The International Simutrans Forum

 

Author Topic: Passengers boarding slower services  (Read 890 times)

0 Members and 1 Guest are viewing this topic.

Offline Carl

  • Devotee
  • *
  • Posts: 1600
    • Website
  • Languages: EN
Passengers boarding slower services
« on: September 04, 2013, 09:59:42 PM »
I have a hunch that something is awry with the system that is meant to prevent passengers from boarding a service which is much slower than the fastest available to their destination.

In particular, this can be observed on some of the Euston-Manchester lines on the recent maps I've uploaded. Because of the way the timetable works, one of the hourly Manchester diagrams looks as follows (apologies for the crude drawing):



So the train heads to Manchester via Crewe, and then back to Euston via Stoke. Another hourly diagram is the same, but in reverse.

The problem is that passengers at Crewe on the northbound trip are boarding to head to Euston - via Manchester. Similarly, passengers at Euston are boarding to head to Stoke, via Manchester. By my calculations, Crewe to Euston on this service should take just over 3 hours, and Euston to Stoke is about the same, when taking into account the wait at Manchester. But in both cases, the fastest journey available is around 1.5 hours. So the code be should preventing them from joining these much slower service options, and instead making them wait for the faster services.

I thought this might just be a phenomenon that needed to wait for data to accumulate and for the journey times to "iron out" -- but even after two month-ends have passed since the instigation of these services, I can still observe this behaviour. So something is amiss somewhere, I think.

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18745
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: Passengers boarding slower services
« Reply #1 on: September 06, 2013, 11:05:13 AM »
Apologies for not having spotted this earlier - the e-mail notification did not work for some reason.

Hmm - the issue might well be that the point to point timings system cannot cope with a ping-pong bat shaped route; it is rather difficult to get the system to work well with odd shaped routes. Quite what to do about this is difficult.

One long-term solution might be to allow convoys to be assigned to lines dynamically, so that one could have one (unidirectional) line on the way out and another on the way back. This would probably be necessary in any event if we were ever to have dividing/joining of convoys or similar.

Offline Carl

  • Devotee
  • *
  • Posts: 1600
    • Website
  • Languages: EN
Re: Passengers boarding slower services
« Reply #2 on: September 06, 2013, 11:15:32 AM »
Thanks for your reply, James. I just wanted to clarify a few of the details here, since I think that (in theory at least) this arrangement of lines is much less complex than one that we dealt with some time ago.

The clockwise trip (Euston-Crewe-Manchester-Stoke-Euston) and the anti-clockwise trip (Euston-Stoke-Manchester-Crewe-Euston) are assigned to separate lines. Call them line 1 and line 2 respectively. This means that every Crewe>Euston trip on line 1 takes three hours, and every Stoke>Euston trip on line 2 takes three hours. So there is no ambiguity (as with examples we've dealt  with before) with there being fast and slow options associated with the same line.

Since the point-to-point system stores journey times between halts on a per-line basis (as I understand it), I'm confused as to why the passengers would ever mistake Line 1 for a viable option for a Crewe-Euston journey. Surely they should simply flag it as a slow option and wait for a faster service -- as they do when faced with the opportunity with the London Midland slow service from Crewe to Euston via Stoke.

Given your comment about the ping-pong bat shape, is the thought that the double-stop at Stockport (both to and from Manchester) is causing the confusion somehow?