← Back to overview
RESEARCH / AIMING

From recorded activity
to a policy draft.

Explore what was observable before momentum events and recorded re-entries. Predictive Aiming connects individual coin timelines with feature analysis and an inspectable policy draft.

01

Prepare the observations

Selected recording segments are grouped into timelines by session and coin. Timing, missing observations and recording gaps remain part of the context.

02

Investigate features and relationships

Examine individual measurements and changes around event windows and control periods. Depending on the selection, wallet context, lagged pairs and combinations extend the analysis.

03

Translate supported findings

The translator filters for compatible features and fits candidate conditions. It extends ENTRY in an existing policy with AND/OR groups; HOLD, EXIT, REENTRY and DROP are retained. Insufficient evidence can produce no draft.

The draft is exploratory. Feature selection uses the completed analysis, so its diagnostics are not an independent validation. Review and delayed replay are separate next steps.

See how it connects

English diagram
Simplified workflow based on the Dev11 AuditFix2 source. Internal formulas and selection criteria are omitted. This is an explanatory diagram, not an application screenshot.