The International Simutrans Forum

Simutrans Extended => Simutrans-Extended gameplay discussion => Topic started by: jamespetts on March 24, 2009, 09:42:14 PM

Title: Would this feature be fun? (Economies of scale with vehicle fleets)
Post by: jamespetts on March 24, 2009, 09:42:14 PM
I thought of a possible feature for Simutrans this morning that I am considering adding to Simutrans-Experimental, but, before I start working on it, would like people's feedback on whether they would find that this would add interest to the game, or make it too difficult/hard to understand. I thought of posting this in "Extension requests", but, since I would be coding it myself, it did not think that it would be quite appropriate, as it would not strictly be an extension request.

The feature would be simple in itself: the more of any given type of vehicle that one has (up to a point), the less that it costs to maintain each vehicle of that type. So, for example, if a player's fleet contains only one Routemaster 'bus, it might cost 1.00c/km to maintain, but, if the fleet contained 100 Routemaster 'buses, each of that type would only cost 0.85c/km to maintain. This would reflect the real world, in which it costs considerably less to maintain a homogeneous fleet than a highly heterogeneous fleet. Players would then have the interesting task of balancing, on the one hand, having specialist vehicles good for one particular thing on specific routes, and, on the other, keeping the fleet similar to reduce maintenance costs.

To make the feature less capricious than it might otherwise be, it could be possible to define in a vehicle's data that, for the purposes of economies of scale on maintenance, it should be treated as being the same as any other vehicles in a list, specified by the pak file creator in a .dat file. Alternatively, there could be some sort of abstract grouping system.

One thing that I have not fully worked out yet is, if this feature was added, how it could be implemented in the GUI to make it easy for users to understand what is going on, and why the vehicle maintenance costs are not constant. Another thing for consideration is whether the number beyond which economies of scale no longer reduce the maintenance costs any further should be universal, or should be specified for each vehicle (or perhaps class of vehicle) individually.

Thoughts on these points, or any other issues relating to this potential feature, would be very welcome. Whether I add this feature to Simutrans-Experimental may well depend on feedback here.
Title: Re: Would this feature be fun? (Economies of scale with vehicle fleets)
Post by: vilvoh on March 24, 2009, 10:10:36 PM
I must admit the idea sounds interesting and It gives a realistic approach. In real life, transport companies usually buy dozens or hundreds of units of the same model but imho, it has a problem. Let me explain.

As far as I've seen and read, people usually choose the most profitable vehicle. Cheap, fast, medium capacity and cheap to maintain vehicles are always more attractive for the players. So, player's fleet is very homogeneous, there's just a liitle variety. As I see it, If you implement that you're somehow encouraging people to use always the same vehicles, because they'll pay less maintenance costs.

Title: Re: Would this feature be fun? (Economies of scale with vehicle fleets)
Post by: The Hood on March 24, 2009, 10:19:06 PM
As vilvoh says, people tend to use the same anyway.  Unlike in real life, simutrans doesn't have the quirky situations that mean that you have non-standard fleets (e.g. the power supply on this line isn't quite good enough, or there is a tighter than standard curve, lower than average tunnel etc, etc).  Interesting idea though...
Title: Re: Would this feature be fun? (Economies of scale with vehicle fleets)
Post by: jamespetts on March 24, 2009, 11:25:12 PM
Very interesting feedback - I had not quite thought of it from that angle before. One of the things that I am trying to do with Simutrans-Experimental is to give players greater reason to choose suitable vehicles for specific tasks: the steam engine physics means that early, lower speed electric locomotives have distinct advantages in some situations; the comfort rating and loading times will give a real distinction between short, medium and long distance services; the catering level will have a similar effect; the difference in the overcrowded (i.e., standing) capacity of vehicles allows players a tradeoff between comfort and capacity at peak usage levels; and weight limits for roads and rails, coupled with a greater variety of track and bridges with different limits and different costs, will force players to make a real distinction between trunk, secondary and minor routes.

In those circumstances, especially with the particularly wide variety of (currently rail, but likely to extend to road in the future) vehicles available in PakBritain (which I am planning to adapt into PakBritain-Experimental to use as the default pakset for Simutrans-Experimental), players might have considerably more incentive to keep a heterogeneous fleet, especially as new vehicles would be introduced quite frequently: more frequently than it would be economical to replace them.

To those who have replied so far: does any of that influence your view of the matter?
Title: Re: Would this feature be fun? (Economies of scale with vehicle fleets)
Post by: z9999 on March 25, 2009, 08:05:25 AM
In simutrans, if vehicle didn't move, it wouldn't cost.
I think this is not maintenance cost but mileage (fuel cost).

So, it doesn't depend on the number of vehicle, but depend on weight, road condition, engine speed (rev/sec) and so on.
Empty bus is lower than loaded bus. Expressway bus crawling along in the traffic jam is higher than usual.
Title: Re: Would this feature be fun? (Economies of scale with vehicle fleets)
Post by: jamespetts on March 25, 2009, 08:17:35 AM
Z9999,

the idea of the "running costs" is to simulate all kinds of variable costs: the 0 costs if not moving is just a programming shortcut to scale maintenance with use (which is somewhat unrealistic, since there is always a fixed level of maintenance required no matter what the use).
Title: Re: Would this feature be fun? (Economies of scale with vehicle fleets)
Post by: Dwachs on March 25, 2009, 08:41:33 AM
Quote from: z9999 on March 25, 2009, 08:05:25 AM
So, it doesn't depend on the number of vehicle, but depend on weight, road condition, engine speed (rev/sec) and so on.
Empty bus is lower than loaded bus. Expressway bus crawling along in the traffic jam is higher than usual.

It does not depend on speed. Regardless of traffic jams, hills up and down, vehicles pay maintenance for the travelled distance. (actually it is payed tile-wise).

What about emulating your idea with decreasing prices? The more often one buys a vehicle the cheaper it becomes?
Title: Re: Would this feature be fun? (Economies of scale with vehicle fleets)
Post by: the almighty snark on March 25, 2009, 08:44:19 AM
 interesting feature, if you decide to implement this then you should also try and make it so that vehicles will cost more to run and maintain whilst there getting older. Since old vehicles usually require more maintenance since there parts are starting to fail.

I think this could be interesting, but it will take away the divergence of a vehicle park(whilst most people already relay on a small array of vehicles.) Offcourse you could also make it so that people will get offers for a certain vehicle so that they can get it cheaper over a certain month. Im not quite sure how to implement this, and make it so that it wont pick vehicles that are already obsolete.

Just a thought, i think this is a pretty good idea!
Title: Re: Would this feature be fun? (Economies of scale with vehicle fleets)
Post by: The Hood on March 25, 2009, 09:02:30 AM
Actually I like Dwach's idea for purchase costs rather than maintenance costs.  Larger orders are certainly cheaper (as are follow-ons) in the real world, and as it is a one-off cost, it can be shown in the depot window and there is less room for confusion.  The only thing to think through would be about re-sale price - I'm not sure how that works at present though.
Title: Re: Would this feature be fun? (Economies of scale with vehicle fleets)
Post by: the almighty snark on March 25, 2009, 09:05:18 AM
Resell cost: How more vehicles of a certain type there are, how lower the price you get when reselling, since the 2nd hand market would be pretty saturated.
Title: Re: Would this feature be fun? (Economies of scale with vehicle fleets)
Post by: z9999 on March 25, 2009, 04:01:26 PM
Quote from: Dwachs on March 25, 2009, 08:41:33 AM
It does not depend on speed. Regardless of traffic jams, hills up and down, vehicles pay maintenance for the travelled distance. (actually it is payed tile-wise).

I'm talking about fuel cost and engine speed of RL.
I'm not talking about simutrans nor vehicle speed.
But this is off-topic, you don't need to reply about this.  :)

Quote from: The Hood on March 25, 2009, 09:02:30 AM
Actually I like Dwach's idea for purchase costs rather than maintenance costs. 

I thought this is a little bit realistic than jamespetts's one.
But in sumutrans, we can buy 100 vehicles and sell 99 vehicles at the same time.
Title: Re: Would this feature be fun? (Economies of scale with vehicle fleets)
Post by: jamespetts on March 25, 2009, 08:58:25 PM
Thank you all for your further feedback. The Almighty Snark - increased vehicle maintenance costs for obsolete vehicles are already present in Simutrans-Experimental.

The suggestion in relation to purchase costs is interesting, but somewhat complicated. Firstly, in reality, one will only make a saving if one orders a large number at more or less the same time. Buying one hundred Routemaster 'buses this year and one hundred in five years' time is not as cheap as buying two hundred now. The same does not carry over to maintenance costs. It would, I think, be quite difficult to implement that in a way that is easily comprehensible to players whilst retaining more or less the present system.

There is no particular reason to reduce purchase costs instead of maintenance costs, though: I am not sure why Z9999 thinks that it is more realistic that way - Z9999, could you explain? My understanding certainly is that a homogeneous fleet is considerably cheaper to maintain than a heterogeneous fleet.

I am not sure that I fully understand The Hood's suggestion that it would be any harder to show differential maintenance costs in the depot window than differential purchase costs - unless, The Hood, you had thought of some ingenious GUI for that? If so, please let us know! Resale costs at present, I understand, are based solely on a vehicle's age and its original purchase price: the newer that it is, the closer to the original price that it gets on resale. The original price is checked by reference to the underlying price specified in the vehicles' .dat files. In reality, bulk discounts would not make a great difference to the secondhand value of vehicles, except to the extent that a greater supply, given a constant demand, will lead to lower prices: but no account in any event is taken of supply and demand in the second-hand market when setting resale prices.
Title: Re: Would this feature be fun? (Economies of scale with vehicle fleets)
Post by: The Hood on March 26, 2009, 10:24:25 AM
I'll have another go at what I was trying to say!

- With purchase costs, it is one off.  The price shown is what you pay, there and then, and that is it.  Simple.
- With maintenance costs, you see a price in the depot, but you pay this every month.  In your system, this varies with the number of vehicles, so is liable to change (not at random, but unlike with increase due to age, in a slightly less easy to manage way).  So for the first vehicle you buy, it gives the full maintenance cost.  For the second, it is reduced, and a lower cost is shown.  But the lower cost applies to both vehicles, and you only saw the lower price quoted for the second.  Slightly less simple, and more confusing, or is that just me?
Title: Re: Would this feature be fun? (Economies of scale with vehicle fleets)
Post by: jamespetts on March 26, 2009, 11:17:13 AM
Ahh, yes, I understand the point now. It should not be too complicated, however. The question is whether, given the interesting balancing of incentives that it would involved it would, if clearly documented, be worth that bit of added complexity. Any views?
Title: Re: Would this feature be fun? (Economies of scale with vehicle fleets)
Post by: The Hood on March 26, 2009, 11:19:32 AM
Personal view - I don't think so.  Happy to be proved wrong though.