Subsidies and Aggressive Trading Practices

If the AI had been coded properly

Click to expand...

I'm not sure what you mean by 'coded properly'. In my view, this means that GPT trades are a viable part of the game. Under this model the AI cannot follow human practices because like you say, an intelligent human will only pay what they can afford. I've already mentioned that it makes no sense to allow the AI to ever have a GPT surplus because this is monies that are not going to immediately develop its empire. Therefore the AI would never consistently allow interesting GPT trades, making this feature pointless. I'd rather have a feature I use nearly every game than one I seldom, if ever, use. The actual limit makes much more sense as the one I should be using since it is tied into happiness. I agree that I'm raising the AI's happiness via the gifts. However, from what I can tell this impact is just about negligible. I certainly haven't had a deal go from 27 to 28 just by giving a measly 27 GPT gift. It would be interesting to experiment and see exactly how much the subsidies are improving your deals 10 turns later in a case where you're reasonably sure only your subsidies have had the majority effect on relations over the last 10 turns.

...it's an exploit because it wasn't intended by the programers. That's the very definition of an exploit and there's no way around it.

Click to expand...

I don't entirely agree. I might be convinced if the limit was fixed but a mutable limit is clearly meant to be reached, and as I keep pointing out, it is not practicable to have the AI even try to approach this limit. So the feature is clearly meant to be used but is in fact unusable by the current system. So I don't see anything exploitational in getting something that was supposed to work to actually work. Did the programmers intend for me to be able to make a decent trade? Yes. Did they expect me to use this method? Probably not. But I'm sure using a feature is more important than how it was used. And as I pointed out in my lengthy post, who are we to say that a proper AI is not supposed to be doing things that it can't afford? Human players run deficits, and the AI runs deficits. I don't see how the particular cause of a deficit has any relevance as I already went at length to point out.

I play Deity and don't find the AI challenging enough. It becomes challenging enough when i add "no tech trading". When i get better at the game, this may not be enough anymore to make it challenging.

Click to expand...

I am genuinely impressed. But I am addressing my comments to those whose choices may influence others. If something makes the game too easy for me, why would I have the right to tell others not to? It's up to them to decide if it makes their lives too easy or not.

No one is going to prove what the developers intended unless the developers decide to tell us. As a programmer, I would give 10 to 1 odds they never intended to create this loophole.

Like Zombie69, I don't much care if it's an exploit or not. Being a lazy person, I've decided it's too much work for me. If that keeps me at a lower level of difficulty, so be it.

Eqqman, there's no way i can convince you, because you've already convinced yourself. Probably because you like doing this but don't want to feel like you're cheating, so you've convinced yourself that this wasn't an exploit so that you could keep on doing it and still feel good about yourself.

For anyone looking at this objectively, this is obviously an exploit. Just look at the number of people who say so in this thread.

If you want to keep thinking that it's not an exploit, so that you can feel good about doing it, that's fine. Just don't try to argue on the forums that it's not an exploit, because it obviously is. Think what you want, but stop trying to prove something that can't be proven.

I don't view this as an exploit, I think it's pure genius. Economic warfare is as valid a form of warfare as any other, and there's a long history behind it!

DaviddesJ said:

What's "proper" is a matter of opinion. I think the designers coded in an upper limit for how much the AI would pay for a resource, with the idea in mind that it would never actually pay that much unless it happens to have the free cash flow. You're exploiting a loophole in the latter constraint.

Click to expand...

That much is arguable. While the upper limit may not have been intended to ever be reached, the more consistent result is that it's impossible to even come close to trading for the upper limit. If, given the example, Asoka is willing to trade 20gpt for a resource, and the most consistent offering is in the 1-5gpt range, this suggests either that the system is operating poorly (ie the AI is designed to efficiently operate its economy too well), or that the AI is trading at intended levels, but that the upper limit on trades is unreasonably high.

Fixing the problem from the standpoint of the broken system is likely impossible. It would mean recoding the AI to incorporate some form of inefficiency into its economy. Presumably, it does reassign worked tiles as needed to run at a state of greatest efficiency. It's safe to assume that was intended (and to undesign, or design it out of the system, would be counterproductive).

More likely, to appease those who feel that the current system can be abused, the upper limit on these trades will be reduced to more realistic values. I assume that the AI will be similary treated for the reverse situation, in which a human attempts to purchase a resource off of an AI, and the AI will, as well, expect a reduced amount of gpt in return.

Naismith said:

No one is going to prove what the developers intended unless the developers decide to tell us. As a programmer, I would give 10 to 1 odds they never intended to create this loophole.

Click to expand...

As a programmer, myself, I find the original intent of the developers to be a moot point. If someone found an original way to make use of my programs in a way I didn't anticipate, I'd be pleased by it. Given that the scope of this particular tactic does not distort the game - it's an add-on feature at most - I would be tickled pink if I were the developer of the game, didn't anticipate this, and found someone made a successful strategy out of it.

But I would guess that this was fully anticipated, because there is a limit on how many GPT an AI will spend. The only thing that can be remotely considered exploitative is gifting GPT to the AI so they can buy your resources; on this, all I have to say is, look at the real world. Countries do this all the time, give subsidies to other nations so they can buy the donor nation's goods.

ownedbyakorat said:

Countries do this all the time, give subsidies to other nations so they can buy the donor nation's goods.

Click to expand...

Aye. It's quite possibly the most realistic aspect of the game, aside from perhaps the map generation scripts.

...there's no way i can convince you, because you've already convinced yourself.

Click to expand...

That's interesting because I feel exactly the same way about people convinced it is an exploit. I'd thought my very lengthy 2nd post was fairly objective and convincing. People concerned about what the developers 'intended' are playing the game with made up rules that the game knows nothing about. Did the developers intend for people to pre chop forests to get a Wonder in one turn? Did they intend people to queue switch to build Workers/Settlers while allowing the city to grow? Did they intend the CS or CoL slingshots? It is inevitable that in a complex game things emerge that are unanticipated. That does not automatically make them cheats. If the developers want to come out and say they disapprove, that's fine.

Just look at the number of people who say so in this thread.

Click to expand...

I'll admit that nobody else seems to think anything I've said makes sense to them, which does surprise me. But I don't consider the 'number of people' to actually be that many. I don't appreciate being told to shut up, however indirectly, and this is supposed to be a place for people to express their views. And at least I'm not indulging in personal attacks. However I'll make this my last post in this thread. If nothing I've said so far rings true to anybody at all, then nothing further I can say will.

Edit: In between my typing this reply to Zombie69 and its posting, some people finally chimed in who didn't totally think I was completely off base. So I'd like to thank them for speaking up. I'll get off my soapbox now.

Eqqman said:

It seems like a silly argument to me to say that everybody should be equally competitive, that's not how life works.

Click to expand...

It sounds silly to me too. Fortunately, I said nothing like that.

What I said is that is that this tactic has a major effect on the game, not in a good way, and therefore I would be happier if it were "banned" (or removed). The differences in skill levels between players will still be large, whether they all use this tactic or none of them do. In general, I think they would be larger if the tactic were not allowed.

I also don't seem to hear people on Deity complaining that the AI is not challenging enough. I can't imagine a player out there seriously handicapping themselves in a game, unless they are playing a lower level maybe. If so, simply move up a level.

Click to expand...

Obviously, if you give the AI extreme advantages in terms of production cost and number of units, you can make the game challenging. The Deity level in Civ4 more or less does that, but, because of the scope of the advantages that are necessary, it also seriously distorts the game and limits player options. You say that exploits like this one are fine and, if they make the game too easy, then players can just compensate for that by giving the AI larger handicaps. But I think the game is more fun if the player has fewer exploits (tactics that give a disproportionate advantage relative to the AIs) and thus a lower handicap level is sufficient to provide a challenge. So I personally play at somewhat lower levels but avoid using various exploits. Certainly, I could win at Deity level while using more exploits. But I don't think that's as much fun.

You give an example that seems clearly erroneous (if you could find a way to get 1000 per turn, for example). In practical terms I've yet to have more than 27 accepted for a resource.

Click to expand...

I think the amounts of gpt that can be collected using this exploit are "clearly erroneous".

I don't think anyone has even mentioned yet the money that can be made by arbitrage; i.e., finding one AI that will sell a resource to you for X, and another that you can sell the resource to for Y, where Y is greater than X.

Eqqman said:

People concerned about what the developers 'intended' are playing the game with made up rules that the game knows nothing about.

Click to expand...

I didn't say not to use this. In fact, i use it myself. All i said was that you shouldn't delude yourself into thinking that it's not an exploit.

Eqqman said:

Did the developers intend for people to pre chop forests to get a Wonder in one turn?

Click to expand...

No, and therefore by the very definition of the term, this is an exploit.

Eqqman said:

Did they intend people to queue switch to build Workers/Settlers while allowing the city to grow?

Click to expand...

Again, not intended and therefore exploit.

Eqqman said:

Did they intend the CS or CoL slingshots?

Click to expand...

I'm 100% sure the developers intended for the Oracle and Great People to actually be used to gain techs. I'm also 100% sure they intended for you to try and get the most expensive tech you could out of it. Not an exploit, the game works here as intended.

Eqqman said:

It is inevitable that in a complex game things emerge that are unanticipated. That does not automatically make them cheats. If the developers want to come out and say they disapprove, that's fine.

Click to expand...

I never said cheat, i said exploit. Those are two completely different things.

DaviddesJ said:

I don't think anyone has even mentioned yet the money that can be made by arbitrage; i.e., finding one AI that will sell a resource to you for X, and another that you can sell the resource to for Y, where Y is greater than X.

Click to expand...

Can't be done, because the AIs were coded to always ask a lot more for a resource than they're willing to give you for the same resource. And because the best deal, whether selling or buying, will be with the AI you have the best relationship modifiers with. If two AIs are equal, then they'll both give you the same amount for buying and they'll both ask for the same amount when selling, and you'll always pay more than you'll receive. If the AIs aren't equal, then you'll just be making a slightly worse deal than you otherwise would with one of them, making this trade even less interesting for you.

Zombie69 said:

If two AIs are equal, then they'll both give you the same amount for buying and they'll both ask for the same amount when selling, and you'll always pay more than you'll receive.

Click to expand...

Is this really true? I thought that the amounts were determined in some way by the size of their economy. If it's only a fixed price that depends only on the level of your relationship, then why do the amounts go up over the course of the game?

I might have to try to look at the code. I don't know so much about how this actually works because I've been avoiding it in my own games.

Zombie69 said:

I'm 100% sure the developers intended for the Oracle and Great People to actually be used to gain techs. I'm also 100% sure they intended for you to try and get the most expensive tech you could out of it. Not an exploit, the game works here as intended.

Click to expand...

I'm doubtful that the developers intended the tactic (I would call it exploit) of avoiding certain early (prerequisite) techs in order to cause a GP to give you a more expensive/desirable tech instead of a less expensive/desirable one. I think it's an inadvertent consequence of the relatively simple system for determining which tech a GP will discover.

I'm pretty sure they did intend it, and i expect that it was used by some beta testers and reported back to the company.

If not then yes, it is an exploit.

Zombie69 said:

I'm pretty sure they did intend it, and i expect that it was used by some beta testers and reported back to the company.

If not then yes, it is an exploit.

Click to expand...

I think they probably knew it would be possible to take advantage of the GP algorithm to manipulate which tech you get. If not when they wrote the code, then, as you say, when it was tested. But that's different from intending it. There are lots of things that must have been known to beta testers, and were consciously left in the game (if only because they are hard to change), and I would still consider exploits. The GP tech manipulation is a good example: it's hard to "fix" without making the system much more complicated, so it doesn't surprise me that they left it in even though the effects seem somewhat undesirable and "exploitative".

ownedbyakorat said:

I don't view this as an exploit, I think it's pure genius. Economic warfare is as valid a form of warfare as any other, and there's a long history behind it!

Click to expand...

Yep. Holland fought a war against the English where Dutch merchants sold weapons and ships to the English at the same time. At the end of the war Englands coffers were depleted and Holland was filthy rich

I don't want to enter too much in the quarrel about if this system is an exploit or not... apart saying that the fact developers intended it or not, is not cause to call it exploit (or not).

So exploit of not, this system has some basic "limitations".
1st: it's limited by the number of resources you can sell
2nd: the price for each single resource is limited by the cap

Sure this system can become very exploitative in very specific conditions... but not always you have those conditions in play.

The main limitation of the current system is that AI cannot "price" a resource in any meaningful way (so the AI cannot understand if a susidised deal is no more good for them).

The main problem with the non subsidied resource commerce is binary:
0. the AI uses all its available GPT as soon as they are available, so the time they have a surplus is very limited: you need to use smartly this window of surplus to erap the best price.
1. for the AI there is no "real" value for a resource: a resource (e.g. bananas) that is available from 3 civs have the same value of an other resource available only from you.

Some resources, at the right time, are invaluable for the military because they can allow or block the development of a strong army eg: Iron in early game, Horses for early mongolian games, Oil in late games.
In a more realistic economic model they cannot have the same value of bananas.
Ai, if has money, always buy your resource even if not strictly needed (e.g. if you already have all happy or healty, why pay for a resource that effects only one of those factors?)

I think the commercial model shall be revised:
- AI shall be able to give a sensible price to resources based on their availability and strategic importance for development
- AI shall be allowed to go in deficit (down to a cap) to get a resource that's strategic for them at the time (and then readjust it's economy to bring the balance into parity)
- AI shall be able to giudicate if an acquired resource is no more needed (at the given price) and renegotiate or cancel the deal

With a more intelligent resource market system "subsidies" can remain in play... they'll have an extremely limited scope.

DaviddesJ said:

Obviously, if you give the AI extreme advantages in terms of production cost and number of units, you can make the game challenging. The Deity level in Civ4 more or less does that, but, because of the scope of the advantages that are necessary, it also seriously distorts the game and limits player options. You say that exploits like this one are fine and, if they make the game too easy, then players can just compensate for that by giving the AI larger handicaps. But I think the game is more fun if the player has fewer exploits (tactics that give a disproportionate advantage relative to the AIs) and thus a lower handicap level is sufficient to provide a challenge. So I personally play at somewhat lower levels but avoid using various exploits. Certainly, I could win at Deity level while using more exploits. But I don't think that's as much fun.

Click to expand...

Amen. Look, it's all about what's fun. If you like using this trading technique, go for it. Determining whether this technique is "intended" or an "exploit" is a fruitless task, although I find this thread very entertaining.

DaviddesJ said:

Is this really true? I thought that the amounts were determined in some way by the size of their economy. If it's only a fixed price that depends only on the level of your relationship, then why do the amounts go up over the course of the game?

Click to expand...

In a nutshell, I'd say the largest factor is inflation. The AI buying and selling rate is probably pegged in some way to that, though I wouldn't doubt that factors such as the size of a given economy are taken into consideration. Though not looking at the AI leader specs in the XML file atm, I'd also hazard to guess that a large portion of the rising prices can be found in there.

DaviddesJ said:

I'm doubtful that the developers intended the tactic (I would call it exploit) of avoiding certain early (prerequisite) techs in order to cause a GP to give you a more expensive/desirable tech instead of a less expensive/desirable one. I think it's an inadvertent consequence of the relatively simple system for determining which tech a GP will discover.

Click to expand...

I feel as though if what you were stating were true, they would have developed a more linear list of offered GP techs in relation to available researchable techs. Given how the list is designed, I can only say that there is no question it would result in the availability of relatively advanced techs while at select points in the tech tree with regards to each Great Person type.

Instead of offering early game techs, then late game techs, then mid game techs, then early game techs, they could have just offered early game techs, then mid game techs, then late game techs, requiring you to progress through the tech tree as it is ordered. That you can blatantly skip certain techs in order to produce more desirable offerings from Great People seems to be a intended feature.

DaviddesJ said:

There are lots of things that must have been known to beta testers, and were consciously left in the game (if only because they are hard to change), and I would still consider exploits. The GP tech manipulation is a good example: it's hard to "fix" without making the system much more complicated, so it doesn't surprise me that they left it in even though the effects seem somewhat undesirable and "exploitative".

Click to expand...

With regards to the first part, yes, there are certainly things that would be too difficult to fix to warrant delaying the release of the game or a given patch (the rounding error produced by whipping with a production bonus, for example). Issues such as chopping were unlikely deemed very important during the development or testing phase, but only later changed as a result of an overwhelming amount of feedback related to the given issue.

With regards to the second part, as I note above, they could have made the offered GP techs more linear. I'm not sure if it would need to offer early->mid->late game techs or late->mid->early game techs in order to keep the system more closely linked to the actual tech tree, but the current staggered tech offerings leads one to believe that some sort of "slingshot" must be intended. That said, manipulating Great People (and The Oracle to a much lesser extent) in order to research relatively advanced techs is probably one of the most interesting aspects of the non-warmongerer side of CIV.

Nares said:

That you can blatantly skip certain techs in order to produce more desirable offerings from Great People seems to be a intended feature.

Click to expand...

Your argument seems to be just that it would be "easy" to do something different, if they didn't want players to be able to manipulate GP techs (e.g., avoid Masonry in order to get Civil Service), so therefore they must have intended and desired to enable such strategies.

I think there are plenty of glitches in Civ4 that seem equally "easy" to eliminate, yet are there. The argument that something seems easy to avoid therefore doesn't carry much weight. I think you overestimate the available resources for eliminating such glitches, and also the desire to do so. They may well have not cared enough to try to eliminate it, but that's quite different from saying that it is "intended".