Why failed alerts must remain visible

The mistake this prevents

A signal service can look accurate simply by deleting, editing or never revisiting its inconvenient alerts. The practical defence is a record committed before the outcome was known, followed by append-only updates.

What the record has to show

At P4, the chart stops at the completed cutoff bar and says that P5 is still unknown. Later, the same record receives one of the named lifecycle outcomes: P5 formed, invalidated before P5, still open at cutoff, or unresolved data.

If P5 forms, a separate observation records whether there was clear follow-through, no clear follow-through, incomplete future data or disagreement between valid interpretations. Positive and negative branches have the same visibility requirement.

How ReclaimRadar keeps the sequence inspectable

• The initial evidence is bound to one setup identity and cannot be silently replaced by a better-looking version.
• Later P5, Burn and result evidence is appended to the original record.
• The original market time stays visible even when publication or result updates happen later.
• Corrections are new timestamped records, not quiet edits.

The live failed case

EURGBP M5 rr-9WPF keeps its original bearish P4 preview and its later “no material follow-through” result. P5 formed, but the later observation did not meet the published movement rule. The record stayed visible.

What the current evidence does—and does not—prove

The public archive contains a much larger historical replay set and a small live set. Replays are useful for inspecting the lifecycle, but they are explicitly labelled and are not presented as alerts published live at their original market times. A small live sample proves the mechanism is visible; it does not establish a performance rate.