The International Simutrans Forum

 

Author Topic: A set of questions  (Read 2827 times)

0 Members and 1 Guest are viewing this topic.

Offline Anderson

  • *
  • Posts: 15
A set of questions
« on: October 26, 2011, 08:59:32 AM »
This is going to be an odd set of questions for a first post, but...
1) Is there any way to simulate the demand shifts which took place more or less from 1945 onward (and/or to simulate periodic variations in demand for certain transit types)?  Simtrans does an amazing job of simulating connectivity issues in a way that a lot of other games (Railroad Tycoon leaps to mind) fail to...but one thing I'd like to see is a situation where demand is falling on its own (forcing the player to improve services, which I can do the math on being expensive in most paksets, or to cut routes...and we all know how well that second option worked out for a lot of American railroads.
2) Are there any actual penalties (in terms of lost business) for either long waits or too many connections?  I know there's a cap on how many connections a passenger will make on a trip, but is there any bonus for better connectivity and/or cutting those waits aside from the smaller crowds at the station in question?
3) In general, is there any reason not to use a car which offers either more capacity for the same operating cost and/or can operate faster for the same cost?  I'm asking because in a few paksets, there seems to be no reason not to use a "roach coach" or a de facto subway car on an intercity route.

Offline Combuijs

  • Web Team
  • Devotee
  • *
  • Posts: 1392
  • Maintainer of maps.simutrans.com
    • Combuijs
  • Languages: EN, NL
Re: A set of questions
« Reply #1 on: October 26, 2011, 09:16:47 AM »
Welcome to the forum!

1) No, that's considered micro-management. With the public player you can of course remove and add factories and factory links, to simulate this, but you have to do that manually.

2) In Simutrans standard there is an option not to route passengers and goods over overcrowded stations (e.g. more passengers than its capacity). So passengers from A to C via B won't travel ("take the private car") if B is overcrowded. In the experimental branch of Simutrans total travel time is taken into account (based on historic figures) and if all alternatives take significantly more time than a private car would take, passengers won't travel at all.

3) No, consider that as a pak error.

Offline prissi

  • Developer
  • Administrator
  • *
  • Posts: 9584
  • Languages: De,EN,JP
Re: A set of questions
« Reply #2 on: October 26, 2011, 09:19:44 AM »
Actually demand is increasing with time, forceing the player to go with demand. Moreover the destinations pax want to travel spread out further. Additionally in experimental long waiting times will have pax walking away and not travelling. In Experimental there is a bonus, in standard not (yet).

Buses are usually geared for intercity service during balancing, while most trains are not. But that is a very pak set specific question, as all those paksets could have a different economy. ALso in next version the reward will be based on the route you travel. Therefore wasting a big intercity bus on local lines will not get good revenue. In experimental this is enforced by other means.

Offline Anderson

  • *
  • Posts: 15
Re: A set of questions
« Reply #3 on: October 26, 2011, 10:07:17 AM »
Thanks for the welcome and for the responses.  I've noticed the tendency towards increased demand...part of the reason I was wondering about these things is that I've noticed, for example, that I frequently need to run money-losing (albeit at a relatively low level of loss) bus lines to generate demand for my trains (which then generate large profits).

And as to the demand issue: I was particularly thinking passenger, not factory, demand, and thinking of handling this via a universal multiplier.

Another question: Though my networks never quite covered the entire map, is there ever a time that the high-capacity, high-cost passenger cars in the basic pak ever make sense to use?  At least early on (pre-1940), I was never able to find a situation where the revenue offset the sheer operating cost.

Finally...looks like I need to report the British128 pak for that goof.

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Devotee
  • *
  • Posts: 18753
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: A set of questions
« Reply #4 on: October 26, 2011, 02:55:14 PM »
Anderson,

the problems to which you refer are largely solved in Experimental. Dealing with them in turn:

(1) the reason for the falling demand is the rise of the private car. In Experimental, there are settings for the proportion of people who have access to a private car over time: the default is based on UK government statistics (with extrapolated guesses for years for which statistics are unavailable). For each passenger journey, the chance of the passengers using the private car is calculated. This depends on the quality of the road connexion between the passengers' origin and destination, and the quality of the alternative connexion. Passengers who use private cars do not use players' transport. Because this depends on the quality of the road connexion, players starting early in the game and playing a single player game without building more roads between towns than are there at the start of the game may not notice this feature. However, if a proper set of road connexions are in place, the level of passenger traffic going to private cars is intended to be as realistic as possible. I also plan to implement at some time in the future a means of tracking congestion much more accurately, and having the private car graphics in the game follow an actual route to a destination (currently, they wander around aimlessly).  The idea is for the public player in a multiplayer game to have a realistic incentive to improve the road network for private cars (as failing to do this will lead to visible congestion that will impair a town's growth). I have recently hit upon a way of doing this without an unacceptable impact on performance, and will be posting on this topic in the Experimental development section soon;

(2) in Experimental, passengers and goods will always take the quickest route to their destination; furthermore, they have a limited travelling time tolerance (a random figure per packet of passengers/goods set according to a pre-set range based on their intended journey distance), and, if the fastest travelling time is less than this, they will not travel at all. Waiting time is counted as well as travelling time towards this total. Each transfer will, of course, add to the time (two minutes as a hard minimum, more if any waiting is involved), so, all other things being equal, routes with fewer transfers are preferred;

(3) in Experimental, there is a comfort value for every passenger vehicle, which affects revenue: the longer the journey, the more significant the effect of comfort. Higher capacity vehicles in the Experimental version of Pak128.Britain ("Pak128.Britain-Ex") have, generally, a lower comfort than lower capacity vehicles; furthermore, otherwise similar vehicles that differ in comfort are likely to have a directly correlative relationship between comfort and price.

Have a go with Experimental and Pak128.Britain-Ex and see what you think - I should be interested in your views as to how well that you find that Experimental's features that were intended to deal with these issues do so.

Offline DatDamnDom

  • *
  • Posts: 15
  • Languages: English
Re: A set of questions
« Reply #5 on: April 19, 2012, 04:09:25 AM »
I'm curious as to the disposition of this issue.