News:

Simutrans Forum Archive
A complete record of the old Simutrans Forum.

Balancing notes from the online game

Started by jamespetts, January 22, 2012, 10:21:11 PM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

jamespetts

I thought that it would be useful to collect a list of balancing observations made whilst playing the new public server so that there is a central place where I can look for this information when I have time to do something about the issues raised.

In no particular order, here are a few balancing issues that have been raised so far that I can recall:


       
  • Tunnels cost too little to build but too much to maintain (both generally and compared particularly to bridges).
  • Railway locomotives cost comparatively far more than railway track: either locomotives should cost less, or track should cost more.
  • Successful railway/ship/road companies (at least, in the 1830s) are able to achieve an annual turnover on capital investment of over 100%,  whereas, at least for rail, this should be not more than 10% (and probably similar for canals).
  • Early steam locomotives are too slow and underpowered (at least after the Rocket - very early steam locomotives really were too slow and underpowered). This is - provisionally - fixed in the latest Pak128.Britain-Ex sources.
If anyone has any other thoughts, these would be welcome!
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

jamespetts

The passenger factor has been reported as being too high.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

AP

#2
With regards rail, I think a fundamental question is, should one be able to build a successful railway as a first move? (I failed twice!) My impression of the pak is that railways need a large number of trains to break even, as a consequence of the slow speeds being achieved by the locomotives, otherwise most track is empty most of the time.

I think the starting budget needs reviewing, up, or down as a consequence. $250k is a very large number of horses and carriages, and quite a lot of sailing ships (wherries especially), but not very much railway or canal. And passenger railways need feeder routes, which cost further money.

I also think that the speed of the game timeline needs to be adjusted. When everything is moving around at  10km/h in 1830, it's like watching paint dry waiting for deliveries to arrive, and would be helped by having the game run at e.g. double speed, if it can be set up to vary with the date or something?

I was also suprised that I had difficulty making an industrial chain run at a profit. Coal/Ore to a steel mill, to a cannery, to a Grocer's, ought to yield a tidy reliable profit. But I had difficulty encouraging the food inputs to yield much produce (orchards giving 6 crates of apples to go with an entire trainload of steel...), and the industries seemed to lose money too. The industrial revolution was built on heavy industry moving stuff around, passengers were secondary. I also thought the map had too little heavy industry by a significant margin, although it had a nice number of towns and decent distances between.

jamespetts

To answer the question - yes, it should be possible to build a successful railway as one's first move (as AEO and dustNbone have in fact done) at the first instance, if one is careful: in reality, after all, most railway companies were founded from scratch, rather than branching out from existing canal companies. The long term plan for starting funds is to require players to borrow all of their starting capital from the bank for a fixed term, with interest: players who want to start small will borrow a small amount (and have a lower risk), whereas players who want to start big can borrow much more, but will have to earn a higher margin in order to make up the interest payments.

The timeline is a tricky issue - there is no provision in the code for it to be varied with time, and, if more interest/detail was added to the 1830s era (I shall have to look at adding some more locomotives to that era, I think, and I am already adding details to the carriages more generally), and things are balanced better, there will be more to do in that era. In particular, for the next pakset release, I plan on making 1830s era locomotives more powerful and faster. We must take account of the fact that players won't be able to be online all the time and the fact that, in later eras, when there is more to do, going too fast will be a problem. A persistent state online game such as this is perhaps suited to a more relaxed pace of logging on for a few hours every night rather than playing intensively as one might with a single player game.

As to industry - I certainly agree that this map has far too few: I shall have to rectify that the next time that I start a map. I am not quite sure why freight transport was not profitable - perhaps the revenues from freight are too low. The pakset does need rebalancing generally, so I shall have to look into the relative revenue of freight, I think. Does anyone have any suggestions in that regard?
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

jamespetts

#4
Another item - not enough variety of steam locomotives in the late 1830s and early 1840s.

Edit:

Further -
* the Jenny Lind's per kilometre maintenance cost is far too low compared with other, later locomotives; and
* town growth is considerably too low.

Edit 2:
Also -
* the map roughness is too low; and
* the mountain height is too low.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

ӔO

#5
some things I have noted while playing.


       
  • Improved wrought iron track does not come early enough / LNWR Crewe type cannot be used to its full potential and meets an early demise to usefulness when Jenny Lind is introduced in 1947/6
  • Freight locomotives generally have way too much power/maint./cost for the amount of freight available. Might be good to balance them against LNWR DX Goods.
  • Needs a weaker and cheaper freight locomotive for early game.
  • LNWR Sharp tank, somewhat of a mystery position with power/traction/cost/maint. compared to other locomotives available in 1850's. It is worse in every way compared to Jenny Lind.
  • LNWR Problem / LNWR Bloomer, imbalanced against each other (possibly switch power/cost/maint. specs between the two)
  • Not enough freight chains that are economically feasible
  • 115km/h Brick viaduct is missing weight limit
  • 200km/h Brick viaduct and Elevated Brick Viaduct, perhaps available too early
  • Brick viaduct bridges need a length limit
  • horse omnibus and hackney carriage could use an overcrowded capacity. Otherwise, it is very hard to notice when a horse carriage line is overloaded and can result in massive refunds that go unnoticed in the Line Management window.
  • IMO, map might be better with 16:9 (or 9:16) ratio
My Sketchup open project sources
various projects rolled up: http://dl.dropbox.com/u/17111233/Roll_up.rar

Colour safe chart:

sdog

Not a note from the online game, but it fits to what AEO said:
There is also a need for an a cheap, but not too fast engine for local trains and intra city trains as feeders. I used my planet well into the 60s (and duplicated it in depot, when i needed more)

Some companies are very difficult to play LNWR has a lot of wagons and engines, in mid 1863 there are only tank engines available for LNWR. LBSCR is very advantageous when Jenny Lind is available, but when it becomes obsolete, one is lacking an engine. Difficult are also MR which has only wagons but no engines and SDR with engines but no wagons (or was it only a freight line?). I know this is likely due to the fact that it is quite a lot of work to paint the rolling stock, i just mentioned it for completeness.

Milko

Hello

My little experience.

The PSComet (steam ship PAX+Mail) running cost should be slightly reduced .... (20 -> 17/18)?

Giuseppe

AP

Quote from: AEO on January 27, 2012, 02:06:22 AMBrick viaduct bridges need a length limit
That one may be deliberate.

QuoteIMO, map might be better with 16:9 (or 9:16) ratio
I agree - most people have computer screens which are landscape, so having a map which can be maximised to fill it means you can see more information.

jamespetts

AP is right that the absence of a bridge length limit is deliberate, but they should not be able to cross deep ocean. I am planning to change the code to prevent bridges from crossing water that is more than 1 tile below sea level, and also prevent raising/lowering of land more than 1 tile below sea level, whilst increasing the cost by a factor of 5 for raising/lowering of land at sea level.

As to LNWR being a difficult company to play: first of all, the primary aim at this stage is not to enable players to play a single company, although it would be good to have that capability eventually (although it would require a very great many more vehicles), but for players to have a good selection of historical vehicles (from any company) in any given era to enable interesting and economically productive choices to be made about which vehicles to use.

A few miscellaneous points: the weight limit of the earlier bridge has been added on Github; and the hackney carriage and horse omnibus not having an overcrowded capacity is intentional, as neither of these vehicles in real life permitted standing.

The other points are noted for consideration.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

Junna

Quote from: jamespetts on January 27, 2012, 09:26:07 PM
and also prevent raising/lowering of land more than 1 tile below sea level, whilst increasing the cost by a factor of 5 for raising/lowering of land at sea level

I think this is perhaps a bit harsh - unless public player is exempt from this restriction? I think it would be better with just a really excessive and prohibitive cost for it (much, much more than just 5x.)

sdog

often enough, cost doesn't matter at some point in the game. then you see those strange constructions, further tilting the game towards the player with the large advantage.

if not too difficult, excempting the public player from this restriction as suggested by Junna, would be very sensible.

AP

Quote from: sdog on January 28, 2012, 05:13:56 AM
often enough, cost doesn't matter at some point in the game. then you see those strange constructions, further tilting the game towards the player with the large advantage

I think this is a significant balancing problem in the later game though, rather than something which should be used as a reason to price things out of reach in the early game.

What is needed is a mechanism to cause successful players with runaway bank balances to have to spend more than unsuccessful players with low bank balances. I suggested in another thread that costs of making changes to infrastructure might be proportional to the number of convoys using the tile each month, for instance - that's just one idea (and not an unrealistic one).

jamespetts

Junna,

as to the public player being exempt - there is a certain tension between the public player's two functions, firstly as map editor at the beginning of the game, and second as government after the game has started. As government, the public player should have just as much restriction on its physical activities as any other player, but some people prefer to use it for map editing purposes. I am beginning to wonder whether a good means of separating the two would be to disapply restrictions to the public player if freeplay mode is selected: in freeplay mode, the public player would be the map editor, and in non-freeplay mode, it would be the government. Additionally, everything would be free to the public player in freeplay mode so that the bank balance is not affected by map editing when freeplay is turned off. The idea would be that one can generate a map, turn freeplay on, edit with the public player, turn freeplay off, and then start playing. I should be interested in any thoughts on that idea.

As to balancing more generally - sdog and AP are correct to identify this as a problem. Fixing it is planned, but not easy. It was Moblet last year who explained at least part of (and possibly all of) the cause of this problem, being the insufficient utilisation based variability of costs. A number of his suggestions I plan to implement, as discussed here, but these will take time, especially as, at present, it is only me doing the coding.

AP's suggestion is an interesting one, but could easily be manipulated by a player just re-routing or stopping convoys for a month just before the change takes place to reduce the cost artificially.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

Junna

Quote from: AP on January 28, 2012, 10:38:39 AM
I suggested in another thread that costs of making changes to infrastructure might be proportional to the number of convoys using the tile each month, for instance - that's just one idea (and not an unrealistic one).

Is that such a good solution, though? A player with high number of convoys is not necessarily the most stable and largest - your network for example, to use the latest online game as a basis, is far larger than mine, but I do have the most frequently travelled railway segment at around 25 convoys a month. Perhaps it would also be able to take into consideration also the general size (in terms of total income and total passengers carried)?

Quote from: jamespetts on January 28, 2012, 12:29:00 PM
I am beginning to wonder whether a good means of separating the two would be to disapply restrictions to the public player if freeplay mode is selected: in freeplay mode, the public player would be the map editor, and in non-freeplay mode, it would be the government. Additionally, everything would be free to the public player in freeplay mode so that the bank balance is not affected by map editing when freeplay is turned off. The idea would be that one can generate a map, turn freeplay on, edit with the public player, turn freeplay off, and then start playing. I should be interested in any thoughts on that idea

I do think that sounds like a fairly good compromise - is it easier or more difficult to do it that way than to exempt the public player entirely?

jamespetts

It is a more desirable way of doing it than exempting the public player entirely, since the public player can then be used as a government in game and have its own playing goals. I can set it to disable increasing/decreasing town size, manually adding industries and such other things, too, when not in freeplay mode.

Edit: As to balancing more generally, the better solutions are the utilisation based solutions in the thread linked above that are planned.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

Milko

Hello James

In the online game "PSComet" running cost is 20, after the latest revision 3.5. Before your last change the value was 8.2, may be that you have inadvertently reduced the cost twice making it too low? (I think less then 15 is not a good choice ...).

Giuseppe

jamespetts

Ahh - it is indeed possible that I have reduced it twice by mistake. Try the latest commit and see whether that seems correct.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

sdog

QuoteThe idea would be that one can generate a map, turn freeplay on, edit with the public player, turn freeplay off, and then start playing. I should be interested in any thoughts on that idea.

i'm not so sure if that is a good idea. the public player mostly has the function of a server admin or game-master now. And that is good i think. Turning the public player into something that could be played in it's own right is so incredibly distant from anything that could be achieved in simutrans that no effort should be spent to keep it possible to do so.

Simutrans is quite a decent network simulation, a straggling economic simulation, but simulating public sector -- seriously?

What you suggest above is i think needlessly complicated, restricts the use of the public player in game administration and doesn't really gain anything.

ӔO

IMO, if exchanging money between players was implemented, all of the public sector simulations can be done.
Usually, public transport funded by the local government will set up a Commission, Bureau, Branch, Division, etc. and operate it like any other business that gets government subsidies. The subsidies can be provided by the public player on a case by case purpose.
My Sketchup open project sources
various projects rolled up: http://dl.dropbox.com/u/17111233/Roll_up.rar

Colour safe chart:

jamespetts

Sdog - why do you say that it is beyond what can be achieved? Can you be more specific?
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

sdog

the public sector works on completely different principles than private sector.

Firstly the source of funds depends on completely other factors, it doesn't generate income in the game. Then, it's decissions are politically motivated. If the public sector wants to raise land, and build a bridge to nowhere, they can do it. (perhaps to boost construction sector, corruption, or having a hamlet with strategically important voters [or a rotten bourough in the early game], or to have a part of society participate). Thus it does not necessarily follow economic principles but sometimes solidarity.

That is for example building a bridge to allow remote vilagers to get to work and hospitals. which is so expensive the tax they pay will never cover it. Economically it would be suicide, in a wider context it might be completely reasonable. There are defense considerations too, having loyal population at a remote strategically important place might be worth quite an investment. (i won't mention the islands i have on my mind ...)

With all those factors influencing decissions, and a practically unlimited source of finances, public player decissions could be completely arbitrary. There's a different class of games that does that, starting with simcity, or tropico ending at those swedish nation simulations (europa universalis or so?). It's also quite a bit more complex than economy.

There is also much less degrees of freedom in making decissions, as a CEO of a private sector company has or rather had in the past. The public player would thus become a game "build a road from A to B using exactly the route [...]"


(just a thought that came while writing this, did you think about value of land? with economic factors influencing it? using costs for houses doesn't do the trick only very roughly [one can't sell land])




edit:
i think i should conclude: investing time to improve where simutrans experimental could be very interesting, that is an economic transport game, instead of starting another thread of development would be much wiser. More so as this would also reduce the usability of the public player that is required for moderating network games.

jamespetts

May I ask - what public player tools do you think are necessary for moderating a network game? So far, I have not had to do much other than build/delete roads on the bridgewater-brunel server.

As to the economic considerations - planned for the future (although not a priority) is to allow the public service player to generate income by raising tax (on players, partly, and also on cities, which would reduce their population growth). The public service player would act as the Simunation's ministry of transport, rather than government generally, and thus have available to it a rather limited budget (for these purposes, the revenue from the tax on transport operators, plus a small proportion of whatever tax is levied on the general population). Its aim in game terms would simply be to maximise growth using the tools available to it (which is why we need to restrict the increase/decrease population buttons, or else increasing growth would be a somewhat trivial task). That would need only relatively minor changes to the code.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

ӔO

if you are thinking about taxes, I would recommend a progressive tax rate.
My Sketchup open project sources
various projects rolled up: http://dl.dropbox.com/u/17111233/Roll_up.rar

Colour safe chart:

jamespetts

That's an interesting idea, although could complicate things somewhat.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

ӔO

without getting too political, just by looking at what the players have in their bank, say 10% out of Lindley would be quite bad, while 10% out of mine doesn't hurt at all.
My Sketchup open project sources
various projects rolled up: http://dl.dropbox.com/u/17111233/Roll_up.rar

Colour safe chart:

jamespetts

There may be something to be said for that - but implementation is the key!
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

sdog

"So far, I have not had to do much other than build/delete roads on the bridgewater-brunel server."
most of that work seems -- an impression from reading this thread -- to stem from your decission not to allow players to delete public roads.

Sometimes the need arises to fix the havoc an abusive player causes. There is also the possibility of a player acting as game master in a more controlled game. Here you could ask moblet, he has a rather interesting set of rules and moderated game on one of Timothy's servers. (I've never joined it, but, as an example there are public motorways)

The public player sometimes needs to remove infrastructure of a player, who abandoned the game, but has a company still running.

It might also be necessary to make larger changes to level the playing field for players who joined later. Eg, if there is a narrow isthmus between two continents, but the area is completely used by old players, widening the isthmus may be such an operation. Building a public way another, or even replacing player owned infrastructure with public infrastructure. Same for making larger cities more accessible for competitors.

It is not so much a question here if you would do it yourself, but that a game-master/admin could do it (can't be sued for it after all :-)

Imaginable would be a setting similar to the scenarios Severous builds.


In the course of a single player game, i use the public player also frequently. For all sorts of operations. This is sometimes building and linking of factories (to make a game more interesting). Beautifying cities, or improving their growth. In early game overcomming obstacles i consider unfair, but to expensive to fix as player. (eg. having dug too deeply, by accident; using workarounds for things that can't be achieved easily in-game, but would not be excruciatingly expensive in real life.)

Quite often i use public player to straighten out diagonal roads close to cities (they don't grow well with those) and fix silly inter-city roads.

also frequent is fixing river sections i accidentally deleted.

(should this and my previous post perhaps be cut-off and put in a different thread, i'm really far off-topic now)

Milko

Hello James

Quote from: jamespetts on January 28, 2012, 05:57:35 PM
Ahh - it is indeed possible that I have reduced it twice by mistake. Try the latest commit and see whether that seems correct.

The ship, completely full, has an income per km (estimated) of 8.3 € / km. In my opinion a good parity between costs and revenues can be found with filling equal to 75%, so at a cost of about 6.2 € / km (the last commit cost is 4.1 € / km).

Giuseppe


dannyman

#29
Quote from: AP on January 28, 2012, 10:38:39 AM
What is needed is a mechanism to cause successful players with runaway bank balances to have to spend more than unsuccessful players with low bank balances. I suggested in another thread that costs of making changes to infrastructure might be proportional to the number of convoys using the tile each month, for instance - that's just one idea (and not an unrealistic one).

You could have the government institute a graduated tax rate: the higher your profits the more money the government wants to skim off the top, whereas if you are losing money, there's nothing to tax, and the government may decide it is in the public interest to grant low-interest loans or an outright subsidy in the interests of economic development.

If you really wanted to go down this road, a smart government would try to tie incentives with deliverables, or as we call them in video games, scenarios.  "The government will grant you $300k grant in order to build a network connecting 12 cities.  If you have not completed this obligation in three years time, you'll owe the money back plus hefty interest.  Do you accept?"


I also note a tension between the public player acting as a would-be government and the public player being a would-be God mode.  I'd say make it configurable, and then down the road make the existing public player "God mode" and make a "government player" that can do stuff based on taxes.  In a multi-player game, the government player could possibly be elected at regular intervals, maybe favoring a player who has gone out of business: the government then has to balance tax policies with delivering infrastructure needs in a fair manner in order to get re-elected.  This political aspect would encourage players to cooperate nicely with each other, though in some games you'd have a tit-for-tat bloodbath.


My existing practice is the public player is used for infrastructure projects within or between cities that might typically be undertaken by the government, and, for example, to clean up after in-game bugs. :)

-d

ӔO


       
  • LBSCR A1 terrier, maintenance too cheap when compared to GWR 517
  • Midland 700, maintenance too expensive when compared to LNWR 17in
  • LNWR Precedent, maintenance too expensive when compared to Midland 156 and GNR Stirling Single
My Sketchup open project sources
various projects rolled up: http://dl.dropbox.com/u/17111233/Roll_up.rar

Colour safe chart:

wlindley

Narrow gauge ways should cost significantly less to build (a little more than half), and slightly less to maintain, than regular gauge track.  For the 100km/h narrow track, I imagine a cost of 100.00 (150.00 now) and a maintenance of 10.00 (comparable standard track maintenance 14.40).  Am really hoping the narrow gauge bridge makes it into the next release.

jamespetts

Quote from: wlindley on February 03, 2012, 11:35:52 PM
Narrow gauge ways should cost significantly less to build (a little more than half), and slightly less to maintain, than regular gauge track.

May I ask - what is the basis of these figures?
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

Isaac Eiland-Hall

Just a quick note: If the idea of having a 'government' player takes off, I would recommend splitting 'public service' into "Server Administrator" and the Government player.

And this sounds really interesting to me.

I frankly get more fun out of building freeways than I do playing the game. On my list of things I want to do is to run a server where I build freeways and public rail and such... :)

jamespetts

If you like building roads, have a look at this thread under the title, "traffic simulation". Not currently a high priority, but I'd love to do it one day...
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.