Where the evidence comes from

Not your statements. The decisions you make here.

Nothing to upload. Nothing to connect.

How it builds

  • Every checkup adds decisions to your record.
  • Around 60 before a tilt is worth naming.

Below that it stays marked “too few to rely on.” We’d rather say that than hand you a confident-looking number.

What we do with it

  • Your choices, and the setups you made them in.
  • Delete on request — actually deleted.
  • Never used to recommend a trade.

Your review

312 decisions · 9 sessions · sample data

63%

of your buys came within 3% of a 20-day high.

When you passed, it was 21%. You act after the move has started.

40
Decisions
11.5d
Average hold
2.4×
Losers held longer

Most frequent mistakes

By count, across 9 sessions. Colour is rank, not severity.

Share of 40 decisions

In detail

You buy after the move.

25 of 40 buys sat within 3% of a 20-day high. Sixteen drew down 4% or more within five sessions.

Your own decisions · 25/40 · no model involved
You hold losers longer than winners.

19.2 days against 6.1. The gap holds across the whole set — it isn’t one or two outliers.

Your own decisions · median gap 2.1×
Seven stops went unhonoured.

Seven positions stayed on more than three days past your own level. Median loss −8.4%, against −2.9% for the ones you cut.

Your own decisions · 7 positions

What this doesn’t answer

  • What to buy next — never.
  • Which method suits you — no evidence yet.
  • Whether changing makes you money — we only promise to show whether the mistakes shrink.

Training

Levels follow your mistakes, not indicators. You clear one by changing behaviour, not by making money.

Chasing highs
Your #1
7 / 20
Holding losers
Your #2
Not started
Broken stops
Locked
Adding into volatility
Locked
Overtrading
Locked

What clearing looks like

Not “20 correct answers.” Your last 20 decisions show chasing down from 63% to under 40%.

Which takes time, and takes coming back. That’s what the subscription is: your evidence changing.