One-screen view of predictability, throughput, misses, and quality signals
University trend points come from persisted source-pair metrics and the published pulse archive. The dashboard is designed to show miss patterns and execution quality, not just output volume.
University now has 12 source-backed checkpoint(s) in the persisted trend store.
These cards bias toward leadership questions: can we trust the promise set, what diluted it, and what actually came out of the sprint system.
Linked table values open the frozen Jira issue set used as the metric numerator. Percentages also show numerator / denominator so the calculation is visible before drilling down.
| Metric | Definition / formula | Direction |
|---|---|---|
| Committed completion | Committed items marked Done / all committed items | Higher is better |
| Committed carryover | Committed carryovers / all committed items | Lower is better |
| Finish predictability | Finish-intent items marked Done / all finish-intent committed items | Higher is better |
| Progress predictability | Progress-intent items that carried as intended / all progress-intent committed items | Higher is better |
| Added load | Items added after planning / all visible sprint items | Lower is usually better |
| Planning quality | Committed items marked Ready / all committed items | Higher is better |
| Carryovers marked Ready | Carryovers marked Ready at planning / all committed carryovers | Diagnostic |
| Bug share | Bugs delivered / all delivered work | Lower is usually better |
| Stories | Delivered items whose Jira issue type is Story | Count |
| Tasks | Delivered items whose Jira issue type is Task | Count |
| Bugs | Delivered items whose Jira issue type is Bug | Count |
| Partial | Committed carryovers classified as partial completion | Lower is better |
| Dependency | Committed carryovers classified as dependency delay | Lower is better |
| Not started | Committed carryovers that never started | Lower is better |
| Workflow mismatch | Done outcomes whose Jira workflow status was still non-final | Lower is better |
| Not-started committed | Committed items that carried without starting | Lower is better |
| Product stories | Delivered Story count, repeated in the execution-truth view | Count |
Historical checkpoints were reconstructed from their published pulse metric tables; frozen source-pair rows remain authoritative whenever both are available.
Choose a consecutive week window to inspect how the delivery system changed over time for this scope.