1 · THE QUICK ANSWER
What happened after the bot said no?
2 · PAIR SCOREBOARD
What happens most often for each pair?
“Best later” is hindsight; “final” is the last recorded checkpoint.
3 · DECISION DIAGNOSTICS
Which rules rejected winners and losers?
A setup can have several rejection reasons, so reason rows overlap.
By rejection reason
By confidence
By market regime
ACTION PLAN · WHAT TO IMPROVE NEXT
Recommendations backed by evidence
“Ready” means ready for a shadow test—not permission to change trading automatically.
4 · REPEATING BEHAVIOUR
Which pairs repeated the same setup?
Results group the pair, direction, recorded setup signal and market regime. Repeated scalp and strategy exposure may still describe one market event.
5 · EXIT SCOREBOARD
Did each exit protect money or leave opportunity?
Compares actual exits with later executable checkpoints without assuming the best checkpoint was predictable.
DETAIL · EXAMPLES YOU CAN CHECK
Rejected setups that later reached £10
The “best later” value is hindsight evidence, not guaranteed capturable profit.
How to read this safely
- All displayed money outcomes are net of the recorded fee model and use executable bid/ask quotes.
- A setup counts as a £10 recovery only when a recorded checkpoint reached at least £10 net.
- Best-later profit is selected after the event. A deterministic exit policy must prove it could capture that result.
- This report is bounded and cached so it cannot repeatedly scan the full production evidence store.
- No result on this page changes paper-trading policy.