The International Simutrans Forum

 

Author Topic: Ideal spacing for city stations / new display option for passenger success  (Read 3108 times)

0 Members and 1 Guest are viewing this topic.

Offline zook2

  • *
  • Posts: 321
The current catchment area for passenger stations has a radius of 12 tiles. But if I understand it correctly, the walking time from the house to this stop becomes part of your overall travel time and thus decreases passenger demand. Ideally you'd have a stop at every house, except that loading times would slow down everything to a halt. Finding a good compromise is difficult, because demand depends on travel/waiting time and distance from the stop is as important as convoy spacing and speed.

I seem to have developed a habit of placing one every 10-12 tiles, which means a 50% overlap. How far apart do you place your stops?

Passenger success ratings for single houses doesn't seem to help me much, either. Unless it could be shown as a colour-coded display; not in the map window, but the normal display (as in the shaded "Shift-2" mode, where industries are red/yellow and public buildings green), with each each house shaded a different colour depending on passenger success rating. The shading mechanisms are already in place, the success rating is calculated, too, and it would help James fight the long, boring afternoons. What do you think?

Offline jamespetts

  • Simutrans-Extended project coordinator
  • Administrator
  • *
  • Posts: 20341
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: Ideal spacing for city stations / new display option for passenger success
« Reply #1 on: February 13, 2014, 09:42:25 PM »
Interesting discussion indeed. It is probably worthwhile looking at how closely that 'bus stops are spaced in real life: they have to cope with similar constraints. The spacing on my local road is rather close by: every 200-300m or so, which would be every 2-3 tiles. 12 tiles is 1.5km, meaning that the maximum walk to a 'bus stop is 0.75km, which is probably a bit far for most people, at least, in the modern age where the car is competition. I suspect that they should not be more than 5 tiles apart in towns.

As to the passenger success rate shading - this is an interesting idea. There is an enormous queue of other work at present which I am slowly coding, and better GUI comes rather below better simulation at present, but I shall bear it in mind for the future. Thank you for the suggestion!

Offline asaphxiix

  • *
  • Posts: 723
Re: Ideal spacing for city stations / new display option for passenger success
« Reply #2 on: February 13, 2014, 09:47:52 PM »
from all I've gathered, which is perhaps not enough, passenger success doesn't seem to be affected by transportation at all.

Offline jamespetts

  • Simutrans-Extended project coordinator
  • Administrator
  • *
  • Posts: 20341
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: Ideal spacing for city stations / new display option for passenger success
« Reply #3 on: February 13, 2014, 10:11:39 PM »
from all I've gathered, which is perhaps not enough, passenger success doesn't seem to be affected by transportation at all.

That is not correct, although passengers can walk to their destination.

Offline Sarlock

  • Devotees (Inactive)
  • *
  • Posts: 1340
  • Languages: EN
Re: Ideal spacing for city stations / new display option for passenger success
« Reply #4 on: February 13, 2014, 10:51:00 PM »
It's very hard to gauge the success rate of passengers as we don't know where the passengers are intending to go each year that we get the statistics.  One year might be higher then the next for no other reason than passengers chose to go someplace more accessible that year.

Having a look through the online game, which is currently in 1768, passenger success rates are in the 10-20% average range for both local and non-local transport in a few hubs that I have chosen.  These are well-served passenger hubs with service to half the map and probably 90% of the nearby cities.  Look at buildings at the edge of station catchment zones and they yield similar statistics.  Then look at buildings outside catchment areas and they are still in the same range.  I haven't spent a lot of time trying to see if there is an optimal station proximity/distance but at first glance it appears not that significant.

Offline jamespetts

  • Simutrans-Extended project coordinator
  • Administrator
  • *
  • Posts: 20341
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: Ideal spacing for city stations / new display option for passenger success
« Reply #5 on: February 13, 2014, 11:11:05 PM »
I am hoping, incidentally, that these figures will show a clearer difference with good passenger service with the passenger generation changes in the next major release, although it should be noted that passenger transport was so slow in 1768 that many passengers would not have found travelling worthwhile.

Offline ӔO

  • Devotees (Inactive)
  • *
  • Posts: 2345
  • Hopefully helpful
  • Languages: en, jp
Re: Ideal spacing for city stations / new display option for passenger success
« Reply #6 on: February 14, 2014, 04:31:55 PM »
There is an ideal range, but nothing exact.

Having the stops very close to each other slows down the service, which means less people are inclined to use it.
Having the stops too far apart causes gaps where people would need to walk, which ultimately reduces service quality.

According to this blog: http://www.humantransit.org/2010/11/san-francisco-a-rational-stop-spacing-plan.html
which I assume is by a transportation expert, 400m to 1km spacing is ideal for buses/trams that have a typical top speed of 40km/h to 50km/h inside cities.



For my own observations of real cities...
Subways/Metro, or similar slightly faster services, with top speed around 80km/h to 100km/h, 800m to 1.2km seems to be ideal.
Trains, with a top speed of 100km/h to 120km/h, typically have station spacing within 2km of each other inside cities and up to 4~6km in regional areas with lower population.

The key factors are population density and service speed.