Unit Trading - bad idea

[dexters]

Originally posted by Spiffor
Sorry Boris, but this one is a firm no-no. If a civ both has the techs and the resources to build a unit, I wonder why this Civ would buy units at all, except in the rare circumstances when it is in urgent need of manpower (this is about as rare as a Civ needing a loan).
A resource requirement to unit-trading is about as 'good' as no unit-trading at all. With this, unit-trading would be about as useless as colonies IMHO.

Come on. All Civs could use more units almost all the time. If you watch AI in debug, only in rare superpower AI civs will you have large reserves.

In most cases it could always use more units.

Spiff, you're too fixated on unit trading from a superior Civ to an inferior Civ. You probably have your "ideal" puppet master game in mind. But I'm thinking more generally here. From experience, there are plenty of scenarios where trading units an AI civ can build is required. This includes bolstering the ranks of a frontline Ally that is in a war. And it could also be a case of replenishing depleted ranks of AI armies as the war drags on. In these cases, you can play as the arsenal of your allies.

Besides, unit trading is not a one way street. We humans may also want to acquire units. And the tech requirement makes it more difficult for humans to acquire highly advanced units and use it effectively in a war. And you know as well as I do humans are still way ahead of the AI when it comes to massing units and making the best use out of all units in an attack.

[Boris Godunov]

Being able just to give units could be a big plus, too.

The Germans enjoy a position of supremacy, but the English seem to be catching up fast. The French are at war with the English, however. Germany doesn't want to get involved in a war, as it would interfere with its social works and create citizen war weariness. They do want to limit the power of the English, though. So instead of allowing the English to conquer the French and thus become more powerful, Germany sends France tank divisions to help it maintain the war. Thus the balance is maintained, and Germany is able to keep its lead while France and England keep each other in check through the slugfest.

Likewise, if Germany were behind England, it could use this situation to its advantage. Sending units to France will allow it to help weaken England without going to war and risk losing cities. This would be a nice alternative for more passive empire builders.

[dexters]

Sometimes, I get the feeling the AI does play international politics. Such as giving a civ i'm attacking a resource, when previously they were unwilling to trade it to them.

I hope Soren beefs up AI's international politics side of things and make it capable of supporting other civs against a common enemy, even if they are not at war with them.

[Spiffor]

Sorry Boris, but your example does work for a few units, but doesn't for quite a few others :

According to your idea, only a Civ that can build swordsmen can buy them.
Same for chariots.
Same for musketmen
Same for infantry ( )
Same for ironclads

Swordsmen, musketmen and infantry are quite useful through the game. Besides, if a resource requirement is needed rather than an 'era requirement', that means you can sell riflemen and guerillas to any prehistoric tribe. If there is any limitation from the client's side, I think this one isn't good.

[dexters]

I'm not sure about resource requirement. But a Civ with a tech without a resource to build it can still trade for units.

I read Boris' post and all he was saying was that if a Civ have The Tech + Resource (as in they just got it a few turns ago) they can immediately buy units and speed up modernization instead of waiting for however many units for their cities to crank one out. In peacetime, AI civs usually don't modernize too quickly, especially those Civs who aren't too large and don't have a lot of gold around.

[Boris Godunov]

Originally posted by Spiffor
Sorry Boris, but your example does work for a few units, but doesn't for quite a few others :

According to your idea, only a Civ that can build swordsmen can buy them.
Same for chariots.
Same for musketmen
Same for infantry ( )
Same for ironclads

Swordsmen, musketmen and infantry are quite useful through the game. Besides, if a resource requirement is needed rather than an 'era requirement', that means you can sell riflemen and guerillas to any prehistoric tribe. If there is any limitation from the client's side, I think this one isn't good.

They answer to the latter problem is to make those units require saltpeter.

Even if a civ had to have the required tech, it would still be useful for it to be able to buy them in bulk from an ally. If I can buy a bunch of units from a friend rather than rush them in my cities (hence keeping my cities free for either building even more units or social projects), I'd consider that an advantage.

[Boris Godunov]

Originally posted by dexters
I read Boris' post and all he was saying was that if a Civ have The Tech + Resource (as in they just got it a few turns ago) they can immediately buy units and speed up modernization instead of waiting for however many units for their cities to crank one out. In peacetime, AI civs usually don't modernize too quickly, especially those Civs who aren't too large and don't have a lot of gold around.

Not really, I was focusing on the resource. I think that, so long as a civ has access to the requisite resources, then trading them more advanced units than their technology allows would be appropriate.

Making it Era-based brings up the ridiculousness that I could transfer, say, nukes to a civ that doesn't even possess rocketry. If they don't have the means to launch a nuke, what good would it do? Or, likewise, selling an ironclad to a civ that doesn't yet have access to coal. How would they power their ship?

[SewerStarFish]

As unit trading was in Civ II, I would think the programmers already have a good model of what they are going to do. The Civ II AI pretty much did not trade units but afforded you the option to bolster an ally in need.

Now your MPP with the most backward civilization could have some umph. It also could be used to amend damaged relations.... or giving a bonus for cultural flips [picture a city defended by foreign troops being more prone to flip to that civ].

[Mad Bomber]

Originally posted by Fosse
I think that the tech advancement that lets one build a unit, for balancing purposes, should be viewed as a basic understanding of how to use that unit. The countries that don't have the capacity to build tanks, still know what they are and how to work them - in Civ terms they "have the Tank advance."

If unit trading isn't restricted to known advances, then the player could never spend the money or time to research or buy military tradition, and still buy hordes of cav... and similar things.

For design sake, the requirement to build the unit must be met.

For civs that don't have the resource or the production base to build those units though, trading is a great - and realistic - option.

I agree and you have hit an important but neglected point. Sometimes a civ can't build a unit because they lack the resources to do so. If the 'gifter' civ can't trade the appropriate resource to the civ then unit trading would be a quite valuable alternative.

As far as mercenaries go, they have been used in some of the earliest battles up to present times. In Alexander's first great battle at Granicus, over half of the Persian soldiers were greek mercenaries. In the American Revolution both the British and Americans used mercenaries (the Americans called them volunteers) And in 1941 the Americans sent a number of fliers to china as mercenaries termed the AVG (they were hardly voluteers) later to be known as the "Flying Tigers". As recently as the Korean and Vietnamese conflicts; Mercenaries have been used or its use suspected; Russia and China were suspected of providing aviators to the countries of North Korea and North Vietnam as well as support personel to the SAM sites used by North Vietnam. In the Israeli war of independence many soldiers came from different countries and could be considered mercenaries.

I like the idea of unit trading it should add more realism and strategic depth to a great game although some measures should be taken to prevent abuses.

[Spiffor]

Boris, Dexters :

I think the main question over the 'limitations' issue is : "What are we concerned with ?".
As far as I can tell, we are not concerned by realism as long as we don't see utterly fantasist outcomes, such as bronze-era people suddenly becoming a nuclear power...

I suppose we are all concerned with having a balanced feature, i.e one which isn't overpowering, and which doesn't make the game's other features useless.

Let's imagine an unlimited unit-trading (i.e Civs trade units between themselves as they please, with the price they like). In what occurences could it become overpowering ?

- it could artificially turn a minor Civ into a major military player, if this Civ has the right friends who give the right amount of units. Used at full potential, unit-trading would effectively allow to fight wars at full power on other continents, for example. It would kill the purpose of having a transportation fleet at all, except if you can only trade units that are already in the client's territory (like in SMAC).
However, using unit-trading to wage a war at full force has drawbacks the supplier isn't always ready to pay : the supplier's units will conquer cities, but these cities will end up in the hands of the client, NOT the supplier. Besides, if the client takes a too big chunk of the supplier's army, the client might become more powerful than the supplier.
In this case, I think game mechanics can balance this out by themselves.

- It could give a military bonus to the player, since the player knows better than the AI how to use units at their full potential. A surge of units would mean, for the human player, a sure victory.
Indeed, to balance the impossibility for the AI to match human wits, the AI should have bonuses when it comes to unit-trading too. This way, the AI would get more units through trade than the player, and the military 'balance' (i.e plenty units badly used by the AI, a few units well used by the human) would be kept.
I think the current AI-to-AI trade bonuses are what we are looking for in this case. I don't know whether they would need to be tweaked or not, but I think the fundamental mechanics are enough.

- It could give to the player an unfair bonus by raking money from 'whoring' its units while the AI cannot. The AI would end up losing its money to a manipulative player, and this can easily be viewed as an exploit.
I think the right answer to it would be to program the AI so that it 'whores' its units just like it 'whores' everything else (techs, resources, luxuries, alliances, maps). To teach how to whore units would give the AI the same advantage as the player, and maybe even better, thanks to the AI-to-AI trade bonuses.
Besides, I think the fear of unit-whoring by the player isn't a good reason to keep unit-trading at bay. Tech-whoring was a well-known abuse of Civ2, and was also an abuse at the very beginning of Civ3, before the AI knew to tech-whore efficiently. Yet, I see nobody complaining about the ability to trade techs

- It could deter a Civ to build units, and concentrate on social buildings in its cities (Boris).
I think the best way to avoid it is to have the standard price of a traded unit significantly more expansive than the price needed to rush-build the unit (like 6 gold per shield or something). If a swordsman costs 180 gold already, the Civs will prefer to build them by themselves. I also think their should be specific modifiers to the price, such as technological gap (higher tech units are even more expansive, and low-tech units are a bargain), as well as the usual price modifiers.


The more I think of it, the more I think this mechanics can be balanced without imposing artificial limits to it. The only really tricky part is to teach to the AI how to unit-whore. If it is impossible, then limitations on unit-whoring would be good

[ALPHA WOLF 64]

perhaps we are concentrating to much on trading, maybe we should consider unit loaning. much as a worker retains its nationality, so could military units. then after 20 turns, the unit returns to its original nationality. Also, you could only loan units to civs you have a MPP with. just imagine the scenario were you have loaned units to both civ A & B, A attacks B triggering your MPP against A. You now have units positioned inside A ready to attack if they arent taken out right away. this would give the AI one more thing to "consider" before it launches another ill-advised war.

[Ghengis-Sean]

Trading Units would a bit of depth to the game, and as others have said it isn't unprecedented in history. Still it should be done with certain restrictions. Namely your "lent" forces shouldn't exceed the number of troops the locals have. A few foreign volunteers/mercenaries is one thing, but twice as many foreign troops as local troops could be a bit hard to swallow.

One way around this would be to allow troop selling at a limited rate per i.e. equipment and advisors, and also allowing players to send units to foreign governments as "volunteers."

this is actually more common that you might think. During the Korean War several hundred thousand Communist Chinese "volunteers" fought and died in the war, even though the US and China technically were at peace.

I'm not sure how the mechanics of "volunteers" would work out, but I think it would require a formal diplomatic agreement, like a right of passage, maybe a volunteer agreement, that is something like a right of passage treaty, but the moment your troops cross the border they adopt your allies color but stay under your control.

maybe

[Mad Bomber]

Trading Units would a bit of depth to the game, and as others have said it isn't unprecedented in history. Still it should be done with certain restrictions. Namely your "lent" forces shouldn't exceed the number of troops the locals have. A few foreign volunteers/mercenaries is one thing, but twice as many foreign troops as local troops could be a bit hard to swallow.

I think that the primary limitation on mercs should be the size of your bank account. Merc's are expensive and if incororated in Civ 3 they should be at least twice as expensive as a normal rushed unit. Unit experience and technology level should also play a role in the cost of the unit.

A formal agreement should be made for a merc but it should not be linked to MPP's or Alliances.

Alpha: I think that an option between unit trading and loaning should be part of the agreement (trading a unit should be more expensive than loaning the unit)

[dexters]

Spiff, it isn't so much an artificial limit but a logical limit.
Having civs who can't build unit X acquire a technology before running the advanced units seems to be more inuitive than your proposed free for all buffet (so to speak). I'm open to the idea of trading the techs away along with the units.

This would also get the realism junkies into the fold since one could presume that you send tech advisors along with your units to school the Civ.

---------
BUT that stuff is really technical. Your long post is a step in the right direction... meaning a step back to see the big picture.

What is the big picture? Why do we want unit trading?

1) It adds depth and strategy
2) It adds another layer to the dipomatic game / puppet master.

That is why people want it.

Firaxis could very easily rig a system such that the AI never trades its units, and basically, build a unit trading feature just for humans to vent their megalomania. AI will happily accept good human trades but never really initiate any of their own. I DON'T WANT THAT and I think everyone in here don't want that system either. That is so 1990's AI.

For unit trading to be include the AI MUST be able to

1) Manage its units better than it can now.
2) Trade amongst themselves as well as initiate trades of their own
3) AI must be programmed to see unit trading as a potential PROFIT center

4) VERY IMPORTANT: AI must be aware of/capable of doing puppet mastering of their own. That is, if the leader AI perceive you as a threat, it will very likely send units as deep discounts, or even for free to a Civ that is close to you, or you are currently at war with.

5) A system of artificial limits will likely be required to prevent human manipulation and trickery.

Those five general points must be met for UNIT trading to be worthwhile for me.

A unit trading system where the AI is sort of aware of it, and does it sometimes, should not be included in the game. Because it makes the human player twice as powerful. Because in the current Civ3/PTW build, human players cannot send direct aid in distant wars, except for gold, techs and resources. With unit trading, propping up Civs that would otherwise have been wiped out becomes ten times easier. And it will

A) Cripple the AI's ability to conduct war B) Make the human too powerful in international politics.

To summarize the AI must be able to manage its units better, hoard units when needed, see it as potential for profit and as well as be capable of puppet mastering itself for the system to beworthwhile.

It is a tall order, and that is why I suspect it may not be inclued. Maybe Civ4.

[Spiffor]

Originally posted by dexters
For unit trading to be include the AI MUST be able to

1) Manage its units better than it can now.

I indeed hope the AI's military prowess will be improved in Conquests, for general reasons. But if not, I think the AI-to-AI trade bonuses will act as a "balancer" the same way as the production bonuses currently do. If an AI can buy 2 units for the price you'll pay for 1, it catches up on the military level the same way as he's doing now.

2) Trade amongst themselves as well as initiate trades of their own
3) AI must be programmed to see unit trading as a potential PROFIT center

These are very important points that come together. Indeed, the ability of the AI to tech-whore and to resource-whore makes diplomacy very interesting in Civ3, and makes the game generally more challenging. If the AI is unable to see units as a potential profit center (which can happen, I suppose such a feature would require quite much code), there must be indeed limitations to unit-trading. I'd think along the lines of maximum price (relative to the AI's income).

4) VERY IMPORTANT: AI must be aware of/capable of doing puppet mastering of their own. That is, if the leader AI perceive you as a threat, it will very likely send units as deep discounts, or even for free to a Civ that is close to you, or you are currently at war with.

Yes, I think it is generally extremely important the AI does that, and assesses threats and friends more efficiently. This way, the AI could give resources and techs to a buffer Civ (like a human can already do), rather than simply switching between war and peace. In my current game, I'm giving loads of iron, coal, rubber, techs etc to AI civs so that they fight the Celts more efficiently. I'm also building railroads into their territory so that they bring these troopsto the Celts more quickly. And it works
Without this understanding, the AI will basically not understand the use of unit-trading, except if it understands it as a source of profit like it already does with the techs. It will make a "so 1990's" AI in this regard, which is sad. However, I'm so fond of the feature itself that I don't think it justifies to remove it But I agree it is a matter of personal taste.

5) A system of artificial limits will likely be required to prevent human manipulation and trickery.

In a situation where the AI can't work properly with the system, yes. I think either the amount of units traded per turn should be limited, or the maximum price should depend of the client's GDP.

Because in the current Civ3/PTW build, human players cannot send direct aid in distant wars, except for gold, techs and resources. With unit trading, propping up Civs that would otherwise have been wiped out becomes ten times easier.

Well, don't forget sending workers to build roads / railroads. This kind of direct help probably explains why the threatening Celts are now losing this war in my current game. I agree I had unexpected results : the Russians and English get the most poils, while the French and Germans suck unlike what I wanted.

A) Cripple the AI's ability to conduct war

Don't forget the AI will be in charge of these units, making unit-trading effectively a balancer between weak and strong Civs More seriously, I see your point, and the feature could prevent the formation of powerful AI empires if implemented. I think the best way to avoid it is to avoid unit-trading to take place too soon. Maybe only Civs connected by the trade network could be able to trade ? That would already give some time for the AI to evolve into powerful or weak empires.

B) Make the human too powerful in international politics.

If the client's enemy becomes angrier everytime the human trades units (which is a purely logical outcome IMHO), the human won't be that powerful, because they will show restraint. After all, unit-trading is for the people who don't want to afford a full-fledged war.