Industrial AI Governance & MLOps worked example

Model Performance Gap with performance gap impact score of 20 score: a worked example

Push performance gap impact score up to 20 score and the picture changes. This example computes every intermediate figure at that operating point. Use it when data scientists or model risk owners need to prioritize models with accuracy, precision, recall, latency, or drift concerns.

The inputs for this scenario

  • Performance gap impact (severity) score: 20 score (raised for this scenario; the documented default is 8)
  • Performance gap likelihood (occurrence) score: 5 score (unchanged)
  • Performance gap detection difficulty score: 4 score (unchanged)

Working through the calculation

  • Applying the documented formula (Model performance gap risk score = performance gap impact score × performance gap likelihood score × performance gap detection difficulty score) to the inputs above produces each figure below.
  • At this operating point the engine returns 10.75 score for model performance gap risk score, the number this scenario is built around.
  • At this operating point the engine returns 20 score for performance gap impact score.
  • At this operating point the engine returns 5 score for performance gap likelihood score.
  • At this operating point the engine returns 4 score for performance gap detection difficulty score.

How this compares with the baseline

  • Against the tool's baseline example, where performance gap impact score sits at 8 score and the headline result is 5.95 score, this scenario comes in 80.67% above the baseline at 10.75 score.
  • It multiplies a performance gap's impact, likelihood, and detection-difficulty scores into one risk priority number for ranking model issues. The value of this scenario is the size of the gap it exposes: that gap, priced out over a year, is the budget you can justify spending to close it.

Results at a glance

  • Model performance gap risk score: 10.75 score (headline result)
  • Performance gap impact score: 20 score
  • Performance gap likelihood score: 5 score
  • Performance gap detection difficulty score: 4 score

Run it with your numbers

  • Every input above is editable in the live Model Performance Gap calculator, which recalculates instantly and can be shared with the inputs intact.

Last reviewed 2026-05-12.