Discussion thread for exploits and strategies

[Thrallia]

I'm just using that point to show one more reason that the gifting exploit is indeed an exploit...because it is being used to bypass the fact that the AI will generally want more of you than it is able to give.

Whether it was intentional or not, the fact that an AI asks so much more of you for a resource than it can usually give for one maintains the overall feel of a negotiation with the AI-that they will give less than they think something is worth.

 

[occam]

I also disagree that there is any difference between "regulable exploits" and "non-regulable exploits".

Didn't you use it as a VERB, and say it was a problem if players were "exploiting that AI weakness of attaching too high an internal value to the resource"? I submit that all players exploit AI weaknesses all the time. YOU do it. The AI is KNOWN to be weak. That is why the handicaps exist, the player is always going to out-build, -maneuver, -trade, etc. the AI with better choices.

Exploiting AI weaknesses is not a regulable issue. I don't know what you are proposing is the regulable issue, but if that is all -- bzzzt you lose.

If exploiting AI weaknesses were sufficient, tell me how to AVOID exploiting the AI weaknesses at, say, troop deployment. Or research path. I exploit the h*ll out of those AI weaknesses.

Do you have some categories of AI weaknesses, DaviddesJ? Why aren't you crusading against people who exploit AI weaknesses of troop placement? If magnitude of effect is your key concern (as you claim), that has a much larger, and more proven effect! I am sure that almost all the GotM winners use the "poor AI troop placement" exploit.

An AI weakness does not an exploit make.

It's quite obvious that the gpt that the AI will offer for a resource is the minimum of "available gpt" and the "gpt value" that the AI assigns to that resource.

The AI remains willing to offer resources at that minimum you describe both before and after this tactic is used (and during!) - no violation there.

I don't think "designer intent" matters at all. If you read my postings, the only thing that matters to me is how the possible AI manipulation affects gameplay.

You are mistaken. Several times on this thread you do discuss the intent or purpose behind the mechanism. Here, I'll quote a few for you:
- "None of the people involved would deliberately design a mechanism where players can get a big advantage against the AI through a counterintuitive tactic of giving them extra gold---I know them well enough to know that." (post #154, and did you mean personally you know them? all?)
- "It's important to have multiple overlapping mechanisms to protect the AI against manipulation." (post #164)
- "This probably would have been adjusted in playtesting...[a]nd so it didn't get as much attention or tuning as it should." (post # 164)

It is heartening to realize that you aren't just not listening to me, you're also not listening to you.


- O

 

[DaviddesJ]

Exploiting AI weaknesses is not a regulable issue.

Of course it is. The GOTM staff can make whatever rules they want.

I don't see any reason to respond to the rest of this. I don't know whether you're deliberately misunderstanding, or not, and I don't really care.

 

[occam]

The GOTM staff can make whatever rules they want.

No, they can't, they can only choose from the spectrum of possible rules. What would a rule banning people from exploiting AI weaknesses in troop placement even look like? Give me a few words of how to write that rule, lol.

Of course, I would also take issue with your insulting implication that the GotM staff isn't transparent and procedurally honest about how decisions are made -- why can't we just safely assume that they will make rules in the same old, usual fair and forthright fashion, and reason from there? Frankly, if you must challenge that assumption to argue something is an exploit, I predict a lack of success. But good luck with that!

- O

PS: I find the way the GotM staff makes rules to be very predictable.

 

[DaviddesJ]

No, they can't, they can only choose from the spectrum of possible rules. What would a rule banning people from exploiting AI weaknesses in troop placement even look like? Give me a few words of how to write that rule, lol.

It's your strawman. You would have to write it, not me.

Of course, I would also take issue with your insulting implication that the GotM staff isn't transparent and procedurally honest about how decisions are made -- why can't we just safely assume that they will make rules in the same old, usual fair and forthright fashion, and reason from there?

The GOTM staff can make rules against such exploits in a fair and forthright way, just as they have always done in the past. There's nothing "insulting" about that. Your attempt to insinuate that any rule that you don't approve of, must be dishonest or unfair, is pathetic.

Since you joined this site only last year, I have about five times as much experience with GOTM rulemaking as you do. I can confidently say that I've never seen the GOTM staff in a situation where they felt a particular tactic was an exploit that should be banned, but they declined to do so solely because the issue was considered "not regulable". Whatever that means.

P.S. If "not regulable" just means "It's hard to write a rule against it," then that certainly doesn't apply here. One can easily write a rule against giving the AI extra gpt just so you can trade with it and then cancel the gpt gift. This is no less clear than many Civ3 rules.

 

[Ribannah]

An example of such a rule would be to forbid the following exploit: you declare war and do nothing else but move an archer onto a forest near one of there cities. Now, without any real effort or skill, you have crippled the AI to the point that it will never truly recover.
Another one is to declare war on an AI that cannot get to your territory or vice versa. It will often stop building wonders and you will be able to get them instead. It may even give you significant tribute for peace (in Civ3 it was very easy to get a city on an otherwise unreachable continent that way). Again, without any effort or skill on your part.

 

[DaviddesJ]

An example of such a rule would be to forbid the following exploit: you declare war and do nothing else but move an archer onto a forest near one of there cities. Now, without any real effort or skill, you have crippled the AI to the point that it will never truly recover.

Are these examples of bad rules that the GOTM staff could implement, but would be well advised not to? Or are these examples of rules that you think they should adopt? I can't really tell what your point is.

 

[occam]

I can confidently say that I've never seen the GOTM staff in a situation where they felt a particular tactic was an exploit that should be banned, but they declined to do so solely because the issue was considered "not regulable". Whatever that means.

Don't get hung up on the word 'regulable' -- my claim is quite clear in post #182:
An AI weakness does not an exploit make.

That is my prediction of how the GotM staff sees things. It takes just a little more substance than targeting some AI weakness, after all the weak AI is why we play with handicaps.

Your reply used to be:
The GOTM staff can make whatever rules they want.

But now we agree that the GotM staff will predictably not make such a rule based on just targeting some weakness in the AI?

I have gone on record with my prediction about the GotM staff's rules... sorry, but I don't know what your prediction is? Please forgive me if at first, it seemed that your stance was they were hard to predict -- and would make whatever rules they wanted. (You only say that in post #154 AND post #183. )

Since you joined this site only last year, I have about five times as much experience with GOTM rulemaking as you do.

Ah, but how do you know what posts I have gone back and read? Even your ad hominem attacks are poorly reasoned.

- O

 

[occam]

An example of such a rule would be to forbid the following exploit: you declare war and do nothing else but move an archer onto a forest near one of there cities. Now, without any real effort or skill, you have crippled the AI to the point that it will never truly recover.
Another one is to declare war on an AI that cannot get to your territory or vice versa. It will often stop building wonders and you will be able to get them instead.

Ribannah, thank you for the honest effort -- it is much appreciated... and now I feel guilty, for I am afraid that I miscommunicated the challenge...

I am claiming that a rule that is impossible to write is: never exploit the AI troop placement weakness in any way. Not some PARTICULAR way of your choice. In other words, when you attack you can't target poorly defended strategic locations, when you are attacked you can't concentrate your quick reaction force to the poorly thought-out point of attack... no sinking the weak amphibious assaults... etc. No second order manipulations based on your general garrison level, how you calculate forces required to attack, city placement, etc. Now do you see that the GotM staff can't really write such a rule without eviscerating Civ?

But let me save you any further effort by naming some of the future hurdles in the "Can Gotm Staff Make This Rule?" lineup.... Self-enforcing rules. Self-contradictory rules. Rules that contain sufficiently strong axiomatic systems that are simultaneously consistent and complete. Rules that are kept secret. Rules with no words. You get the idea, have fun with it.

The general point is:
Saying the GotM staff can make whatever rules they want just doesn't sound very thoughtful, since it is trivially false -- some rules just can't be made.

=====

Since that is such a such a minor point, I am abandoning it henceforth, and you may have the last word on that subject if you so desire.

- O

 

[DaviddesJ]

Saying the GotM staff can make whatever rules they want just doesn't sound very thoughtful, since it is trivially false -- some rules just can't be made.

No, it's trivially true. They can make any rule. If it can't be written down, then it's not a rule, after all.

I'm very much in agreement that there are many things that can't effectively be banned, and we shouldn't try. (As you undoubtedly know, from reading the last several years of GOTM threads. RCP is a good example of something that seems impractical to ban.) But the gpt exploits certainly don't fall into this category.

I think the purpose of this thread is to discuss what the rules should be, not to play a guessing game about what rules the staff will or won't adopt. But the latter is pretty easy to answer. I think that if the gpt exploits come to play a big role in many submitted GOTM games, there's likely to be some sort of rule adopted. Until it is shown that they make a big difference in practice, there won't. So the most important question is the empirical evidence, not the philosophical arguments.

 

[Gazaridis]

Quick question about a chopping 'feature'.

Say you chop something and have a lot of hammers left over. There is a way for you to be able to 'save' those hammers (by having the city not produce any hammers) until a few turns later and then spend those hammers then.

Is this disallowed or allowed? You do lose the hammers you gather during those turns. But it does enable you to rush out say a unit that only becomes available a few turns later.

Also, slightly related, is it OK to chop when what you will be producing will only take one turn anyway? This means that if whatever's next is under 30 hammers then that comes along in one turn one whipped person rather than 2 turns or 2 whipped people.

 

[DaviddesJ]

Say you chop something and have a lot of hammers left over. There is a way for you to be able to 'save' those hammers (by having the city not produce any hammers) until a few turns later and then spend those hammers then.

I don't think there is. Perhaps you should explain exactly what you're doing.

 

[grangerm]

You can save overflow by producing wealth, research or culture. I'm not sure if there's a decay over time like there would be for regularly produced shields, but I tend to doubt that.

I would say there's nothing wrong with doing that.

 

[Thrallia]

one thought in regards to the gifting exploit...if it were a 'feature' and not a bug, shouldn't it be present in warlords? cause in warlords, you can trade a resource to an AI for say, 5 gpt, and then before you leave the negotiating table, they are suddenly able to trade 3 gpt again.

Obviously, that means that they are able to offer more, they just choose not to.

 

[Shillen]

one thought in regards to the gifting exploit...if it were a 'feature' and not a bug, shouldn't it be present in warlords? cause in warlords, you can trade a resource to an AI for say, 5 gpt, and then before you leave the negotiating table, they are suddenly able to trade 3 gpt again.

Obviously, that means that they are able to offer more, they just choose not to.

I don't think that's conclusive evidence. If they get a new resource they can likely work more tiles than they could before, so that could be the reason they now have 3gpt available. I think we'd have to look into it a bit more before saying Firaxis attempted to fix the problem with Warlords.

 

[DaviddesJ]

There's definitely a deliberate change in Warlords, where the AIs re-evaluate their budget after each trade, and not only once a turn.

This doesn't prove that gifting is an "exploit".

 

[Thrallia]

Would the amount of cash and gpt you can see an AI has before and after they are your vassals be enough evidence?

I played a test game for WOTM1, and warred against Brennus. Just before he capitulated to me, in diplomacy I could see he had 3 gpt available and 280 gold in his treasury. IMMEDIATELY after he became my vassal, on the same turn, I saw he had 324 gold in his treasury and had 19 gpt available.

I would contend that that shows that the amount of gpt and gold they show in diplomacy is only the amount they are willing to part with at that time, not the total amount they have available. This in turn, would mean that the AI are choosing to not offer more than 3 or 5 gpt for a resource unless really really want a resource(at least, in Warlords)

 

[AlanH]

Couldn't it also mean that they were able to change their sliders as a result of the change from war footing to vassal status?

 

[DaviddesJ]

I would contend that that shows that the amount of gpt and gold they show in diplomacy is only the amount they are willing to part with at that time, not the total amount they have available.

I don't understand this distinction. We know how the AI manages its economy. (You can look in the SDK code if you want.) They allocate a certain amount to research, maintenance, etc. If the balance is positive, then this is what they show for trade.

The difference in Warlords is that, when you make a trade with the AI, they immediately re-adjust this allocation. So, if you give them gpt, that doesn't mean their available gpt will go up. It might go down, if they increase their research rate.

 

[Thrallia]

I guess it could mean they changed it, although I'd be surprised if it changed by that much before we even left the negotiating table.

Either way, I have been tending to be more warlike in Warlords, so that I can get vassals who have to gift me enormous amounts of gpt for the rest of the game.