Public methodology register
High-frequency excess predicted wait
- Version
- settled-predicted-excess-wait-v1
- ID
- network-regularity:excess-predicted-wait
Definition
- Definition
- The difference between waiting time implied by settled predicted arrival spacing and waiting time implied by scheduled spacing for eligible high-frequency route–stop hours.
- Formula
- Predicted EWT − scheduled EWT, where each EWT = Σ(gap²) ÷ (2 × Σ(gap)). Gaps are ordered independently for scheduled and predicted arrivals before aggregation.
- Units
- seconds, displayed in minutes
- Population
- Eligible high-frequency bus route–stop hours with ordered scheduled and settled predicted arrivals.
- Measurement window
- Eligible completed service days in the latest 30 days.
- Coverage
- Only service days that pass prediction-matching quality are included; every result retains its matched-arrival and stop counts.
- Publication eligibility
- At least five matched arrivals in a route–stop hour and an average scheduled headway from two through twelve minutes.
- Scope
- Completed service days in the latest 30 days that satisfy the prediction-matching quality threshold; route–stop hours require at least five matched arrivals and an average scheduled headway from two through twelve minutes.
- What this number can support
- Difference in wait implied by predicted spacing versus scheduled spacing. It is not observed passenger wait, actual door-opening regularity, or a low-frequency punctuality measure.
Implementation status
- Source candidate
- Implemented in this source candidate.
- Public production
- This metric version has not been independently verified in the public production deployment.
The calculation is implemented in this source candidate. This record does not claim that the same version has been independently verified in the public production deployment.
Sources and identifiers
- Method
settled-predicted-excess-wait-v1Settled predicted excess-wait method - Current API
reliability-evidenceCurrent network reliability public API
Exclusions
- This is not an observed passenger-wait survey, and it does not include low-frequency route–stop hours or rail without supported realtime observations.
- A negative result is not labelled better service; it only means the predicted spacing implied less wait than the imported timetable for the eligible population.
Uncertainty and interpretation
- Predicted arrivals are the last supported predictions near scheduled time, not recorded door-open events. Feed and matching quality therefore remain attached to the result.
Current evidence
Change history
No registered material change is recorded for this metric version.