The Myth of “No Short Turns” (July 2022)

This post includes short turn counts for the month of July on the major east-west streetcar routes downtown. See also:

Updated August 14, 2022: Charts of travel times on King between Strachan and Dufferin have been added to show that although there were congestion problems, they existed only on specific days due to special events, not pervasively through the month of July.

According to the TTC CEO’s Report, short turns (a situation where a vehicle does not reach its scheduled destination but instead turns back at an earlier point) were all but eliminated in May 2019.

This is not to say that short turns should not exist. They are an inevitable part of transit operations where delays can occur, and are essential to restoration of regular service. Back in 2019, the TTC’s problem was that they were used very frequently either as a lazy way to manage service or in response to unrealistic schedules. Now they occur but are not reported.

Meanwhile, other problems with service such as bunching, gaps and missing vehicles are not reported or tracked (at least publicly) at all.

There is no way to avoid saying this: the reported level of short turns is a total misrepresentation of what actually happens on the street as any regular rider knows. Management gets to claim they have eliminated a problem, but in fact it persists.

  1. Methodology
  2. 501 Queen Eastbound at Kingston Road
  3. 504 King Westbound at Spadina and at Bathurst
  4. 504 King Eastbound at Parliament
  5. 504 King Eastbound at Dundas/Broadview
  6. 505 Dundas Westbound at Lansdowne
  7. 505 Dundas Eastbound at Parliament
  8. 506 Carlton Westbound at Lansdowne
  9. 506 Carlton Eastbound at Coxwell

Methodology

From TTC vehicle tracking data, it is possible to count the number of streetcars passing any point on the line. In order to determine how many short turns occur at a specific location, counts on either side of a turnback will reveal the answer.

For example, if the screenlines for counts on Queen are defined as Coxwell Avenue and Woodbine Avenue, then the difference in counts shows how many cars short-turned at Woodbine Loop.

For these analyses, the counts are grouped by hour and by day through a month. Next, all weekdays, Saturdays and Sundays are consolidated to show the difference between types of day. The weekday counts are broken out by week to catch short-lived effects.

Friday, July 1, Canada Day, is counted as a Sunday. Note that this means that there are six “Sundays” and only five “Saturdays” included in the totals. That is the reason the count of trips within the month is higher for Sundays than for Saturdays.

An important distinction in any analysis is between overall averages and a detailed view of operations. TTC has a bad habit of reporting stats, when they do so at all, on a monthly average basis. This blends together periods when service is good with periods when it is very bad giving the impression that things are going fairly well. Riders, of course, encounter and are angered by the bad times which happen too often and fairly predictably.

The raw data are at a minute-by-minute, vehicle-by-vehicle level. In the charts here, I have tried to strike a balance between “information overload” with too much detail, and high level views that obscure what is happening on the street.

Continue reading