Analysis of 504 King: Part IV – Monday, December 4, 2006

This is a fairly ordinary day with no particular upsets on the route.  The weather was “rain and snow in the late evening”.  This is a good example of how service manages to be fouled up even when nothing in particular is going on.

This day’s data contains no major delays, although some small increases in link times are evident westbound through the core during the two peaks.  These are much smaller than the effects visible on Friday, December 1.

The two minute headway wave eastbound through the morning peak is poorly regulated despite many short turns, and there are a few gaps with no immediately obvious cause.  What is particularly noticeable is that service leaves Roncesvalles eastbound in bunches and stays that way through the inbound trip.  Link times show no evidence of congestion until, oddly, between Crawford and Portland around 09:00 well after the peak is over.

The afternoon peak is somewhat disorganized, but not badly compared with other days.  There is some congestion westbound through the core, but no theatre district effect in the evening because it is a dark night for clubs and theatres.

A large gap opens up westbound just after 20:00 and it travels across the city to Dundas West widening to almost half an hour on the return trip southbound at Marion and Roncesvalles.  This gap is not filled until eastbound from Roncesvalles Carhouse, and even then headways through the evening are erratic.

King Dec 4 Graphic Timetable

King Dec 4 Headways Westbound

King Dec 4 Headways Eastbound

King Dec 4 Link Times Westbound

King Dec 4 Link Times Eastbound

One thought on “Analysis of 504 King: Part IV – Monday, December 4, 2006

  1. Thanks for posting all your analyses, Steve. I don’t know how many of your readers have the patience to look through all of these, but I find them quite interesting and am looking forward to more.

    One thing really stuck out to me in the December 4 data: specifically, the westbound headways from roughly 2:00 to 7:00 PM. When we look at the moving average, we see a headway that almost perfectly matches the supposed 4:00 headway shown in the Service Summary, all the way along King Street from River to Jameson. Yet the underlying data behind the moving average show a headway that fluctuates much more than the average implies, even a fairly steady average based on a relatively small number of points.

    For example, at King and Fraser, the nearly constant 4:00 average headway lasts from roughly 3:00-7:30 PM, but that includes more than a dozen bunched cars (headways of a minute or less) counterbalanced by half a dozen gaps of 9:00 or more (up to 12:00).

    I think this sums up nicely a lot of the complaints that people have with reliability. The average tells us that things are running smoothly, but try to tell that to the potential rider at King and Fraser who, expecting an average 4-minute headway, happens to luck out missing the car immediately preceding the 12-minute gap.


Comments are closed.