Defining a new economic system for future versions

[Breakable]

...
With that, what would there be left to do for an unspecific economy simulator?

The amount of goods is not everything. It depends how complex you want the system to be, but I see other things that also matter:
Price,
Production halt (Strike, Workers dead, Revolution),
Creation/Destruction of sources/consumers,
Account management,
Business AI,
Workers,
other?

The minimum addition is Price, all the others are optional, or can be left for later.

So a basic economic simulator would have to have:
1)Goods management
2)Price management

This seems quite simple to implement

[Fade]

Excuse me but english isn't my native language so I can't write what I precisly want to say. I'll try to make my point more accuratly below.


Every component can be reduced by size complexity to adapt a solo game. Some people just have fun with a light economic system. In the multiplayer perspective, you may have economic wars the player can be part of it (even initiate it). You may also have empire/corporation/single NPC doing automatics tasks. This may need economc simulation if player realy want to have fun with this side of the game.

Here is something to think about :
ES - PC : like a physical engine, produce price computation based on offer and demand themselves bases on resources, consumers, risks, stress, cultures, media information, etc. Like a nervous systems, produce virtual needs. This is very complex to implement. But a economic entity could be described like this :

Type : Planet
Culture : A
Population : 10000000
Produce : Food (X unit/std year)
Consume : Default (means automatic needs computation based on other carateristics)
Organisation : Dictature
Techlevel : middle
etc

NB : you may push the limit of the simulation very far, for exemple, self modifying data, auto civilians war (for exemple when there is not enough food, increasing demands for arms, etc).

ES - IA : Take needs and compute actions based on factions, planet (rebelions ?), and other economics entities and issus needs.

ES - MR : This is the Mission Runner. Its mission is to fulfil a need by issuing missions and report missions consequences.
First, it will choose if a mission has to be run by VS. If something have to be done where no player is present, a dice roll (success/failure(piracy/lost)) may do the trick. ES - MR will then tell ES - IA about it and tell to ES - PC about stock modifications. This information would be issued after a mission success, even if it's not a ES-IA mission, VS could put missions results and player "out of the blue" initiatives can be treated like missions results.

I hope you'll understand everything

[lee]

...
With that, what would there be left to do for an unspecific economy simulator?

The amount of goods is not everything. It depends how complex you want the system to be, but I see other things that also matter:
Price,
Production halt (Strike, Workers dead, Revolution),
Creation/Destruction of sources/consumers,
Account management,
Business AI,
Workers,
other?

Ok, but how would you implement these independently from any specific application (game)?

Production halts would have to have to do with events in the game as well as the creation and destruction of sources and consumers. What the others are for, I'm not sure.

The minimum addition is Price, all the others are optional, or can be left for later.

Yes, but prices are also linked to the specific application because you need to deal with the factors the application provides.

So a basic economic simulator would have to have:
1)Goods management
2)Price management

This seems quite simple to implement

Well, what exactly is goods management and price management?

Prices can be calculated going by what it costs to produce something and by how much profit can be made. That depends on the factors specific to a particular application.

What goods management is also depends on the application. In one case (like in widelands or settlers II), it's producing goods at places and storing goods in warehouses and using many carriers to transport it across roads in single units without cost. In another case, it's producing and storing goods at places and transporting them to and from other places in containers carried by starships which would cost.

In both cases, you have to keep track of places, goods and transportation --- that's a lot of data, eventually. You could make use of a database in both cases. You could consider a DBMS as an economic simulator. But instead of adapting the simulator to a particular application, the application would use functions the simulator provides. The particular ruleset of the economy to simulate is specific to the application.

[lee]

This may need economc simulation if player realy want to have fun with this side of the game.

The game needs a good economy simulation if players are to have fun with it.

Here is something to think about :
ES - PC : like a physical engine, produce price computation based on offer and demand themselves bases on resources, consumers, risks, stress, cultures, media information, etc.

If you put factors like this into the simulation, you also have to give players ways and means to influence these factors. They might have TV stations or studios to produce movies and shows.

But I don't think it needs to be so complicated. A simple economy with true production, consumption and transportation --- using the already available goods, places and ships, eventually extended by goods needed to get a complete chain (what are ship hulls made of?) --- would be a great step forward. Once that is implemented, allow the player to have his own stations and give him good ways to automate things. Then it's his choice if he wants to fly around in a Plowshare or an Ox to do some trade, if he wants to be a mercenary or if he wants to build his own empire and conquer the universe. Faction AIs would take care of what the other factions do. Flying a ship should feel like flying one (Take a look at btrl --- not very realistic maybe, but it feels like flying. Unfortunately, you'll always will run out of missions and then the game is over.). That would make for a great game.

[Fade]

Ok, but how would you implement these independently from any specific application (game)?

You can implement independtly the engine, but the game universe has to be translated by rules.

Where are the resources ? What does it take to build 100T of goodies ? Where are the consumers ? What are the cultures ? Who needs arms ? Who has the technology. This is game specifics.

[Fade]

Of course, if you want the players to do what IA do, you'll have to deal with crafting

[lee]

Ok, but how would you implement these independently from any specific application (game)?

You can implement independtly the engine, but the game universe has to be translated by rules.

Where are the resources ? What does it take to build 100T of goodies ? Where are the consumers ? What are the cultures ? Who needs arms ? Who has the technology. This is game specifics.

Sorry, but I don't understand what the engine is supposed to do: Read a ruleset from a database, apply the ruleset on game specific data stored in a database and update the data accordingly, while the game reads the data from the database as needed?

That's probably possible, but appears still quite specific to a particular game. And how would the engine interact with other parts/aspects of the game, for example an AI that tries to figure out if more places producing something should be created or if something else needs to be done? Counting the number of such places would be easy, but how would the AI decide if the production of something is insufficient not because the number of production places is too low but because (other) resources are not being delivered or not produced in sufficient amounts? To decide that, the AI would become dependent on its ability to interpret the ruleset the economy engine is using ...

[Breakable]

There no need for the Economics system to be too complex and to try to figure out the game rules.

As I mentioned before the minimum requirement would be price and goods management. For this amount of complexity there would be no need for an AI. Just set the prices to reflect demand/supply, and the rest should be done by the game's faction AI.

Well if we try to implement a full featured economics simulator, with capability to create new sources of goods and revenue, then some sort of collaboration with the game system is required, but I think it can still be implemented on independent basis. For example, if you see that demands for some goods is really high and you want to build a new factory, you should ask the game system : "build factory for type of good", and then the game's system would be responsible for finding an appropriate location, building, protection, etc.

It even does not matter much if the system is not very effective, because there is no perfect AI in any game currently, and it still fun to play to beat it. Actually if the AI was too good that would be a problem for playability.

[lee]

As I mentioned before the minimum requirement would be price and goods management. For this amount of complexity there would be no need for an AI. Just set the prices to reflect demand/supply, and the rest should be done by the game's faction AI.

The AI I mentioned would be part of the game, not of the simulator.

Well if we try to implement a full featured economics simulator, with capability to create new sources of goods and revenue, then some sort of collaboration with the game system is required, but I think it can still be implemented on independent basis. For example, if you see that demands for some goods is really high and you want to build a new factory, you should ask the game system : "build factory for type of good", and then the game's system would be responsible for finding an appropriate location, building, protection, etc.

You could make it the other way round by having a game AI as part of the economy simulator, but a game AI would have to be game specific. And when the game doesn't know anything about the economy, how could it find an appropriate location to create a place of production? It couldn't even build the place because building something takes resources, and the game won't know about that.

It even does not matter much if the system is not very effective, because there is no perfect AI in any game currently, and it still fun to play to beat it. Actually if the AI was too good that would be a problem for playability.

Economy doesn't need an AI, but AIs need economy. And they better be pretty good because it gets boring quickly if they suck.

If players are to have the means to take over the universe, they will need to be able to make use of automation. That automation would be at least as smart as any of the AIs, so it would be very difficult for the players to loose.

[Fade]

Sorry, but I don't understand what the engine is supposed to do: Read a ruleset from a database, apply the ruleset on game specific data stored in a database and update the data accordingly, while the game reads the data from the database as needed?

That's probably possible, but appears still quite specific to a particular game.

Yes. I think the ruleset can be shared among any spaceopera game but the dataset is game specific. That's why the system you just described seems to be game sepecific.

And how would the engine interact with other parts/aspects of the game,

The engine compute prices. Prices, including transport and work prices, once given can be computed by AI in order to take actions to maximize profit. AI doesn't have to be a part of the economic simulator, but if it is, it will not be the price computer.

for example an AI that tries to figure out if more places producing something should be created or if something else needs to be done? Counting the number of such places would be easy, but how would the AI decide if the production of something is insufficient not because the number of production places is too low but because (other) resources are not being delivered or not produced in sufficient amounts? To decide that, the AI would become dependent on its ability to interpret the ruleset the economy engine is using ...

I think that any action that could be taken by player (now or in further game versions), like crafting, installing facilites, etc should be processed, of course, by a game dependent AI. So I don't see a problem.

Anyway, such things like creating new economic entity (factory, or other thing) could be human "GM" decisions. But that's only details about implementation.

What we want is that if a player comes to a system counting 1000 people with 10000T empty store, food price will be high, if it's full, food price will be low. What is the limit ? What is the real price ? Many things should be taken in consideration. That's the Price Computer job (called ES-PC in my previous post). It depends about consumtion, available stock, future stock predictions, risk management, etc). You don't need to have game specific variables to describe it, only game specific data. What are the costs of energy ? What is the real price of transport (including costs of lost stocks, etc).

[Breakable]

Economy doesn't need an AI, but AIs need economy. And they better be pretty good because it gets boring quickly if they suck.

If players are to have the means to take over the universe, they will need to be able to make use of automation. That automation would be at least as smart as any of the AIs, so it would be very difficult for the players to loose.

If you take a look at the best games out there, such as civilization, starcraft or fallout the AI's are really simple, even idiotic I would say. You can cheat them over, and that makes you enjoy the game more.

For example in starcraft a professional player can beat up to 7 ai players at once by using some exploiting tactics on a specific map.

You can also beat fallout 2 in less than 20 minutes, if you know what to do.

I think what is most important is to have a believable system, that is easy to implement, and enjoyable to play.

[Deus Siddis]

If you take a look at the OLDEST games out there, such as civilization, starcraft or fallout the AI's are really simple, even idiotic I would say. You can cheat them over, and that makes you enjoy the game more.

For example in starcraft a professional player can beat up to 7 ai players at once by using some exploiting tactics on a specific map.

You can also beat fallout 2 in less than 20 minutes, if you know what to do.

I think what is most important is to have a believable system, that is easy to implement, and enjoyable to play.

Fixed.

[Fade]

If you take a look at the OLDEST games out there, such as civilization, starcraft or fallout the AI's are really simple, even idiotic I would say. You can cheat them over, that's why you have to play on internet against human players to have fun.

For example in starcraft a professional player can beat up to 7 ai players at once by using some exploiting tactics on a specific map.

You can also beat fallout 2 in less than 20 minutes, if you know what to do.

I think what is most important is to have a believable system, that is easy to implement, and enjoyable to play.

I can't agree more. We have to look closely to economics. A little for the solo game, but for multiplayer persistent game, we have to really go far. Economic freeks are standing outhere waiting to build mass of money in order to bring down empires to their feet. They may succeed if we have a realistic (or very believable, at least, economic system) and they will fail because we'll also have powerfull IA to counter them. Right ?

[lee]

Yes. I think the ruleset can be shared among any spaceopera game but the dataset is game specific. That's why the system you just described seems to be game sepecific.

The ruleset would have to be game specific as well. Each game has its own things that are being produced, transported and consumed, and each game is using its own rates for that.

The engine compute prices. Prices, including transport and work prices

Prices are game specific as well. There are games that don't even use/have prices --- see widelands, for example.

I think that any action that could be taken by player (now or in further game versions), like crafting, installing facilites, etc should be processed, of course, by a game dependent AI. So I don't see a problem.

It doesn't take an AI for that. And it's game specific as well.

Anyway, such things like creating new economic entity (factory, or other thing) could be human "GM" decisions. But that's only details about implementation.

So you won't only be playing the game but also be a game master who constantly analyses a dozen or so different economies (one for each faction) throughout hundreds or more systems and manages the creation of production places for them at the same time?

You need to consider details of the implementation when thinking about an idea like this. Still the questions I've been asking haven't been answered ...

What we want is that if a player comes to a system counting 1000 people with 10000T empty store, food price will be high, if it's full, food price will be low. What is the limit ? What is the real price ?

There is no limit or real price. Any currency is worth exactly as much as the people using it believe it's worth, that's all to it. Actual numbers to start the simulation with can be set as a relationship, like setting the price for something very basic (like water) to some number (like 1 water costs 1 credit). You need to do that once for all substantial (as in: not produced from other goods) goods, and the production costs of everything else can be calculated from that, including the production costs of water. Then you run the simulation, and prices will evolve. One thing you will have to do is keeping a reasonable limit on the total amount of money available.

All of this is game specific.

It depends about consumtion, available stock, future stock predictions, risk management, etc).

You can only deal with what you have.

You don't need to have game specific variables to describe it, only game specific data. What are the costs of energy ? What is the real price of transport (including costs of lost stocks, etc).

Can you give an example for that?

There are no stocks, no risk management and no currency, no energy costs, no transport costs in some games. Even if you have all that, it's all game specific. In vegastrike, the only thing you have is currency --- and none of the other things.

It's probably sufficient, maybe necessary, to keep the economy rather simple because the universe is huge and has or could have a great number of units that are interacting with each other. The outcome could be quite interesting, and due to the large number of units and a pretty large number of different goods, the implementation of a simple economy already strongly suggests to use a database (like mysql) to maintain all the data that would be involved. I would go as far as to keep track of every container, but I don't see that a good economy simulation will ever be implemented anyway.

[primate]

Why would a llama even think to care about trading goods that get traded on a huge scale by the Ox and other massive freighters? The cost would be extremely high per unit for the llama compared to the Ox. Why would anyone buy such items in such small quantities when it's obviously they would either not need it at all or need massive amounts of it?.

Not even to mention that we dont use shuttles to transport between ships that are obviously too big to dock directly or land on a planet.

Regarding the viability of small scale trading ships, smaller ships may have the advantage of being better at blockade running (which would make trading more exciting for players that's for sure) and hence higher "siege" pricing. Larger ships are also likely to attract more attention, and thus require a higher military/protection overhead. Large ships cannot evade attackers as easily as a llama, so defense overhead is higher. Limiting the way big ships can dock would also help. A big ship should park nearby a planet or base and then smaller ships should make low profits as a sort of goods ferry.

Anyway sorry, bit off topic... great thread, someone needs to put the ideas down in the wiki

[Breakable]

Playing online is more like working online (if we are talking about MMORPG). You have to use a lot of 3'rd party tools and guides, read forums, buy stuff just to be effective, and not let your friends down. Where is joy of playing there? PVP is a little more reasonable, but still if winning is fun - you have to invest a lot. Of course there are advantages/disadvantages everywhere.

And yes, I like older (or you can call classical) games more. And new games does not seem to have better AI either.

If we go back to topic, so I still think you can implement an Economy system separately from the game engine, because if you are not able to then you are doing something wrong.

Regarding the viability of any system the proof is in the pudding of course. So please instead of empty arguments think how you can actually define a model that works. I think I'll spend some time on that.

[lee]

Regarding the viability of small scale trading ships, smaller ships may have the advantage of being better at blockade running (which would make trading more exciting for players that's for sure) and hence higher "siege" pricing. Larger ships are also likely to attract more attention, and thus require a higher military/protection overhead. Large ships cannot evade attackers as easily as a llama, so defense overhead is higher. Limiting the way big ships can dock would also help. A big ship should park nearby a planet or base and then smaller ships should make low profits as a sort of goods ferry.

Have you recently tried to outrun a pursuer in an Ilama? And large freighters would have effective defenses so that it would take a lot of small fighters or a capship to shoot one down.

Anyway, look at this thread: http://vegastrike.sourceforge.net/forum ... such+thing. One of the things it shows is that implementing an economy involves setting up more reasonable ship sizes.

[lee]

If you take a look at the best games out there, such as civilization, starcraft or fallout the AI's are really simple, even idiotic I would say. You can cheat them over, and that makes you enjoy the game more.

What's the point in playing when you can cheat to win or when you already know that you only need to do this and that and you will win (within like 20 minutes) for sure? It just makes a game boring; you might try it for a few times as long as something new is there to be discovered, but after that, you're done with it.

[lee]

PVP is a little more reasonable, but still if winning is fun - you have to invest a lot.

You could play btrl or FS2 online. PVP is just like deathmatch which you can play in many games online. DM is one of the most boring types of games for me, especially when playing online because the lag is too much a factor --- and what is to it anyway?

And yes, I like older (or you can call classical) games more. And new games does not seem to have better AI either.

Well, safemode posted good ideas about how to create faction AIs. If there were those and a good economy simulation and if one the things players can try to achieve in the game is taking over the universe, that could make for an awesome game. If the AIs would turn out not to be smart enough, they would have to be improved --- or other players would take their places when playing online (letting aside that a game could take too long).

If we go back to topic, so I still think you can implement an Economy system separately from the game engine, because if you are not able to then you are doing something wrong.

That you could do something doesn't mean that it could be done with reasonable effort, and it also doesn't mean that it is reasonable to do so. And that you do something wrong doesn't mean that you could do something else if you could do what you are doing not wrong.

So please instead of empty arguments think how you can actually define a model that works. I think I'll spend some time on that.

I'll rather wait until the developers have made some decisions towards that. Economy is easy to define, you can do it outside the game provided that you have access to all in-game data needed and that you can give the game (or other parts like AIs) access to all economy data as needed.

But before a model can be defined, there are decisions that need to be made: Will players be able to have their own stations, fleets, sectors etc.? Will it be possible to build new stations? Will the ship stats be revised? Is it allowed to come up with more cargo types if needed to make a production chain? Will there be more types of production places so that not too many cargo types will have to be produced at one type of place? How is cargo transported (In single pieces? In containers and eventually paracontainers? What's the size of the containers? How do you figure out how many containers fit into a particular ship?). Will there be population? Is there going to be a concept of "happyness" that can apply to units and to population? Is it an option to use a DMBS like mysql to keep track of the economy data? How are dormant systems to be dealt with (You can't just freeze the economy there. How are goods transported in dormant systems?)? How do you deal with systems that haven't been created yet (Is there some limit so that they can be created when a game begins? Or should they just left be empty and open for being developed during the game once "discovered"?)? What about asteroid mining bases, do they provide in inexhaustible supply, or will they be hollowed out at some time and then be destroyed, salvaged, turned into something else or abandoned? Will new asteroids be spawned or will there be a limited number of them?

And so on ... It would be fun to work on it, but whatever work you or I would put into it would very likely be for nothing. Currently, the game cannot even be saved, not with details. It's not even prepared for having an economy, and it might take another 20 years to implement one ...

[jorganos]

Hi there. Another newbie piping up.

I don't think we want to simulate the entire economy, but only the trade.

Trade (transportation of goods) occurs where local supply fails to satisfy local demand.

Production usually is a combination of various resources, and usually located where one of the most important resources is abundant and the rest can be transported (or established) conveniently.

Resources would be: infrastructure, energy, work force, one or more types of raw or preprocessed material.


Productivity can depend on: availability of resources, efficiency of resource use, downtimes.

Trade deals with the problem of the availability of resources, and may aid efficiency of resource use (e.g. by providing luxuries and recreation for the work force, or keeping downtimes short providing spare parts and replacements).


(Production can be run in different modes: high output with low efficiency of resource use, low resource consumption with longer production time (or greater investment in infrastructure or recycling), exceeding standard maintenance intervals for temporarily higher output risking longer downtimes, etc. - factions may use this to balance short term tactical or strategic benefits against greater maintenance or repair cost.

This could be used to alter an economic model reflecting a threat scenario (e.g. as part of a story-line). Alternative resources can be investigated and introduced, too.


Here is one way to make a small privateer merchant (feel) part of the ongoing economy, possibly without having to simulate it in detail:


X2 and X3 have a type of mission where a base is giving out "tickets" for a required commodity. (Usually explained by sabotaged deliveries, additional demand for the product, etc...) IMO this is a niche for freelancing merchants. There ought to be a financial penalty for not providing the contracted goods, too.

A contractor might subcontract parts of these tickets to smaller ventures, giving even beginning merchants a chance to participate in such an economy.

Sensibly, such a contract should read:
Provide at least 800 units of tungsten metal to hull plating facility X07 by <reasonable deadline>.
The facility will buy up to 1200 units for a price of <usual regional price plus small markup>.
Failure to deliver at least 800 units will incur a penalty of <missing volume multiplied by maximum regional price plus compensation for production downtime>.


Failure to provide the goods will cause another ticket to be launched, with a less reasonable deadline and maximum regional price, paid from that compensation...

A sufficiently high-ranking merchant might post subdivisions of such a contract in regions where the goods are produced, or even harvested (e.g. by mobile miners).


Such a model wouldn't require to keep track of every single unit of a commodity throughout the universe.

A threat indicator might reduce the number of supply lots that make it to the factories, thereby raising asking prices.

A rival using a locust scheme on some resource (buying up all the available goods) might raise resource prices. Such market imbalances may also lead to an over-abundance of goods, resulting in facilities having downtimes or shutting up for good.

One might try to simulate all of this, or one could simply decree a certain economic situation for a given area and modify prices accordingly.

[Fade]

Trade (transportation of goods) occurs where local supply fails to satisfy local demand.

That's not true. Trade occure when someone can make profit by selling goods (or working time). Even if the local supply fails to satisfy local demand.

It can even happen on first necessity goods, especially food. That's happening on earth right now and people in poor countries are dying because of that.

[safemode]

keep it in relation to the game. Trade as far as the game is concerned would only deal between two units. So we treat what goes on internally in a base or planet as a black box, all we can deal with is what happens on the outside. General functions will have to deal with internal fluctuations when appropriate, if we care to even get that minutely detailed. We might not have to if the rest of the trading is engaging and dynamic enough enough.

So when talking about the economic system of dreams and fancy, you have to deal with it in terms of what the game is and simulates and is plausible to simulate. I wouldn't go any smaller than the units as they currently exist when dealing with individual economic simulation. All other aspects of the economy have to come from this or come from general functions using this.

You could put forth the idea of higher level "units" like faction ai's, or commerce center's (special), etc, to help an economic system along, since those sit above units.

Getting into discussions about real economies without considering the limitations of the game is just wasting your time. The idea of the thread is to figure out how to bend reality around the limitations to produce something that functions in all the ways you think it should function.

[loki1950]

Boiling that down we have to code this in a reasonable time frame and not have it take over.

Enjoy the Choice

[Deus Siddis]

Boiling that down we have to code this in a reasonable time frame and not have it take over the galaxy in a sea of economic imperialism. So we should make the ISO become ruler of the confederation in first half hour of gamplay and then there is no need for merchant pigs and an economy built by greedy capitalist dogs. Players will mine, move and fight for glorious motherworld or be exiled to an ice planet to play repetitive tetris-clone gameplay for the rest of their lives or until they load a previous save game.

Forget the Choice

I agree whole heartedly comrade!


Seriously, asteriod, rock planet, gas giant and nebula mining, establishing and protecting colonies/stations, fighting epic battles, exploring planets, all of these things are so much more interesting than 'space trade' which isn't even that realistic anyway, even with the jump network. So we shouldn't get too worried about this one aspect of the game, it just has to require some kind of skill and be balanced.

[lee]

The idea of the thread is to figure out how to bend reality around the limitations to produce something that functions in all the ways you think it should function.

It should function as an actual simulation: with a production chain so that nothing comes out of nothing, with consumption and with transportation and trade of goods. If you save the game and then reload it, it must be restored to the same state it was in when it was saved.

It doesn't have to be complicated, just simple static rules like "place type X produces good Y out of these goods within this much time and consumes that much of those goods". If you assume population (other than generically as in saying "it's the population that consumes") , that would already make it more complicated.

The point is actually having production and consumption by given rules. It's "having the real thing". The point is not making it more complicated than needed for that.