had.
The hour after your exit is context for you, not evidence against you.
What the grade actually considers
Five inputs. All of them existed at the time of the trade.
The coach reads all of that and writes one paragraph. It interprets the figures, it does not produce them. Entry price, hold time, size and pattern counts are computed from your data before the model sees anything.
- Entry timing against the reconstructed chart: 1-minute candles, 15-second bars on scalps, from 15 minutes before entry to 60 minutes after exit. Where your fill sits inside the move you bought into.
- Exit timing, read only against the chart as it stood before the sell went through.
- Token provenance: bundler-launch flag, launch time, deployer wallet, and a 1-10 risk score. Buying a coordinated launch is a different decision from buying an organic one, even when the charts look identical.
- Position size, read against how you have been sizing lately rather than against a fixed number that means nothing on your wallet.
- How the trade sits against your recent trades and a rolling 90-day profile: what you closed just before it, and what you keep doing.
The letter is relative to you
Size is the clearest case. Four SOL is a rounding error for one wallet and a month of rent for another, so a fixed threshold would grade the wallet rather than the decision. Size is read against your own recent trades and your own 90-day profile, so the same position can grade differently for two traders. That is correct.
The profile does the same for behaviour, and every closed trade is checked against the 10-mistake taxonomy. A trade that reads as unremarkable in isolation sits differently when it is the fourth REVENGE_TRADE this month, and repeat offences are flagged from the second occurrence with a running count.
You will not find a points table here. The letter is a judgement over those inputs, and writing it out as arithmetic would make it look more exact than it is.
One grade is close to noise
A single letter on one trade is not worth much. You will disagree with some of them.
The distribution is the point. A hundred graded trades, sorted and counted, turn "I think I overtrade at night" into a named hour and a named leak. The Trader Report Card is recomputed on every graded trade: one page, your worst hour, your worst category, your single biggest leak.
None of it is a signal. It is a record of what you have already done, with the mistakes named and counted.
What it does not do
- It is not a 0-100 score, not a trade quality index, and not a radar chart. One letter, A to F.
- It does not use anything that happened after your sell. The chart runs 60 minutes past the exit so you can look at it. The grade cannot.
- It does not grade you against other traders. Size and behaviour are read against your own recent trades and your own 90-day profile.
- It does not grade open positions, and no grade implies an entry or an exit. Every letter is a post-mortem on a trade you have already closed.
- It does not publish a points table. There is no threshold to farm, because the letter is a judgement over the inputs rather than a sum.
Where the grade points you next
A letter on its own is a verdict. What makes it useful is the tag beside it, drawn from a fixed taxonomy of ten named mistakes, and the count of how many times you have made that one before.
The mechanics of how a trade gets from closed swap to graded entry are on how it works.