Transit, “soaking,” and you

Bunched streetcars on Spadina
TTC streetcars bunched at Spadina northbound at Queen

There is a problem of intentional bunching of streetcars and buses by a small few, but a very obvious few, TTC employees. In TTC-speak, this is called “soaking” — where one driver intentionally follows another vehicle closely in order to “soak” the lead bus or streetcar of passengers while having a very easy, passenger-free run.

After a very blatant display of soaking last week (for which, to its credit, the TTC has promised to look into), I finally found the impetus to bring this little-known, yet infuriating, practice out into the open. Bus and streetcar bunching is common, and a very old problem for transit systems everywhere. (The old joke about the Montreal Tramways Company is that their streetcars were like bananas: green and yellow and come in bunches.) In most instances, bunching is an unintended consequence of passive route management, traffic congestion and road conditions, uneven passenger loads due to ridership surges (such as high school dismissals), uneven loads on converging route branches, and/or poor scheduling. It is frustrating to the average TTC passenger because it makes a mockery of posted schedules and increases travel times. Bunching is most common on long, busy suburban bus routes, but sometimes even occurs on lightly travelled or short routes as well.

However, bunched buses on a busy, frequent route such as the 29 Dufferin will often “leapfrog” each other, and it generally works. Once in a while, other ad-hoc measures by frontline TTC employees are taken to mitigate the effect. I have seen two drivers decide to let the first bus at the start of a busy rush-hour trip run “doors closed” for a few stops and start picking up passengers further downstream, allowing the second bus, running “local,” to make the initial pick ups. This reduces crowding and waiting times down the line.

Short turns can be an effective way to reduce long gaps between vehicles and to address bunching, and are one of the only on-the-ground tools for managing a streetcar route (though unscheduled short-turns have drawbacks as well, just ask TTC customers in Long Branch or the Beaches). But attentive operators can make an extra effort, too. I have been on late streetcars on which the driver opens the rear door to anyone with a valid pass or transfer in order to speed loading times; at other times, operators of crowded, late vehicles encourage waiting passengers to wait a minute longer to get the next streetcar, which will offer a more less crowded, more comfortable ride.

This is where soaking, that special sort of bunching, comes in. A few bus drivers will obviously stay glued to the bumpers of the buses before them, taking on as few passengers as possible, leaving the bus in front to become a packed sweat box. Some drivers on shorter routes will pay little heed to the posted schedules to accomplish the same goal. This was a common problem a few years ago on the 90 Vaughan route. Instead of the posted 15-minute schedule during the evenings and weekends, buses would run 10 minutes apart, then 20 minutes apart, as one of the two drivers consistently ran “hot” (early), resulting in a reduced passenger load.

An experience last week Friday compelled me to finally write about this issue, which has been in the back of my mind for months.

My new regular commute is on the 506 Carlton Car. Eastbound in the AM peak, service on the 506 can be a bit spotty, but not by any means intolerable. In fact, I enjoy the slower streetcar ride over the multiple bus and subway transfers I used to rely on for my commute.

The diversion of the 505 Dundas Cars on College between Lansdowne and Spadina has been a blessing and a curse – the additional streetcars supplement the 506, but slow all traffic down, particularly at Spadina Avenue eastbound as they make the switch to head south.

On that Friday morning, I dashed out of my home, about 15 minutes late and found that I had just missed the previous eastbound 506 at Dufferin Street. According to NextBus, the next car from High Park will not arrive for over 10 minutes.

But before the next car from High Park arrived, two streetcars simultaneously emerged from the short-turn loop at Lansdowne (something that NextBus has yet to be able to predict). The first car, #4078, was already half full arriving at Dufferin. I could see the other streetcar right behind, and decided to wait to board that streetcar.

The first car loaded up, closed its doors and proceeded through the transit-held green light. But instead of stopping for an amber light turning red, the next car, #4021, accelerated through the amber/red light, leaving me to wait another seven minutes for the next car — which was of course, standing-room-only by the time the car reached Bathurst.

Along the way to work, I periodically checked Darwin O’Connor’s “Trans See” website, which is more detailed than the basic NextBus site, as it accurately lists vehicle fleet numbers and run numbers as well as the GPS arrival time estimates for each stop.

Incredibly, car 4021 kept within 60 seconds of 4078 all the way to Gerrard and Coxwell, when 4021 was finally sent for short-turn at the loop at Coxwell and Queen. I found it frustrating that up to that point, there was no attempt to even out service, causing a headache for riders and for the operator of #4078.

The solution is simple: the operator of the second car should have been instructed to wait at Lansdowne and College for two or three minutes after the first car left the short turn loop. It isn’t, pardon the pun, rocket science.

Transit advocate Steve Munro told me that he has seen many cases of soaking show up on his detailed route analyses, and it is not uncommon that the same vehicle and/or run number pulls the same stunt at the same time several days in a week, or gets away with it for consecutive trips in the same run.

Thankfully, TTC head office is aware of this problem this small minority of operators cause. I contacted TTC Communications Director Brad Ross while writing this post, and he assured me that the practice “is not acceptable and we work hard to stop it when we see it.” Furthermore, he promised to look into my specific example above, for which I provided additional information.

I do not understand the reasons why an operator would engage in such blatant soaking, apart from a strange sense of selfishness or apathy. It is certainly a problem only amongst a small group of operators, but it looks terrible from a customer service standpoint.

My hope for this post is that the TTC will rise up to the challenge to end soaking altogether, and that passengers themselves will be made better aware of this practice and call it out. That’s probably the only way things will change.


  1. The other way to deal with it, as Munro has long advocated, is to move to headway-based management on frequent routes, especially on streetcar lines.

    An (in my opinion and probably very conceptually) easy, automated, way of doing this would be to install programmed signals at various points along the route that would hold a streetcar or bus until the scheduled headway has passed since the last passing vehicle. 

  2. When I see two ‘soaked’ buses together, I always take the second one.  Colour me spiteful, but really am I going to get on a packed car/bus and stand, or get on the empty one and sit? If the buses are that close together it’s not like I’m losing much time taking the second anyhow.

    But this is a really serious problem on the streetcar lines in particular, since there’s no real option to ‘leapfrog’ a packed bus (not until those socialists on Council approve those flying LRT’s of course 🙂  )

  3. Technically streetcars are supposed to stop at every stop where people are waiting even if they are packed to the rafters. Thankfully some drivers don’t do that when they are full.

    However for an virtually empty streetcar to skip someone waiting at a stop is definitely against the rules.

    Yesterday I added a feature to TransSee which will show off-route vehicles on the Google map with an O marker. This is useful to see short turning streetcars and unscheduled diversion. You can see them all the time on 504 King shortturning via Bathurst and Fleet.

  4. I noticed the “O” as well for a short time tracking the 511 last night. I noticed one NB streetcar making the right onto King Street that went from 2 to O, and guessed it was going to Russell Division for the night. Nice work on the new features. Now when will the TTC finally release the bus data? As a professional “systems analyst” I know I could spend hours just watching it, like watching an ant colony.

  5. When I was taking the Dufferin Bus from Queen to Bloor for middle school and high school, I always figured they were bunched together because it was a tough neighbourhood. 🙂


  6. The solution is simple…they should setup an automated analysis…any vehicle running too close for a certain amount of time should get flagged…then it should be forwarded to the onstreet route management people, or radio’d to the driver to slow down…if they don’t…put them on desk duty for a week…

  7. I used to live Toronto and was visiting yesterday. I waited for a southbound streetcar at spadina and harbord, when EIGHT streetcars headed north passed by, bunched together.  It was outrageous.  Others at the stop laughed with me at how awful that was. This has to stop.

  8. I’ve always understood soaking not as causing trouble for the operator who is being tailgated, but for the operator behind the one doing the tailgating.

    1. The first streetcar picks up just as many people as it would whether there was another directly behind it or not. (Assuming it’s relatively on schedule, and about 5 minutes behind streetcar 0)

    2. The second streetcar runs nearly empty and ahead of schedule. (Assuming it left early, and is less than 1 minute behind streetcar 1)

    3. The third streetcar gets soaked, and has to pick up twice as many (very irritated) passengers as it should, which slows it down even more. (Assuming it left on schedule, but is facing a double-sized headway of more than 10 minutes because of the jackass operating streetcar 2)

  9. John,

    I guess you are right about streetcars (though I was trying to do my part by waiting for streetcar #2 last Friday and perhaps make life slightly easier for streetcar #1).

    But for buses, bus #1 certanly does get the short end of the stick because otherwise the bus behind would at least leap-frog and even things out a bit. Streetcars can’t do that. But if regular streetcar passengers knew that car #2 would stop and pick up those choosing not to crowd and slow down car#1, then my point with regard to streetcars would stand.

  10. @Barron

    Yesterday afternoon (around 7pm perhaps), a streetcar ran into mechanical problems at Dundas and Spadina, which caused a delay. That’s probably why you saw 8 streetcars in a row. 

  11. @Barron

    I think it’s important to realize the impact that traffic conditions and unforseen/unforseeable road situations have on bunching. For example, five days ago I saw a minor two-car accident at the corner of College and Spadina, and one of the drivers refused to move his car off the tracks because he wanted to prove where he was on the road when the accident occurred. As a result, streetcars were stopped in two directions, and I saw six northbound streetcars piled up on Spadina. One of the operators came out to tell the car driver to move his car, and did his best, but got nowhere with him. It was a full ten minutes before the car got moved, and several hundred people were waiting inside the streetcars that were delayed. Sure, these passengers might know why the delay happened, but I bet you that plenty of people farther down the line saw six streetcars coming up Spadina and thought something like “stupid TTC, can’t even do their jobs right!”

    It’s something that I’ve always wondered about: when people arrive late somewhere after driving, and blame the traffic, people are understanding and it’s acknowledged to be no one’s fault, but when they’re late on transit, even if they used traffic-impacted streetcars or buses, the frequent reaction is “yeah, transit sucks.” (And really, it’s cars causing the congestion that makes it hard to run surface transit.)

  12. I used to live off Dufferin and it was pretty common there (luckily I lived a short walk from the subway so I’d only hop on the bus if I saw one coming), but now that I have moved out to Jane and Annette it’s becoming quite clear how infuriating this can be for riders.

    I have to cross Jane in order to catch the station-bound bus, but the lights take a long time to favor the east-west traffic and if one is coming and you’re on the other side of the street, you’re out of luck (I ran the red once as the bus was pulling up to pick up passengers and the driver threatened to kick me off). Normally it wouldn’t be such a problem but Jane seems to have a big problem with soaking. Between the Jane buses bunching up, mixed with the Warren Park bus as it joins the Jane route, I’ve counted up to four buses riding each others coat tails — and this is at noon.

    Obviously this then effects northbound service where the platform gets so packed that some riders have to wait outside of the station on Armadale. Then a couple buses show up at the same time and one will take on riders while the other sits out of service for 10 minutes. The station does seem to be the best place to “reset” the schedule/distance between buses, but the only thing worse than sitting there with no bus not knowing how long you have to wait, is sitting there with the bus right in front of you and not knowing how long you have to wait.

    I’ve seen some TTC employees (or volunteers? They seemed in their late teens) go around the Jane subway platform trying to talk with riders and explain why buses do these sorts of things, but it seems like a pretty ineffectual practice — When a stranger walks up to you on the subway and asks for a couple minutes of your time, most peoples inclination is to ignore/dismiss them.

  13. tailgating streetcars are often directed to short-turn once they have passed through the downtown core.

    this creates an incentive for passengers to get on the first streetcar to avoid being possibly short-turned on the tailgating streetcar. such a choice, of course, contributes to bunching.

  14. The term “soaking” comes from the idea that the operator of a car behind a gap created by a “soaker” is working really hard, sweating and hence “being soaked”. This term has been around for as long as I can remember, and it’s not a new phenomenon.

    However, there seems to be little effort to minimize the problem even though the TTC can see where every car is, and someone managing a route would notice common patterns from day to day. Some within the TTC who acknowledge this problem have observed that the cheapest additional capacity can be obtained simply by spacing service.

    Another reason the first car in a parade really gets hammered is that there’s a good chance the second one won’t make it to the terminus, but will be short-turned. If I am waiting for a King car, I will board the first of three allegedly bound for Broadview Station because the other two may never get there, and I will have to wait at the short turn point for the next through car.

    Things on King and Queen are particularly bad now because with King West closed in Parkdale, all of the road traffic has shifted to Queen. Meanwhile, the traffic signal at Ronces has not been adjusted to give more green time to the east-west flow, and traffic can back up for blocks. This sort of change should be co-ordinated with road closures, but it isn’t.


  15. Once a streetcar or bus gets to near-max capacity it should be allowed to change its route sign to SET DOWN ONLY, stopping only at stops where an onboard passenger requests a stop. I’ve seen this done in Waterloo, where a bus driver requested permission to do so from Control.

    This would be easy to do on the new streetcars which will have LCD signs, especially since the old ones frequently don’t have the scheduled sign set right, leading to many King cars being 504 from the front and 512 from the rear…

  16. As an annoyed/frustrated passenger on a soaking streetcar, I get tempted to pull the “request stop” chord intentionally at every stop, just to get the streetcar to slow down.

    I’m sure the operator would be able to pick me out very easily (since the car would be empty). I would invent a dark stick with a hooked end to reach the chord, so I can inconspicuously pull the chord.

  17. How can passengers tell the difference if vehicles are soaking versus if vehicles are trying to be on time but stuck behind a slow streetcar leading the parade?

  18. Anybody who knows people who work in transit knows that this is actually a huge problem and causes significant stress not just for the passengers but for the drivers who resist engaging in these shenanigans — and, hence, end up being the soakee’s. TTC already has sufficient technology and staff to address this issue; what they don’t seem to have is any appetite for addressing the issue. Addressing the issue would probably mean acknowledging that many staff in the supervisory ranks charged with ensuring routes run efficiently are simply not doing their job. And of course the TTC would rather have everyone believe that the real customer service problems have to do with errant front-line employees.

  19. I’ve seen Mississauga buses, particularly the #19 Hurontario, skip picking up passengers when there’s another bus coming right behind it.  Definitely seems to help with speed and crush loading.  The drivers seem good about it too.

    Also, I’m sure I’ve seen some of the Mississauga buses with ‘Drop offs only” once they’re inside of Toronto city limits on their way to Kipling/Islington.  As mark Downling said, this sort of thing might be useful when buses are full….

  20. When bunching occurs, I agree with those who favour letting the first streetcar run “doors only” or “drop off only” thereby “accelerating” the service on the line.

    With the alternative of delaying the second streetcar, this car may eventually become the head of a bunch and necessitate delaying a third streetcar. These delays will continue to cascade through the entire line and result in slower overall service.

    As for the passengers cramming into the first car in a bunch to avoid getting on a streetcar that may short turn, the TTC could make it a policy to only short turn a streetcar if there’s another one right behind it. In practice, this mean that the car at the head of a bunch is the one likely to be short turned, and the passengers who have to get off will only have a short (or no) wait.

    It’s incredibly frustrating that the TTC doesn’t adopt such cheap and simple solutions.

  21. @Darwin, he should have had his cell camera running. I’m sure once the TTC says that there was nothing wrong/can’t prove anything/just tosses the issue aside because they don’t want to deal with it, Sean could go to the mainstream media and they could run another ‘expose’ of poor customer service as an empty streetcar leaves a waiting passenger.

    Anyways, once the TTC moves to a fare-paid honour system, they could theoretically run multiple streetcars together and have them stop in sync to operate as a mini-train (especially since streetcars can’t pass each other like buses can).

  22. College car. I’ve been taking it back and forth for years, and it ALWAYS shows up in pairs. I won’t run for the car anymore, because the other is always not more than 2 stops behind.

    And then you wait for 15 minutes for the next one.

    Always. Middle of the day, rush hour, you name it. Ask anyone who stands at the corner of Carlton and Yonge.

  23. This is why all public data such as transit, mapping, budgets, etc should be made available and accessible digitally.  Not only made available but also provided in a format that is easy to import into various applications to analyse and scrutinize.  It’s human nature for employee’s whether they be in the public or private sector to act in their own self interest at times.  One way to even the playing field to hold people accountable by utilizing information such as this example.

  24. @Ben Smith

    “they could theoretically run multiple streetcars together and have them stop in sync to operate as a mini-train”

    The same hourly passenger flow is given this way, but passengers would note a drop in frequencies and increased wait times.

  25. The various TTC idiocies have a lot to do with why I found it took the same time to walk to Union Station from Parliament and Carleton as take the 506 and subway. Happy to keep my $5 return for the trip too.

  26. This will continue forever because the TTC have no idea what customer service is. They’re just there to drive the bus, and the passengers are just an inconvenience

  27. One problem with not getting on the first car in a “train” is that it is usually the second (empty) car that is short-turned – of course the first car does not wai to let you on, and may be too full to do so anyway) so you then have to wait for the NEXT one.

  28. The 35 Jane busses are awful for this. I know it’s not simply a case of traffic causing bunching because at Jane station they park their fat asses on the side, wait for one poor bastard to finally crack and pick up customers, and then 2-3 jane empty busses will ride out on his tail

  29. It is a matter of supervision or lack thereof. Plain and simple. Crack a few heads, make examples of these operators (loss of pay) and you may see an improvement. Otherwise, nothing will change. Ever!

  30. Short turns is a different issue DavidC, and has a lot to do with the fact that the City and the Police Service refuse to give proper traffic signal and rule enforcement priority to mass transit at rush hour which then causes schedule slippage and short turns.

    The scramble at Dundas/Yonge is all very well but at least some extra time should be given to Dundas E/W at rush hour as streetcars are currently imprisoned in traffic jams, and TPS don’t adequately patrol Yonge to Bay to move along people picking up gadgets from Best Buy etc. I witnessed this impact on the 505 first hand on Friday evening, it was farcical.

    Towing vehicles should be very prominent on King, Queen, College, Dundas too – and used.

  31. Skip-stop / “leapfrog” operation is a great idea, unless you’re trying to get to a destination that is served by only one route variant, in which case it’s endlessly frustrating and actually increases travel times. Nothing is more annoying than having the “wrong” variant pull up to your stop while the “right” one speeds on ahead in the left lane. This is actually a serious consideration on the Dufferin route, which is busy and has some pretty major route variants.

  32. As a life-long transit user and an operator I find this discussion interesting, many posters really seem to understand many of the major issues that cause the bunching phenomenon. 

    I think it is important to point out that the vehicle being “soaked” is the one following the “soaker”, the bus or streetcar that is early and on the bumper of the vehicle ahead. The most common way for a soaker to do little work is to “jump the end” (leaving before the time listed on the waybill at the end/beginning of the line). This is against TTC policy but it often goes unpunished, that said generally this is only done by “certain bad apple operators” for which there really aren’t that many and who are well know by their coworkers and supervisors. 

    I personally feel that the riding public and TTC Operators share the most in common and that both sides would benefit greatly from some mutual respect and understanding. We are both negatively effected by the same traffic, the same policies, the same over crowding, the same bad apples, etc etc etc. If passengers wish to make it an “us vs. them” environment then everybody suffers and again we’re no further ahead. 

    I’d say the main reason TTC vehicles bunch up is because of the way the schedules are written in Toronto. Unlike almost any other city in the world there is often zero minutes of recovery time given at the end of the line, this has the benefit of reduced headways (on paper) and cost savings due to fewer vehicles being required to run the same frequency of service. The downside to this is that if a vehicle gets late it will often say late because there is no padding to the schedule. Not to mention that there is zero time allotted for the operator to use the washroom but calls of nature must be answered! Many afternoon and evening crews come with no break or meal period, requiring any dining to be done at the end of the line where possible. 

    My most important point to all of this is that there are many reasons, not excuses, for why operators do the things we do, getting one over on  passengers is not a goal 99% of us share. I’ve been and will continue to be on both sides and can appreciate the frustration that passengers feel but unless you’ve driven a mile in the operators seat you will never fully understand. To those that keep an open mind, trying to understand the full picture goes a long way and I thank you for your consideration.      

  33. I was northbound on Spadina, when two streetcars pulled up, the first was sufficiently standing-room (by no means packed), and as I was closer to the back, I walked to the streetcar directly behind it.  The driver refused to allow me (and several others to board) — infuriating.  I did however take the opportunity to glare at him through the back window all the way into the station!

  34. Excellent comment by Pat… How TTC schedules are written has a LOT to do with the amount of soaking taking place. From what I know, the technology and staffing exists to address the issue… But the issue wouldn’t loom so large if TTC’s schedules were more sensible. Transit users may or may not that over the past few years, some schedules on surface routes have been tightened considerably… even though traffic has gotten worse. There are routes where drivers would need to go well over the speed limit in order to complete the run on time. Less time means more hurried (and stressful) interaction with passengers…and certainly less willingness to wait for the passengers running to catch the bus… which of course further feeds into the increased stress of all parties involved. I don’t want to deny that there are some very rude operators out there… but I’d say the TTC’s insistence on unrealistic schedule is probably the major reason for deteriorating relations.

  35. “We run for those who wait, we cannot wait for those who run.”
    These are many of the same people that whine when service intervals are longer than expected. People practically throw themselves under the wheels to board a leaving vehicle when the next three are already in the station!

  36. This happens constantly on the Eglinton eastbound at Bathurst around 8 – 8:30 in the morning and the Bathurst southbound bus at Eglinton from 5 – 6;30/7p. It is especially infuriating for those who have difficulty standing up on crowded buses because of infirmities that may not be visible or who are sensitive to the heat produced on these crowded buses.
    None of the Eglinton buses ever seem interested in taking on passengers. Two buses sailed by as I was trying to get to the bus stop which is down about a block away from Bathurst and then 5 buses come by, packed to the rafters with people.
    And no one ever apologizes for the 20 – 30 min wait one frequently has before this caravan of buses appears, especially on Bathurst. The TTC schedule posted at Bathurst at Eglinton on the NW corner doesn’t even post any times, it just says frequent service, which apparently is 20 – 25 min. apart. Undoubtedly because of the previous caravan that went sailing by 25 min ago.

Comments are closed.