News:

SimuTranslator
Make Simutrans speak your language.

OpenPak128 Economy & Balancing

Started by Zeno, September 04, 2011, 10:22:52 PM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

Zeno

I've been thinking all around the economy system of the pakset, and I've decided to investigate in the following direction with the goal of making balancing easier:

1) Drastic reduction of bonuses: perishable goods and pax will have 4%, post & goods 3%, others 2% or 1%; bulks and similar 0%. Before they went from 0 to 18%.
2) Remove bonus calculations for vehicle balancing: as long as it's a smaller bonus now, it won't have a great impact on vehicle RC.
3) Create a "clean" new excel sheet for these "simplified" calculations

Actually I'm at step 2, having tested trams (results were satisfactory) and will test road vehicles now.
If you have suggestions, ideas or comments, please feel free to post them here.

VS


My projects... Tools for messing with Simutrans graphics. Graphic archive - templates and some other stuff for painters. Development logs for most recent information on what is going on. And of course pak128!

Zeno

I hope this bonus reduction will also reduce the problem of modern airplanes earning zillions compared to old ones, same happened with high speed trains. The problem of a big bonus is that you get a huge (I mean exponential) difference between profits at different speeds. That makes almost impossible to know the theoretical profit of a vehicle (well, you always have empirics :S), or at least makes your calculations extremely sensible to error (it will multiply the possible error by 18% per each km/h over the "bonus" speed!!!).

My point of view is: imagine two vehicles transporting same amount of passengers, one at 100 km/h the another at 200 km/h. With 0% bonus they would get paid the same for each travel, but the fastest one would take enough advantage by traveling at double speed (aprox. half of time), so in 1 year it could take up to double the passengers transported thus much higher profit.

Anyway my idea is to leave that 4% just to give a little bonus to fastest vehicles over slower ones.

Vonjo

#3
I have an idea:
Automatically set price by makeobj or DAT generator. That is create a special costum build makeobj or DAT generator for pak128, witch automatically calculate the price based on capacity, power, speed, etc. So, anyone create vehicles, track, etc doesn't have to think about its price. And it will always be balanced, no excel required.

sdog

in a lengthy discussion of James and moblet on the economic model*, one of the conclusions was that the speedbonus system is not economically sound at all. It appears to have been introduced as an incentive to buy new vehicles instead of having old ones run indefinitely, and later missinterpreted as a bonus system for prompt delivery.

It's perhaps best to drop speed boni alltogether in the pak.


*The discussion was in regard to experimental, but should still apply here. Experimental has some elements built in that cause the desired effect too (but might be too heavy for standard)
quick routes attract more pax, too long journey times give refund to pax (player loses money earned already), very high obsolescence cost for vehicles. [this just as additional information to put the above said in context]

VS

I suspect that Frank has done some work in this area and concluded pretty much the same - bonus only makes things harder to calculate for authors.

My projects... Tools for messing with Simutrans graphics. Graphic archive - templates and some other stuff for painters. Development logs for most recent information on what is going on. And of course pak128!

Václav

Quote from: Zeno on September 04, 2011, 10:22:52 PM
I've been thinking all around the economy system of the pakset, and I've decided to investigate in the following direction with the goal of making balancing easier:

1) Drastic reduction of bonuses: perishable goods and pax will have 4%, post & goods 3%, others 2% or 1%; bulks and similar 0%. Before they went from 0 to 18%.
2) Remove bonus calculations for vehicle balancing: as long as it's a smaller bonus now, it won't have a great impact on vehicle RC.
3) Create a "clean" new excel sheet for these "simplified" calculations

Actually I'm at step 2, having tested trams (results were satisfactory) and will test road vehicles now.
If you have suggestions, ideas or comments, please feel free to post them here.
I am sorry if it woould sound bad but I am worried about that it will push people to create more really cheating (and maybe also unreal) objects (it means with minimum running costs) - because to make profit with aircraft and not only with it will be unaccessible - without any special addons that will bring profit back to game.

I use rule If it is not prohibited, then it is legal but not much.

Currently some planes are very crazily ballanced (mostly in value and also running costs) and make profit with them is at least very hard. So, please, correct this at first - and then return profit to aircraft. I think that aircraft should make more money than other ways. And that some new planes make more money than previous? It is right.

Chybami se člověk učí - ale někteří lidé jsou nepoučitelní

Zeno

@VaclavMacurek: that's why I leave that 4% bonus; airplane profit will be still big, but it won't be *huge* like now.

@sdog/VS: I wanted to keep that 4% just as incentive, not being used for calculations (that means, faster vehicles should have a small extra profit, but yet I'm not sure if 4% will be fine)

Also I'm wondering if it would be good to lower a little bit some infrastructure maintenance costs, but anyway that can be applied later on if needed. That would help to lower a little the difficulty level of the pakset, don't you think so?

Václav

And by the way, if I wrote about crazy ballancing of planes, I thought mainly too high price of some ones in comparison with some else - for example about half million against fifty thousand and so on.

Chybami se člověk učí - ale někteří lidé jsou nepoučitelní

Fabio

At least regarding pax service, speed matters for ticketing price, to a certain extent.
E.g. a Turin-Naples (900 km) has two railways connections:
- ordinary intercity connection: 9:30 hours at 63 EUR
- high speed link: 6:30 hours at 132 EUR
- a direct flight between the two cities takes 1:30 hours and costs 134 EUR with the national carrier, 90 EUR with a low cost carrier

Zeno

Sure fabio. Faster service, higher ticket price (and higher RC!!!), but not necessarily higher benefit: the high link may have more expensive tickets, but maybe with lower benefit; anyway faster speed means more trips, and even with lower benefit per trip could result in higher total benefit... just thoughts :)

sdog

That higher ticket prices for high speed links would be nice to be reflected in the game. But this would require an extension to the engine. It does not really work with the current implementation of speed boni. Zeno's approach of a very small bonus might work as a good compromise though.

Zeno this approach:
QuoteFaster service, higher ticket price (and higher RC!!!), but not necessarily higher benefit: the high link may have more expensive tickets, but maybe with lower benefit; anyway faster speed means more trips, and even with lower benefit per trip could result in higher total benefit
doesn't really solve fabios dilema either, you forget that you might run a fast train completely empty and then don't compensate the higher costs through higher ticket prices. Same for fast train moving on a slow track.


trashhead

You can resolve fabios dilemma in two ways, leaving the speed bonus and making faster vehicles to have a higher maintenence and higher running cost. Then again you will have to change all the costs for all vehicles and to look for not to install cheating vehicles in the pakset. And the other way, is if ticket price is too high, there wont be too many passengers as zeno said, so it should have a restriction on the number of passengers that travels in that vehicle, so less people will arrive in the station in comparison with station of less expensive transports, in the same way a factory would prefer to deliver their goods using the faster vehicle only when its necessary, paying a higher transport cost only when its needed. This second could be done in experimental, though.

So Zeno approach in reduce the speed bonus its the easiest way, and it will work fine....i guess... ;D

greenling

sorry the OpenPak128 Economy & Balancing it fully wrong!
i Have be look on some Vehicles and there hat i be get a scratchattac!
Then you want be creat a new OpenPak128 Economy & Balancing then must you exchangen some loadingcapaction !
Opening hours 20:00 - 23:00
(In Night from friday on saturday and saturday on sunday it possibly that i be keep longer in Forum.)
I am The Assistant from Pakfilearcheologist!
Working on a big Problem!

sdog

i'm surprised you care for balance at all greenling, i thought you put everything you can find for 128 size together?

Bughu Baas

#15
It's the first time someone tells me that the bonus is *not* for a fast delivery, but for the intend to scrap old vehicles.. :o

Then, why not implement the simutrans experimental (9.12) vehicle maintenance system into normal simutrans? If you use some old movable traffic obstacles, you have to pay for it ;-)

(Nevertheless, the balancing in PAK128 is gruesome...)

trashhead

Ok, the question is: how can i help you Zeno?

Zeno

Mostly testing. I will post a first set of vehicles with new stats very soon, probably next week. By testing it and telling which some vehicles are more profitable, if there are some vehicles which can not have any profit, and such things.

Other ways to help is to share your ideas on balancing, although I'm basing my job in Napik's original excel balancing sheet, which is in the SVN and allows us to have a point to start at.

In any case, thanks for your aim :)

sdog

haven't seen that excel sheet in the svn, where is it?

i've started a bit of my own balancing:
http://forum.simutrans.com/index.php?topic=7981.msg76002#new

PatrickN

I quit Simutrans several years ago.  I have always blamed the speed bonus, but there are other problems.  Every six months I download the latest stable game and pak128 and try it again, but it always ends with me being disappointed. ::'(

My observations:  Maintenance costs are too high for most of the vehicles, and the price paid for delivery of goods is rarely enough to cover the maintenance costs.  Passengers are the only thing I can find that actually makes a profit.  Every industry I've tried to connect operated at a loss unless I cheated and used public mode to build roads.  Trains, planes, and most of the ships are useless.  Canals are money pits and only to be used by people who like the color red.  When I start a game, I have no choice but to use Public mode to make my own industry chains because if I let the game make them at start up, half of the industries were missing factories because the elevations needed for trees, sawmills, and some others don't exist.  I either have to use Public mode to make hills tall enough for those industries, or I have to override elevation limits to place them wherever I want.

The worst part is that since maintenance costs so much that I never make much money, so when the speed bonus starts cutting down my profits, I go red fast and can't afford to replace old trucks with faster ones. :-[

My suggestion: Remove speed bonus completely and double or triple the payout for all goods.

Painter, in and out of retirement.

sdog

PatrickN, how small are your maps typically?

PatrickN

I've always used 128x128, but I see now that the default is 256x256, and that works better.

One problem somewhat solved. ;)

Painter, in and out of retirement.

Zeno

Hi PatrickN, thanks for your comments (and nice to see your here!). I'd like to give some answers to your suggestions:
1) Planes are by far the most profitable vehicles in the game, due to their bonuses (and it's the main reason because why I wanted to drop bonuses). It's easy to see a plane earning more than a million per year if you have pax enough to fill it.
2) Ships can be very useful, but it absolutely depends on the map (e.g.: on an islands map they become key), and they are mostly profitable
3) Passengers are the only profitable good: this is true, again the problem of bonuses.
4) Impossible to make profit --> You can always use the 1.5 multiplier and you'll have the pocket full of bucks ;)

Solutions: the bonus drop shoud fix points 1, 3 and 4, that's the main reason for me to think of a new full economy review.
About industry distribution, it's mostly a matter of the game engine, nothing to do with the pak, so I won't be able to help with this sorry.
Thanks again for your comments :)

greenling

I do in the evening be look in the Database from openpak128 and then do i make here puplic what we must be change in the openpak128!


Opening hours 20:00 - 23:00
(In Night from friday on saturday and saturday on sunday it possibly that i be keep longer in Forum.)
I am The Assistant from Pakfilearcheologist!
Working on a big Problem!

sdog

PatrickN,
the issue was most likely the small map size. Industry will have difficulties finding a proper spot for two reasons:
- Unless the map is very rough it is unlikely it has all climate zones in such a compact size.
- Every industry has an exclusion zone around it, preventing new industry to form too close nearby (you can try to experiment a bit with industry spacing in the settings.)

Pak128 is also quite difficult on a small map. This is due to the lack of the network effect. Your network becomes more profitable the more entities you have in it.
Station infrastructure is quite expensive in pak128, but tracks not. If you have many stations at short distance you have a hard time to get back the infrastructure costs.

If you have difficulties running a large map on your computer try a very oblong map. 196*512, or so. With a few long distance bulk routes you can easily get money. (especially if you can load trains on roundtrips.) You can also turn of trees to improve performance.


PatrickN

Map size isn't an issue anymore.  It was on my old laptop, which is why I always went with a 128x128 map size.  I went to a larger size (I think it is 512x512) and after I played with the settings a little bit as far as elevations for ground types, the industry chains are all there.  After playing a little bit, I am making a lot of money from passenger bus lines, and my trains are making money as long as I use a small number of stations and only with long distances.  I'm only five years in, though, so I haven't started seeing the effects of the speed bonus yet.

I was my biggest problem. :-[

Painter, in and out of retirement.

VS

The problem with climates is serious, though; not everyone likes to play with all of them. I guess the way out is making graphics that expect blending with (back)ground, and more variety. Tree plantations could get more climates easily with new fields etc...

BTW: I have always used sdog's solution, long but narrow maps.

My projects... Tools for messing with Simutrans graphics. Graphic archive - templates and some other stuff for painters. Development logs for most recent information on what is going on. And of course pak128!

Fabio

Quote from: VS on September 19, 2011, 09:30:33 AM
The problem with climates is serious, though; not everyone likes to play with all of them. I guess the way out is making graphics that expect blending with (back)ground, and more variety. Tree plantations could get more climates easily with new fields etc...

Industries and climate discussion moved here: http://forum.simutrans.com/index.php?topic=6267.msg76646#msg76646

trashhead

after playing pak128 for a while i personally believe IMHO that the production capacity and maintenance costs are the main problem for industry transport.

for example, I build a network with a food chain to a supermarket and in five years the food factory was completely full, meaning that i have to sell my grain trains and change it for only one truck!, the same for meat and fish, the only one with i could make a steady profit its with the food going to supermarket. I was unable to recover my investment in five years, since ten years it is normally the payback calculation.

The same happenned with coal, i have to put a single train for three coal mines and one thermal plant since the coal production wasn't fast enough to replenish the coal in the station, making  my train hang a long time and with maintenance costs that high, is unable to make a profit, even when my train just have only three coal wagons.

So the conclusion is that only when two connected industries are far enough you can expect to make profit, and that is not real, because in real world you would prefer bring your needs from the shortest distant, not the longest.

sorry for my bad English

sdog

balancing out demand and supply is one of the main points of interest in this pak, this is meant to be difficult.

if you struggle with this, you might want to try pak64, it has much higher overall production rates, and is a bit easier to make a proffit. (the difficulty lies in effectivenes, coping with the high demand for transport)

Zeno

Quote from: trashhead on September 28, 2011, 09:21:30 PM
after playing pak128 for a while i personally believe IMHO that the production capacity and maintenance costs are the main problem for industry transport.
Production and maintenance are a though feature, rather than a problem.

Quote from: trashhead on September 28, 2011, 09:21:30 PM
for example, I build a network with a food chain to a supermarket and in five years the food factory was completely full, meaning that i have to sell my grain trains and change it for only one truck!, the same for meat and fish, the only one with i could make a steady profit its with the food going to supermarket. I was unable to recover my investment in five years, since ten years it is normally the payback calculation.
You should use cheapest tracks and small stations, and this is specially mandatory at the very beginning of the game. Infrastructure maintenance usually makes the difference between profit and bankrupt. Later on, with some bucks in your pocket, you'll able to build more advanced facilities and networks.

Btw, we all know the pak is a bit though, and difficulty is to be eased a little bit with the next balancing updates. Just keep an eye on this thread ;)

Erik

I like the chancing speed bonus gone away. (Or at least least less prominent.)
I've also to often that a vehicle isn't profitable any more before I have enough to get new vehicles.

About the problem of the people could use vehicles indefinitely.
Look to the real world.
What is a reason for a company to replace his vehicles?
1. Rising maintenance costs.
2. Can go faster.
3. Stronger engine to pull more and/or accelerate faster.
4. ...

The latest two points are already relevant to Simutrans.
1. There are some development on experimental about this.

2. Because the vehicle can go faster, it can transport more. Of course this is relevant when there are enough to transport. Also with trains, a slower train gets in the way for a faster train. When it happens to often I will replace it by a faster train.

3. I've it often enough with bus lines. The oldest bus on the line cause traffic jams because it accelerate to slow and can't take the fast and busy traffic.


greenling

Erik That´s it very heavy to declare it.
Opening hours 20:00 - 23:00
(In Night from friday on saturday and saturday on sunday it possibly that i be keep longer in Forum.)
I am The Assistant from Pakfilearcheologist!
Working on a big Problem!

Zeno

@Erik: I agree mostly with what you say, but none of these factors can be changed by balancing; they compete to development instead.

prissi

Just a note about speed bonuses. Next version in standrad will base the speed bonus one actually speed allowed by the route and the weight of the convoi. As such, it would make sense to have some 50 km/h buses in later game years.