THE RECEIPTS.
Every completed bus traversal, translated into hourly congestion patterns, weekday comparisons and the journey-time percentiles that matter for reliability.
Connecting to overnight GPS results…
When does the hill bite?
Each cell shows the selected statistic pooled across the same weekday and hour in the report window. Hover or tap for the observation count.
| Loading… |
The weekday face-off
Select any pair of weekdays. The chart compares matching hours; the headline comparison uses the selected time band.
All-against-all weekday matrix
Difference in the selected metric, in minutes: column day minus row day. Positive means the column day was slower. Both cohorts need at least 20 observations.
| Loading… |
The timetable-planning view
Daypart summaries across all seven weekdays, so AM, interpeak and PM conditions can be compared directly.
| Loading… |
Daily results
Mean, minimum, maximum and every requested percentile, with observation counts for each actual date.
| Loading… |
What happened on that day?
Actual individual date/hour results — useful for examining a particular Friday afternoon, not just the pooled Friday pattern. Only hours with observed traversals are listed; omitted hours are not assumed to have zero delay.
| Loading… |
Hour-by-hour detail
One row for every weekday and hour — all journey-time statistics and the underlying sample count.
| Loading… |
Is it getting better?
Completed Monday–Sunday weeks and comparison against the immediately preceding week, where both have observations.
| Loading… |
C8 dwell adjustment
A→D times deduct the greater of an assumed 90-second C8 stop allowance or observed qualifying stationary C8 dwell. These are not added together. A→B and A→C are unaffected. Original observed times are preserved in the CSV. This is a modelling assumption and may remove some traffic delay at C8.
What these numbers do — and don't — show
Based on real BODS SIRI-VM GPS fixes intersecting our directional virtual gates. Times are interpolated between fixes, not stopwatch measurements. Different gate sequences are different corridor measures and should not be added together as separate buses.
Percentiles use all individual qualifying journeys within the selected group, not an average of daily percentiles. A sample below 20 is indicative, especially for P85/P90/P95; missing feeds and empty days are never treated as zero congestion. Very slow or unusually gappy traversals should be reviewed before external quotation.
Initial collection began on Sunday evening, 18 October 2026. That date is partial, and individual subsequent dates may also have AVL outages. Comparisons are descriptive and not adjusted for traffic incidents, school terms or changes in demand.
← Back to the live downhill derby