The International Simutrans Forum

In the thread http://forum.simutrans.com/index.php?topic=6613.0 the idea came up to collect all the features and dynamics which influence pak set balance and gameplay. I have started with a list of features which I could remember - this list was meant to be expanded and detailed further with the help of other people.

The entries marked with (?) are not (yet) existing, but come up in discussions now and then, and might be considered for future expansions and/or Simutrans Experimental.

Original post below:

---

I'll start a list of the dynamics/features, and the ones with better knowledge of STE might expand from this:

Vehicles

Achievable top speed
Acceleration
Waiting times
Capacity
Comfort
Price
Running cost
Maintenance
Pathfinding
Speed bonus
Ecological impact (?)


Ways

Weight limits
Speed limits
Signaling
Maintenance
Electrification


Stations

Capacity
Goods/mail enabled
Waiting times
Price
Maintenance
Catchment area


Depots

Price
Maintenance


Industries

Production rate
Storage capacities
Power supply
Workers


Economy

Taxes (?)
Inflation (?)
Loans/interest rate (?)
Goods price/demand fluctuations (?)
Max. allowed time of transport for goods ?)

Thanks Hajo. I'd appreciate it if we can limit this particular thread to Standard, perhaps deal with Experimental in a separate thread.

What I'm trying to outline is a functional understanding of how Simutrans models reality and how the player interacts with the simulation engine. This is a bit different to just documenting every feature; rather than document Simutrans as if it is a world in itself, I'm trying to build a picture of which parts of the real world it includes, and which parts it ignores. I'm looking only at the transport and economic modelling aspects, especially as, in the first instance, I am interested in finding a good, and hopefully simple, methodology for balancing paksets. I'll now try to start on the next level of detail, and also to cover some of the things that are not explicitly modelled in the sim. If anyone has done any evaluation of the significance of features that are or are not included, please share your reasoning and conclusions. Please also correct anything that is wrong. Units or brief descriptions in blue. I have some questions typed in maroon colour - if there is a wiki or link to the answer just say so.

By saying Not modelled I am not implying that it should be included. I am just observing that is a real world phenomenon that is not explicitly replicated in Simutrans Standard. Please add any more that you can think of.

Vehicles

Achievable top speed
Acceleration
Waiting times
Tare Weight tonnes
Capacity units of cargo, also tracked in tonnes
Comfort impacts passenger revenue?
Price
Running cost $ per km
Maintenance $ per km?
Pathfinding
Speed bonus average speed must increase in later years to earn the same amount of revenue
Depreciation is this a fixed % of original purchase price or % of residual value?
Instantaneous speed on an uncongested way is this a function of acceleration, horsepower, gross mass, and gradient?

Not modelled
Ecological impact
Fuel/energy consumption
Finite life
Fixed costs ($/hr or $/month)
Running and maintenance cost dependent on instantaneous gross mass
Increased maintenance costs with usage or age
Decreased performance with usage or age
Effects of obsolescence

Ways

Weight limits
Speed limits
Signaling
Maintenance $/month
Electrification

Not modelled
Finite life (where applicable)
Increased maintenance costs with usage or age
Upgrading (must be wholly replaced)

Does infrastructure depreciate?

Passengers & Mail

Trip generation any origin-destination combination possible on the map is equally likely, except for industry workers
Pathfinding limited number of interchanges, pax will walk very short distances
Trip revenue EDIT:$/km, Standard calculates on Manhattan distance, Experimental on Euclidian distance
Private car alternative a % of pax will use a car if public transport not available
Happiness (pax only) how is this calculated?
Time limit?

Stations

Passenger capacity
Goods/mail enabled
Waiting times how does this work?
Price $/month
Maintenance $/month
Catchment area

Not modelled
Finite life
Increased maintenance costs with usage or age
Upgrading (must be wholly replaced)

Depots

Storage capacity
Price
Maintenance $/month

Not modelled
Finite life
Increased maintenance costs with usage or age
Upgrading (must be wholly replaced)

Industries

Inputs
Consumption rate
Outputs
Production rate
Storage capacities (JIT mode) production halts if output storage is full
Power supply increases production rate if input storage is overfull
Workers generates passenger trip demand; no impact on industry output
Raw material suppliers contracted to supply particular customers assignment is equally likely to any customer regardless of distance

Not modelled
Closure of "old" industries
Failure of industries that are not supplied
Changes in industrial efficiency over time

Generation of new industries Can someone point me at the algorithm for this?

Not modelled
Most raw materials can only be produced in particular locations

Cargoes

Units e.g. tonnes, crates, m³, all also tracked in tonnes
Revenue $/km/unit multiplied by speedbonus
Perishability [insert one-liner for how this works, or insert factor into revenue above]

Not modelled
Alternative means of transport if not served by the player

Economy

Not modelled
Taxes
Inflation
Stock market
Goods price/demand fluctuations
Variation in fuel costs
Max. allowed time of transport for goods

City Population

Growth rate determined somehow by level of passenger transport provided?

Finances

Starting cash
Player can run an unlimited cash deficit subject to bankruptcy condition
Bankruptcy occurs if Net Worth < 0 at three consecutive month-ends

Not modelled
Loans/interest rate
Credit limit

Reported
Bank balance
Maintenance costs total
Operating costs total & per vehicle
Operating profit total & per vehicle
Revenue total & per vehicle
Assets total & per vehicle
Vehicle purchase expenditure (less proceeds of sales)
Infrastructure purchase expenditure
Powerline purchase expenditure
Net Wealth (Book Value)
Margin operating profit / revenue

Not reported
Return on capital
Total depreciation
Maintenance costs per vehicle

Landscape

20 buildable height levels
Tiles either flat or sloped at a standard gradient

Not modelled
Adjustments in trip revenue due to geographical barriers

Stochastic Events & Uncertainty

Not modelled
Major disasters
Collisions
Reliability/breakdowns
Industrial action

Game Events

No win conditions
Player loses upon bankruptcy player may ignore and continue

Scenario capability
...

What I can answer...

Comfort isn't in standard. It's partially modelled by speed bonus - or that was the intention, ages ago. (Sp.b. is fixed per convoy according to max speed.)

Running cost and maintenance are merged into one factor (per tile), same for buildings (per month).

Speed bonus is configurable separately for different way types. The weight of bonus is set on a per cargo type basis though. This leads to instability on waytypes with greater speed difference or boring game for small differences...

For ways, ownership (sort of?) is also significant in early stages of game: free public roads encourage road transport, up to saturation or economical limits ;D

Neither vehicles nor infrastructure are affected by age at all. Value of vehicles decreases, though, so it is always preferable to keep them stockpiled after purchase - sometimes to the point where building helper ways and driving half the map can be cheaper than "sell here / buy there".

Pax / mail generation has also rule that attractions generate return pax for incoming. No time limit.

Private cars in standard don't drive to any destination, it's a random walk constrained by nodes of road graph.

Waiting times - if you mean what I think, then it does not interact with transported cargo. It's essentially a full load timeout, useful for network "stability" - mainly prevention of lock-ups and jams.

Depots have no limit for storage capacity, or if there is one, it is for technical reasons. Or did you mean warehouses? Terminology...

New industries - a bit lengthier to explain. Triggered by city growth at exponentially increasing levels to linearise trend set by exp. growth (?). The basic unit considered here is a satisfied end-consumer input (>=1 supplier for particular input). An end consumer with N inputs needs N growth events. If there is no unsatisfied consumer, a new one is constructed. Chains needed to supply the particular input are generated without much constraints. Not sure if this is exactly what happens when generating the initial map. Example: materials wholesale takes planks, steel and concrete - 3 inputs.

Revenue - there's the code analysis (http://forum.simutrans.com/index.php?topic=3521.msg34455#msg34455) (thanks, Knightly!). That gives the hard numbers, but in short, you have a linear formula where the slope is given by speed bonus % and x is proportional to considered speed. That yields revenue per unit per tile.

Perishability - none. Once in transfer, cargo will exist as long as there is any valid route. Modelled partially by speed bonus - too slow transport won't pay off.

Ufff. Feel free to correct what is wrong - I'm not a manual :)

Quote from: Hajo on January 28, 2011, 10:51:39 AM

Acceleration

You can't define this directly (at least in standard); only through power (and gear) you can change the acceleration within a certain range.

A few brief comments on the list (reproduced with comments)

Vehicles

Achievable top speed
Acceleration
Waiting times Experimental only
Tare Weight tonnes
Capacity units of cargo, also tracked in tonnes
Comfort impacts passenger revenue? Experimental only
Price
Running cost $ per km
Maintenance $ per km?Not modelled separately from running cost
Pathfinding
Speed bonus average speed must increase in later years to earn the same amount of revenue
Depreciation is this a fixed % of original purchase price or % of residual value?
Instantaneous speed on an uncongested way is this a function of acceleration, horsepower, gross mass, and gradient?

Not modelled
Ecological impact
Fuel/energy consumption
Finite life
Fixed costs ($/hr or $/month) Experimental only
Running and maintenance cost dependent on instantaneous gross mass
Increased maintenance costs with usage or age
Decreased performance with usage or age
Effects of obsolescence Experimental only

Ways

Weight limits  Experimental only (for now)
Speed limits
Signaling
Maintenance $/month
Electrification

Not modelled
Finite life (where applicable)
Increased maintenance costs with usage or age
Upgrading (must be wholly replaced)
Experimental models way constraints also

Does infrastructure depreciate?  No

Passengers & Mail

Trip generation any origin-destination combination possible on the map is equally likely, except for industry workers
Pathfinding limited number of interchanges, pax will walk very short distances
Trip revenue $/km multiplied by speedbonus?
Private car alternative a % of pax will use a car if public transport not available Experimental only
Happiness (pax only) how is this calculated? This is not happiness per se, but whether passengers have a route to their destinations and their origin stop is not overcrowded; they register as "unhappy" if they can get to their destination, but the origin stop is overcrowded
Time limit? Experimental only

Stations

Passenger capacity
Goods/mail enabled
Waiting times how does this work? Experimental only
Price $/month
Maintenance $/month
Catchment area

Not modelled
Finite life
Increased maintenance costs with usage or age
Upgrading (must be wholly replaced) These can be upgraded, but at no less cost than replacing
Also not modelled: price tolerance
Also not modelled: choice of route by comfort of vehicles on the route
Also not modelled: the effect of prestige/publicity

Depots

Storage capacity Not finite
Price
Maintenance $/month

Not modelled
Finite life
Increased maintenance costs with usage or age
Upgrading (must be wholly replaced)

Industries

Inputs
Consumption rate
Outputs
Production rate
Storage capacities (JIT mode) production halts if output storage is full
Power supply increases production rate if input storage is overfull
Workers generates passenger trip demand; no impact on industry output
Raw material suppliers contracted to supply particular customers assignment is equally likely to any customer regardless of distance

Not modelled
Closure of "old" industries Experimental only
Failure of industries that are not supplied
Changes in industrial efficiency over time Experimental only

Generation of new industries Can someone point me at the algorithm for this?

Not modelled
Most raw materials can only be produced in particular locations Partly modelled - industries can appear in particular "climates" (that is, heights on the map). There has been some discussion of making climates depend on latitude as well as altitude.

Cargoes

Units e.g. tonnes, crates, m³, all also tracked in tonnes
Revenue $/km/unit multiplied by speedbonus
Perishability [insert one-liner for how this works, or insert factor into revenue above] In Standard, only the speed bonus simulates perishability

Not modelled
Alternative means of transport if not served by the player

Economy

Not modelled
Taxes
Inflation
Stock market
Goods price/demand fluctuations
Variation in fuel costs
Max. allowed time of transport for goods Experimental only

City Population

Growth rate determined somehow by level of passenger transport provided? Determined by proportion of generated passengers transported, proportion of generated mail transported and proportion of goods demanded by industries in the city supplied. In Experimental, the proportion of electricity demanded supplied is also relevant, as is congestion.

Finances

Starting cash
Player can run an unlimited cash deficit subject to bankruptcy condition
Bankruptcy occurs if Net Worth < 0 at three consecutive month-ends

Not modelled
Loans/interest rate Experimental only
Credit limit Experimental only

Reported
Bank balance
Maintenance costs total
Operating costs total & per vehicle
Operating profit total & per vehicle
Revenue total & per vehicle
Assets total & per vehicle
Vehicle purchase expenditure (less proceeds of sales)
Infrastructure purchase expenditure
Powerline purchase expenditure
Net Wealth (Book Value)
Margin operating profit / revenue

Not reported
Return on capital
Total depreciation
Maintenance costs per vehicle This is reported in the information windows for individual vehicles

Landscape

20 buildable height levels
Tiles either flat or sloped at a standard gradient

Not modelled
Adjustments in trip revenue due to geographical barriers

Stochastic Events & Uncertainty

Not modelled
Major disasters
Collisions
Reliability/breakdowns
Industrial action

Game Events

No win conditions These are present in scenarios
Player loses upon bankruptcy player may ignore and continue

Scenario capability
...

*****

Edit: Other dynamics not modelled in Simutrans

• Effect of competition on price/revenue
• Effect of changing lifestyle patterns on catering revenue (and, in Standard, catering revenue generally
• Economies of scale in relation to maintenance in having a homogeneous vehicle fleet
• The tendency for houses to be built a long distance from places where people can work only where suitable transport (whether public or private) is available
• Different classes of passengers
• State subsidies
• The effect of changing safety regulation on the usability of older vehicles at particular points in time (can be approximated by obsolescence in Experimental)
• The tendency for newly introduced vehicle types to be less reliable than more well-established types
• The requirement of larger aircraft to use longer runways
• National boundaries
• The market in secondhand vehicles
• The requirement to invest in more sophisticated and more expensive signalling for higher speed rail lines
• The requirement to have more sophisticated (and expensive) signalling equipment for very high capacity rail lines
• Different types of ticketing
• The practice of many transport companies in times of old of building/designing their own vehicles, and the consequent importance of the ability of the company's present chief mechanical engineer
• The changing relative costs of unskilled labour, skilled labour, fuel of different types, materials, etc. over time
• Maintenance cost of railway signalling equipment
• Seasonal variations in traffic demand
• The unsuitability (but the concurrent possibility) of using steam traction on underground railway networks
• The effect of social convention on perceptions by passengers of certain types of passenger vehicles (and, in Standard, comfort generally)
• Administrative overhead costs
• Different potential passengers having different walking distance tolerances
• Differential local economics (e.g., certain towns generating more 1st class passengers, etc. -  note, as above, classes of passengers generally are not simulated)
• Various factors choices of routing (in Standard, passengers/goods choose the route with fewest transfers; in Experimental, they choose the fastest route, and do not take account of other factors)
• The effect of the Internet on the volume of mail produced
• The ability to change the formation of trains mid-journey (switching an electric with a steam locomotive to continue the journey into a non-electrified part of the network, etc.)
• The effect on productivity/efficiency of the availability in the general (and local) economy of jobs other than those in transport companies and the consequent availability or otherwise of able employees
• The effect of state regulation on overheads, changing over time
• The tendency of passengers to prefer travelling overnight (where overnight journeys need to be made) on vehicles that have sleeping accomodation as opposed to those which do not
• The change in traffic patterns (particularly of passengers) at different times of day
• The fact that it takes time to build infrastructure (how much depending on the nature of the infrastructure and the point in history), and that the exact cost and time of so building is uncertain at the outset
• The possibility of passengers driving to a transport hub and taking public transport from there (e.g., "park and ride", and the concomitant impossibility of doing the same at the other end of the journey
• Taxis
• The effect of the wind speed and direction on the motion of sailing ships
• The inability of tall ships to pass beneath short bridges
• Opening bridges
• The inability of double-decker 'buses to pass under low bridges or tunnels (can be emulated for tunnels but not bridges using way constraints in Experimental)
• The effect of a vehicle's age/state of repair on its actual/perceived comfort to passengers (or, in Standard, comfort at all)
• The time that it takes between ordering new vehicles and their delivery, and that that time varies between different types of vehicle
• Fare evasion
• The necessity to deal in some way with the fact that much of the land on which one wants to build is owned by other people who may not be willing to sell it, and, if they are, may demand a wide range of prices for it (and the possibility of compulsory purchase as a means of avoiding this issue, which I suppose is implicit in a very simplified form already)
• Cycling (including cycle hire schemes)
• The ability to load/unload freight/mail from trains without the trains stopping in some cases
• The disruptive effect of maintenance/new construction on existing networks for the duration of the works
• The limited physical lifespan of horses
• The variance in performance between different horses
• The variation in performance/comfort of vehicles depending on how well that they are maintained
• The possibility of price competition
• The effect of facilities at stations on how long that passengers are prepared to wait at them
• The possibility of passengers staying at hotels en route on very long journeys
• Advertising revenue generated by permitting advertisements on busy stations/vehicles (and the variation of that revenue by how busy that those stations/vehicles are)
• The effect of public information produced by transport companies on the flow of passengers
• The possibility of mergers/demergers
• The possibility of jointly owned infrastructure/vehicles
• Leasing of vehicles
• The effect of traffic congestion on particular roads on the time that it takes to use those particular roads (and traffic congestion generally in Standard)
• The desirability to tourists and enthusiasts of riding on heritage vehicles
• Effects of the weather on demand and on disruption
• Transmission losses in power lines
• The necessity for electric forms of transport either to generate their own electricity or connect to an existing power grid
• The impact of a transport company's local access to fuel on its fuel costs
• The impact of the quality of fuel available on the performance of vehicles (particularly prominent with steam locomotives)
• The necessity for substations for electrically powered transport
• The impossibility of many modern industries existing at all unless connected to the power grid
• Global-scale transport (i.e., international shipping and aviation over thousands of miles, impractical to simulate at the Simutrans scale)
• The impact of different road layouts and markings on traffic flow
• Roundabouts and other sophisticated road junctions
• Toll roads/bridges/congestion charging zones, etc.
• The tendency for industry to develop where there is already a good transport link
• State regulation requiring certain freight carriers to be "common carriers", having no choice in the freight that they transport
• The ability to generate revenue from busy stations by using the space to let to commercial enterprises (retail, etc.)
• Short-term events affecting traffic (sporting events, special consignments of freight, etc.)
• Construction traffic
• The effect of ways on the comfort of land-based vehicles (and comfort generally in Standard)
• The effect of a convoy's braking ability on its speed
• The effect of vehicles' and ways' rolling resistance on the power needed to move (over) them at any given speed

That should suffice for the present. As with Moblet's original list, that something is not simulated does not mean that it ought to be: but for each item it would seem sensible to decide whether it should be ignored, simulated or its absence compensated for elsewhere.