Civilization 4 - Abstract Trade and Commerce Models

[Fosse]

I know that many people here want detailed and incredibly functional trade and commerce models. Civ has, however, been a game of many high level abstractions, and it stands to reason we might not get to (or have to, depending on your perspective) control every last coin of currency or ignot of iron that trades hands.

Since that possibilty exists, I think it might be fun to discuss how we would like trade and commerce represented in Civ 4 assuming we will have little or no more control of it than we have in the past.

I'll get things started by submitting for discussion a little system I thought about on the way home from work today. One clarification: I am using the word trade to mean exchange of resources, gold, technology, and other valuables as we currently see in Civ games. I use commerce to mean the money generated by your empire that you get to use (ie, trade arrows).


Fosse's Commerce Proposal

I would like to see the ridiculous-when-you-think-about-it system of worked tiles generating commerce done away with. To Civ 3, people find gold on the ground, and can find more of it when someone builds a road.

I propose that no commerce be generated by tiles. Insead, the number of citizens in each city will help to determine that city's "commerce potential." This potential would represent the exportable goods the city generates as its people go about their life.

My next step is to make every city engage in commerce with every other city by default. This means that the Romans are ALWAYS trading with Carthage. The citizens of Athens are constantly bringing Greek goods to Persia, etc. This constant activity generates gold for the player based on the commerce potential of the two cities, and that gold represents the government's cut (tarriffs and taxes and whatever else).

Basically, this means that SMAC's system of bases engaging in commerce with bases when under treaties or pacts would always be in place. You can think of your empire as having constant trade routes with everyone else. But what is the player's role in increasing commerce? And how will roads help?


I think that the amount of commerce generated for each party should be increased for the distance between the two cities, and decreased by the travel time ( in movement points) to get from one to the other.

That means that, until roads are build, you can only enage in commerce with near neighbors. Building roads to those neighbors makes the trading more lucrative for both parties. Undertaking huge road building efforts to connect, say, Europe and China, will result in lucrative routes.

I think that this would simulate goods from distant lands being rarer, and thus more valuable, while accounting for the costs of transportation. Some cities will be so distant from each other that the travel penalty will negate the commercial potential until rails and fast ships, and so on.



That's it, in a nutshell. I haven't touched on trading resources or techs, or what have you, and I wouldn't even know where to begin coming up with functional formulae for all of the results. This just struck me as a neat way to model trade.

Some miscellaneous points:
Civs could embargoe other civs, reducing commerce both ways. This, and smuggling of goods, could be representd by using a double movement point cost. So it's easier to effectivly stop commerce between more distant Civs, and harder to shut out your neighbor.

Civs could possibly enact tarriffs on other Civs, so that they earn 150% of the commercial revenue from a trade route than they should. This would anger the other Civ and generally reduce the amount of commerce the route produces (again, increasing the "movement cost" by some percentage).

I have no idea about programming and calculatio time, and how possible it is to have trade rates calulated between every city. My guess is that it would take a lot of resources. It could work like SMAC, where each Civ's cities are ranked by size and then matched up for trade, or the system could be abstracted so that each Civ constantly trades with each Civ, and only the average distance of all cities from Civ to Civ are calculated... something. Maybe the more programming knowledgable have suggestions?

Anybody have a different idea, or comments on mine?

[lajzar]

I had a lot of free time at work yesterday, and wrote up quite a detailed model for gold income. I divide gold income into two distinct areas. Internal trade is the classic trade arrows on the tiles (which probably does need a revamp). External trade is the area I detailed, and followed a model similar to yours.

I scrapped it when I realised that once rails, and especially air, are discovered, the trade income would get ridiculous. You'd suddenly find yourself in a situation where you have near-instantaneous (in game time) travel between every city on the map/continent. The other reason I dropped this idea is that it gives you nothing to intercept in a trade model vis a vis piracy.

Another thought I had was for civ2 style caravans. the catch is that the only benefit is in luxuries, spread evenly over all your cities over a 20 turn period. This would make it reasonably easy to avoid unhappiness - possibly too easy. otoh, you wouldn't get ridiculous amounts of gold and science from those, removing the silly tactics common to civ2.

[Fosse]

I certainly see the need for diminishing returns on the travel time factor in order to prevent rails and airports from simply bringing in tons of commerce. But that is something that I think can be balanced.

I don't see the need for piracy when we are talking about the gold generating commerce of trade between two civs in my proposal - remember that this is something out of the player's hands and that resources are not being traded. My idea is compatible with resource trade routes on the map, like CtP or GalCiv, which could be intercepted.

What model did you come up with? I'd be interested in the details, and maybe someone on these forums has an idea to make it compatible with "near instantaneous" travel of the modern age.

[Switch]

I really like that model. I've never liked that building roads increases the commerce from that tile...especially since it really looks ugly to have roads, and more so railroads, all around your nation.

Perhaps a way to make rr not be a problem for income is just to make distance an inverse multiple, ie 1/x ,x= turns to reach...and to negate the rr problem (since you can't have 1/0), just make the function, say, 1/(x+y), where x is the number of turns and y is some fraction of the distance between the cities. Obviously a more complex and realistic formula could be made, but this is simply a basic idea so that travel time doesn't make trades overly lucrative.

[lajzar]

ok, the model I made (from memory, so details might differ from my original).

Each city has internal trade and external trade. The internal trade component is much the same as the current tiles = arrows system. I wasn't working on that area.

For external trade, there is nothing until you gain the "trade" technology. Certain technologies (trade, corporation) and city structures (rail depot, commercial dock, airport) give extra trade routes.

For each trade route that you have, the computer calculates an income between the city and the neighbouring cities. Foreign cities are only considered if you have a trade diplomatic agreement with that civ. Beyond that, cities are chosen against to the nearest linear (not travel) distance.

The actual trade income for each route increases with population and decreases with travel time. The actual travel time had some detailed rules involved, depending on whether both cities had certain improvements (don't use the air calculation unless both have airports etc).

Of course, this system didn't allow for intelligent selection of cities, and overseas (ie over water, not necessarily foreign) trade would be effectively zero.

One modification I also considered is that the total number of trade routes is capped as above (instead of civ2 capping at 3), and you need caravans to create a route. Caravans give NO initial boost when they arrive. Instead, their normal income is spread over 20 turns or so, gradually diminishing. Repeatedly trading with the same city also gives diminishing returns, and because teh value of an established route diminishes over time, eventually reaching zero, there will always be a reason to build more caravans.

[chrispieX]

I really like fosse's idea. Taking trade out of the players hands and making it part of the reality of civilizations I think is very clever and realistic. Obviously it would be good as said, if players can influence it, maybe try to kill it or demand more of the profits from it.

I also think it would be good if individual cities have a 'commerce level', which is built up over time by the flow of luxury goods into the city and perhaps other things. It's this commerce level that is used to decide how much trade can go out of this city. This encourages people to not only collect luxuries, but to then create and encourage trade based on them.

It always seemed a lot more logical to me that luxury goods would be a way to generate money, rather than happiness, this money can then be used to do many things, including increase happiness.

[Xorbon]

I don't mind Civ 3's system of resource trading for the purposes of trading important (i.e. strategic and luxury) resources. I think it's important to be able to control whether you want certain civs to have access to these resources. Fosse's idea would then only apply to the trade of 'other' resources, like sugar, salt, timber, etc.

As far as caravans go, I don't want to see them brought back. They're a micromanagement pain in the ass.

I thought up an idea at one point that would bring back caravans, but not as units. A city would build caravans like they would any improvement or unit. However, once it is built, instead of a unit appearing, a screen would pop up prompting the player to choose among various options. Options would include destination city, and commodity to be carried. Once the player makes the necessary choices, a trade route would be automatically created (after several turns) between the player's city and the destination city. The number of turns it would take would depend on travel time and tech level.

Having said that, I think I'd prefer an abstract system (like Fosse's).

[Trifna]

I agree with what Fosse wrote, but I think that you also need to consider other factors such as:
- Transport technology of the civs which change everything
- The level they're at in term of production and technology (more production and better products -> you export more)
- Which ressources they each have on their territory
- Else?

So a simple meter of commercial balance could be done ("You receive X, You send Y")

[Fosse]

I don't think that strategic resources (ie, with military usefulness) should be traded "automatically." Luxuries, perhaps, since it's really the money from the commerce of luxuries that make more people happy (a country that has a lot of incense doesn't riot when they run out because their homes don't smell as nice, they get mad because they no longer have a product to sell).

Lajzar's proposal is interesting, and certainly less processor heavy than mine would be (trade for nearest cities, rather than all cities).

Trifna, to try to throw out some possible answers to your points:

Transport tech: Perhaps a few benchmark techs would simply adjust the rate at which transport times are calculated. A civ that discovers map making suddenly gets slightly more profitable routes on their end, while their partner's stays the same. A civ that discover's automobiles gets another boost. These changes could be small, since they'd be coupled with other improvments (rails, for example).

Tech and production levels: If a city's (or a civ's) production level is made to be part of the vague "commerce potential" I talked about, then this would seem to solve that. Yes?

Resources: I wouldn't want iron or oil traded this way, but perhaps a bonus for each luxury in your resource that the other guy doesn't have (big for you, small for them, and reversed when they have the resource). Then I could put a tariff on spices if there are none in my territory, and anybody with spices gets their "spice bonus" reduced and mine gets increased.

[Trifna]

Control over trade: I'm not sure the player should have a control over trading. It'd depends on the political system and perhaps other factors. In certain cases like a liberal market democracy, you'd need to make an embargo to block it. Also, depending on the system, the trading could give money to the state (you) and/or boost your production (more money thus more investment).

Tech and production levels: Yes, your commerce potential would do. But having a trade balance would a still be useful


But really, my main concern would be how to bring this in a simple model. I know I would accept the complexified model, but I also know I take more of this than the average...

[Fosse]

What do you mean by having a trade balance Trifna?

I don't think that the model is too complicated as long as direct player interference is left out. If the player has to tweak every variable then my plan is no good. If they can try to influence trade patterns indirectly, then that strikes me as fun.

[Xorbon]

Trifna makes a good point about more advanced technology and better products. I'd like to see civs produce various specific products which would be traded between cities and civs. Tech level would affect which products are available, and which are in demand. Examples of products could include precious metals, salt, textiles, stone, cotton, wool, timber, food, automobiles, machinery, electronics, etc., etc. Some products could require certain city improvements (such as autos requiring factories), and others could require access to specific terrain types (salt could require access to desert terrain). Production and trading of these types of goods would be automatic. Things like embargoes, tariffs, subsidies, and free-trade pacts would be ways for the players to affect trade.

Of course, any system of trade would have to be simple enough and not too 'bulky' for the human and AI to use, even if the computer code itself is complex.

[Trifna]

I think that if we go with subsidies and all that, it's gonna get complicated. We better simplify by puting more or less protectionism, simply. Of course, this can imply treaties where both parties engage each other into less protectionism, embargos...

Fosse: When I say trade balance, I mean the relation between what goes out and what comes in. So if you sell 30M$ to the exterior and buy 35M$, you have a negative trade balance (deficit) equal to 5M$. Just like a household that buys more (outcome of money) than the income.


But we have to be careful with anything to do with this since humans didn't knew economic theories alot before pretty lately. Of course, they still managed to organize some parts like favoring trade by treaties or agreements, blocking others' trade, etc. So we have two general questions to solve here:
1- How could the trading be put simply
2- How could we do this without anachronisms


So, considering these two factors, here's what I'd say:
1- You get something from trade (should it be money or else)
2- You can always try to find new deals
3- You can always block trade (at least TRY to)
4- Favor/deteriorate trade by roads, good diplomacy or else
5- The three above have some options that are added with technology (treaties, greater production...)
6- Exports and imports both have different effects (you benefit from exterior products but also pay for them (how? to be seen), you're paid for your exports and lose the product... which sometimes you don't need anyway). This last point is the tricky one: should it be simplified? How should it be? Everything seems on this last point.

Perhaps we could simplify (6) by considering it as an exchange of products, so each trader gets a different amount of money based on the advantage gained from the exchange (which depends on 1 to 5). To complete the system, a second aspect would be added: the possibility of seeing your economy invaded by the exterior somehow or to invade someone else's (based on some factor, a bit like two person in a bed trying to bring the blanket on their side). Economic vampiricism. So (6) would be highly simplified by cuting it into two pieces instead of trying to bring complete economic laws.

[Sir Og]

Originally posted by Fosse I would like to see the ridiculous-when-you-think-about-it system of worked tiles generating commerce done away with. To Civ 3, people find gold on the ground, and can find more of it when someone builds a road.

I propose that no commerce be generated by tiles. Instead, the number of citizens in each city will help to determine that city's "commerce potential." This potential would represent the exportable goods the city generates as its people go about their life.

Your proposal is identical to the current model since the tiles worked are equal to the number of citizens. And roads adding trade is a perfectly good abstraction of getting more trade from improved infrastructure.

Originally posted by Fosse I certainly see the need for diminishing returns on the travel time factor in order to prevent rails and airports from simply bringing in tons of commerce.

The total commerce generated by a certain civ after the invention of advanced seafaring, railroad, flight etc. should be significantly higher than the ancient times commerce levels. But there should be a corresponding increase in the "expenses", such as support cost (for both units and city improvements), more money needed for research etc.

[Fosse]

Originally posted by Sir Og


Your proposal is identical to the current model since the tiles worked are equal to the number of citizens.

Not quite. What I've asked for is the number of people to be one of several determining factors in how much potential commerce a city generates. Other factors can impact the potential, such as improvments and tech.

The amount of potential commerce is not then simply added to your treasury, as in Civ 3, but instead used to determine actual commerce between cities.

Thus, it is the interplay of various relevant factors between cities that generates commerce. Not worked tiles.

And roads adding trade is a perfectly good abstraction of getting more trade from improved infrastructure.

No it isn't.

It's an unrealistic model (roads in the middle of nowhere make you rich), with unrealistic results (roads, roads, everywhere!). It leads to an ugly map and is micromanagment intensive.

Trade infrastructure simply existing doesn't do anybody a damn lick of good. It needs to make travel from one place to another easier if it's going to have any impact whatsoever.

The total commerce generated by a certain civ after the invention of advanced seafaring, railroad, flight etc. should be significantly higher than the ancient times commerce levels. But there should be a corresponding increase in the "expenses", such as support cost (for both units and city improvements), more money needed for research etc.

That would be a perfectly satisfactory way to balance things in my opinion.