// METHODOLOGY · PILLAR 05
Readiness drift detection.
Scoring drifts. Evaluators diverge. Findings stop closing. The record looks healthy until incident discovers it. Drift detection is the discipline that closes that gap.
Readiness drift detection — Readiness drift is detectable change in performance against the unit's calibrated baseline — surfaced before it becomes a gap that an incident discovers. Drift detection is what turns the silent failure mode of legacy readiness systems into a measured operational signal.
// 01 — THE QUIET FAILURE MODE
Why drift is the dangerous failure.
A loud failure — equipment breaks, exercise stops, incident occurs — generates immediate attention. The organization responds. The doctrine handles loud failure well.
A quiet failure — scoring drifts a fraction of a unit per cycle, evaluators diverge by tenths, findings take a week longer to close each quarter — generates no attention. The organization continues. The record looks healthy. Then the incident finds the accumulated drift, and the after-incident review reveals that the readiness data was lying for two years.
Drift detection is the discipline of catching the quiet failure before the incident does.
// 02 — THE BASELINE THAT MATTERS
Self-baselined, not industry-baselined.
Drift is meaningless against an industry benchmark borrowed from elsewhere. Different rubric. Different evaluator pool. Different operational context. The benchmark is fiction; the comparison cannot be calibrated.
Drift is meaningful against the unit's own calibrated history. The same rubric, the same evaluator pool (calibrated), the same operational context. OCTAAR's longitudinal intelligence layer accrues this baseline cycle by cycle. Drift is measured against the only honest reference: the unit's own measured past.
The honest benchmark is the unit's own calibrated history.
// 03 — TWO KINDS OF DRIFT
Unit drift vs evaluator drift.
Unit drift: the unit is changing. Performance is degrading or improving against its own baseline. Surface area: training, personnel turnover, equipment, leadership transitions, mission tempo.
Evaluator drift: the unit is unchanged; the scoring is changing. Surface area: evaluator turnover, rubric reinterpretation, calibration lapse, observer fatigue.
Both produce signal that looks like 'readiness has changed.' The discipline is in separating them. OCTAAR's inter-rater reliability tracking is what makes the separation possible — without it, every drift looks like unit drift, and recalibration is invisible.
// 04 — WHAT HAPPENS WHEN DRIFT IS DETECTED
The closure loop.
A drift event in OCTAAR is not a chart. It is a finding. It has a severity, a domain, a routing, and an assigned action. Unit drift triggers a unit-level finding and a unit-level action. Evaluator drift triggers an evaluator-pool finding and a calibration action.
The drift event closes when its action closes. Closure is evidenced — the recalibration session happened, the unit completed the remediation, the next cycle confirms the drift has reversed. Drift that does not reverse stays open. Drift that reverses but reopens is escalated.
// READ NEXT
// Last updated · · OCTAAR Methodology Team
// FAQ
Direct answers.
How long does it take to build a baseline?
Can OCTAAR predict drift before it happens?
Is drift always bad?
How does OCTAAR distinguish drift from a one-off bad cycle?
// READY