Discussion thread for exploits and strategies

[DaviddesJ]

Is it an exploit because the AI will not steal Workers from you? Is it an exploit because the AI will not retaliate and kill you because it has superior forces (in terms of quality and quantity)? Is it an exploit because, if initiated early enough, the AI will accept peace if it loses one of its units?

All of the above. It's an exploit (by my definition) because it provides a disproportionate advantage. It provides a disproportionate advantage because the AI doesn't defend sensibly against it, and responds stupidly. If the early DOW and worker steal triggered the AI to launch all-out military buildup and retaliation, it obviously wouldn't be very effective. But the AI is programmed so that it just accepts peace and then lets you steal a worker again.

It's detrimental to the game because it just means you have to give even higher handicaps to the AI to make the game challenging, if you use the strategy. From my point of view, the game is more fun if the human has relatively smaller handicaps with respect to the AI, but there are fewer loopholes. I think this is a common but not universal view.

By the way, did you design or test the game?

No. But you might see some moderate similarities between my views and, say, Sirian's (who did). [Note: I'm not claiming Sirian would endorse my views on this particular point, or any other.]

 

[kingjoshi]

Eqqman, why are you so hard defending this isn't an exploit? What is your definition of an exploit? I ask because I want to know what "techniques" you think should be banned, if any. The 'technique' mentioned by DaveMcW above which was determined as an exploit and banned by ainwood, do you agree with that?

I do think this should be banned and is an exploit. At first, I couldn't finger exactly why and some things you always have a general feel for but can't explain. The fact that the AI is completely helpless to this is an important factor, IMO.

If the AI is helpless against the worker stealing technique as in the subsidizing trade, then that too is an exploit. But it seems that sometimes, Civs protect their workers from the start. Maybe it's a risk they take. I don't know if it's the leader trait or what factors come into play for the AI to decide that. Also, the success of the technique depends upon many other factors (being able to find workers, protect the worker on the return, being able to end war, etc) but even then, it is still somewhat of an exploit.

But with the trade subsidies, basically a loophole in the AI programming was found and is being exploited. And there is no possible way for the AI to adapt (with the worker-stealing, the AI later will protect all their workers, or at least most do).

A loophole was found, using subsidies, that overrides the AI's programming on how much money they should spend on trades given their current GDP/etc. It handicaps the AI's gold/research in a way they are completely helpless.

You may have no problems using a bug, glitch or an exploit in a game, but recognize the common way in how the term 'exploit' is being used here. When playing NFL2k5, there was a 'glitch' in which you could call a defensive play and choose a defensive end and reposition him and then attack and get a sack almost every time. It was because the offensive linemen were helpless because of how the AI was programmed in who they should guard and how. It's not a bug per se, but using that is definitely an exploit.

One thing in many football games that some consider an exploit (some disagree) is when they make a quarterback ridiculously fast that defending against them is nearly impossible. Vick, as incredibly fast as he is in real life, cannot do the things they made him do in video games (and Randall Cunningham and others before Vick). But when playing a game (in person or online), many groups of people would have rules stating you can't play as that team or you can't use that player in certain ways. The reason is because of the imbalance and faulty way the game was created. One game, you couldn't play with the Ravens' defense. The game the designers created had flaws and as a community, we agreed on rules on some of them (and some had more controversy).

Here, the trade-subsidy is a clever technique indeed. It definitely uses some of the strategy open to the player (trading resources) but goes beyond just that. It finds a hole in the programming of the AI and exploits it. It's not just exploiting a strategy any Civ can try (CS slingshot), or finding strengths and weaknesses about leadership attributes, finding a clever way to go through the tech tree, etc. It's finding a loophole in the AI programming. You may be fine with using it, but I don't see how you can't see that it's an exploit.

 

[malekithe]

Mostly good stuff...

...Here, the trade-subsidy is a clever technique indeed. It definitely uses some of the strategy open to the player (trading resources) but goes beyond just that. It finds a hole in the programming of the AI and exploits it. It's not just exploiting a strategy any Civ can try (CS slingshot), or finding strengths and weaknesses about leadership attributes, finding a clever way to go through the tech tree, etc. It's finding a loophole in the AI programming. You may be fine with using it, but I don't see how you can't see that it's an exploit.

I would argue that any game in which your opponent is an AI reduces to an attempt to find loopholes and ways of exploiting that AI. How do you beat a person at chess? You find a weakness in his play. How do you beat an artificial opponent in chess? You find a weakness in his programming. Why do we use stacks of death in Civ 4? Why do we reach friendly relations with an AI and then basically ignore any defenses against them until we backstab them? It's because we're able to infer and find weaknesses in their game, their programming.

However, precisely because this is a game, and involves friendly competition (one of the defining characteristics of a game), we're able to decide which exploits we will allow and which we will not while operating within the confines of friendly competition. This is when abuse of an exploit can turn into cheating. If man has created rules that outlaw a particular behavior, breaching those rules (in the context of a competition) is considered cheating. The important fact is, absent the man-made rules, the simple act of exploitation cannot be considered cheating.

So really, what this thread attempts to define is those acts of exploitation that will be deamed cheating for the purposes of this competition. There can and will exist exploits that will not achieve this status, for a variety of reasons. The reasons for not banning an exploit can take many forms. It could be deemed so difficult or take enough time to enact as to be considered "strategy" by those organizing the competition. It could be ineffectual to the point of giving no decided advantage to those who choose to engage in the exploit. Or, it could simply be difficult to define or enforce. Stacks of death, would fall into this last category, as would many of the other things that numerous people feel the AI should be able to handle better. Worker stealing belongs in this category as well, along with, in my opinion, the trading exploit.

 

[DaviddesJ]

Or, it could simply be difficult to define or enforce. Stacks of death, would fall into this last category, as would many of the other things that numerous people feel the AI should be able to handle better. Worker stealing belongs in this category as well, along with, in my opinion, the trading exploit.

How is, "Give extra gpt to the AI, so that you can trade for more gpt in return, and then cancel the first deal asap," hard to define?

 

[malekithe]

How is, "Give extra gpt to the AI, so that you can trade for more gpt in return, and then cancel the first deal asap," hard to define?

In your definition, you're including intent. Is it still an exploit minus the inent to benefit? If, through some serendipitous means, I manage to get more from an AI than I otherwise would, is that still considered an exploit? Say I had signed a peace treaty recently, gifting him some amount of GPT for 10 turns. Then, I trade a resource to him for more GPT than he would have had without the GPT deal. My intent may have just been to renew pre-war trade deals. The effect was I got more GPT from him than I would have. To boot, the peace treaty deal is guaranteed to expire through no involvement of the player.

In my opinion, it runs into similar problems to the worker steal, you have to gauge the inent of the "exploiter". Any attempt to completely ban the behavior, would necessarily restrict otherwise "legal" scenarios. This doesn't mean that it shouldn't be banned, just that I've struggled to come up with a suitably narrow definition for the physical manifestation of the behavior in question. Many behaviors can be defined through intent, but intent is a very difficult thing to police. It doesn't work very well when creating a rule-set (and it often results in laughable enforcement, ie. Basketball's intentional foul rules).

 

[DaviddesJ]

In your definition, you're including intent. Is it still an exploit minus the inent to benefit?

The "exploit" is giving the AI gpt for no benefit, in order to get it to pay you more. Intent is irrelevant: what matters is that you're giving the AI gpt. The definition is clear, imho. If you happen to be giving gpt to the AI for some actual reason, and you get a benefit where it gives you more gpt in return, that's not a problem.

 

[Eqqman]

What I mean to say is that the whole intention is to promote diversity. In some instances, certain actions would promote a more uniform game between all players. Chop-rush, for example, was so strong as to almost necessitate a chop-rushed start. To an extent, post-nerf, it still does, but certainly there are cases when a decision must be made (health benefits/production vs "free" hammers).

As I just said (maybe not directly enough), I don't think this particular technique removes diverse strategies, but creates them. The primary benefit is to put more cash in my pocket, which I can then use to employ in the strategy of my choice. There are few strats I can attempt mid- to late- game that won't benefit from me having more money in my pocket. Sure, 'everybody' will be 'forced' to use this method to get themselves extra money. Then they will use this money to do whatever they like, giving more diversity than you would have if everybody was broke.

Eqqman, why are you so hard defending this isn't an exploit? What is your definition of an exploit?

I already said why in my lengthy footnote, and I no longer have a definiton for that word nor do I care what anybody else chooses to use for a definition. Getting into the nature of exploitation was a mistake I made in the other thread that I regret now. All that matters to me is that the game has certain techniques you can use to facilitate play, which may or may not warrant banning by players, that should be discussed on a case by case basis.

If you want to know why I'm working hard to defend this particular technique,

1) it makes the game more enjoyable for me to play, and
2) does not make the game so easy for me to play that I am no longer challenged.

As an example, I'm still unable to win the game at Monarch level. It seems to me something truly ban-worthy would be egregious enough that any crappy player can be a decent or great player, not one that helps medicore players be only slightly less mediocre. I'm highly concerned about a group of well-intentioned players complaining to Firaxis that they need to 'fix' this, which has ramifications well beyond the game of the month. You can't very well tell people to 'do whatever you want in your own games' when you've taken that option away from them in a patch.

I ask because I want to know what "techniques" you think should be banned, if any

I'll say there are some things that do seem inappropriate to use in a community setting and leave it at that. I'm not going to say more because I don't want to get boxed into a 'you're against 'X', therefore you should also be against 'Y' ' argument. I want to discuss this based on its own merits- I'm not really a black-and-white outlook guy, things should be kept to a case-by-case basis.

It handicaps the AI's gold/research in a way they are completely helpless.

Again, some people are making these statements with NO hard data whatsoever. Please do not read ANY hostility into this text that cannot accurately convey my tone of voice or body language. How many games have you played where you actually used this technique and tried to squeeze every coin out of it? What was the demonstrable impact on the AI? What was the demonstrable impact on you? Did you win faster than you normaly do? Was it significantly faster? Was this a Noble win or a higher level? I've given explanations in the other thread as to why the 'completely helpless' theory is not true. And I am still waiting for somebody to participate in this aspect of the conversation. This is the conversation that will benefit people reading the boards, not further esoteric discussions on the meaning of the e-word that go nowhere. I suspect most people claiming 'completely helpless' are just talking out of their *** and have no idea what the answers to the above questions are. Besides which, it is a sweeping generalization that is going to be nearly impossible to prove. I'm already convinced it's false; when I have AI players that get their GPT back in the plus column the next turn after I pull the subsidies, you just can't come to me saying how much I harmed them. If you want to pull up a saved game where you got the AI so far in debt that all its military units went on strike and you rolled over their now empty cities, then at last we've got some hard evidence to discuss. My follow up question will be: how easy is this to reproduce? In my opinion ban-worthy techniques are usually ones that can be done by players of any skill level at nearly any difficulty setting with nearly any map settings. If it takes expert-level play to really put the screw on, then less reason to ban it.

...that overrides the AI's programming on how much money they should spend on trades given their current GDP/etc.

I don't want to repeat myself ("too late!", the audience cries) but I'll very briefly summarize points I made in the other thread:

1) How much an AI player 'should' pay for a resource is completely subjective. Calculating this price involves numerous variables, one of which is the most tricky of all: opinion. There is really no correct answer to what is the right price, so I don't approve of banning a technique because people disagree on what the 'right' answer is.

2) Assuming 1) was not true and there was a correct price, it is infeasible to program the AI to keep this cash on hand every turn in the hopes that it can spend it on a trade. Even harder if the AI wants multiple resources. It is more of a screw to force the AI to keep cash on hand to pay whatever you think is the proper price than to have the AI spend time in a deficit. The programmers wrote code to allow the AI to pay a 'good' price for resources, which because of the inherently flawed trade system I cannot access. I'm dammed if I do or dammed if I don't. Choosing to use subsidies, I violate the 'intent' of the AI's budget. No subsidies, I can't participate in the intent to have the AI pay good prices. I'm curious as to how often the AI retains the cash to pay the max price on its own. I'm willing to go out on a limb and say nearly never- maybe 1% or less than the time. I've certainly yet to see it. Maybe you can catch it early game when the max price is low anyway because you haven't built up enough friendship.

So the summary is, in my opinion this is a technique I can use to fix a part of the game that is broken. I see a difference between this and the football example you give, but others probably don't. In situations like this, one side has little chance to convince the other. But just because you can't convince me doesn't mean I'm not open minded. People like to accuse others of not being open minded just because they are unconvinced. Should anybody start producing actual data to support terms like 'game-breaking' or 'completely helpless AI' I can't wait to take a look at it. But enough already with going around and around on the meaning of the e-word.

 

[malekithe]

The "exploit" is giving the AI gpt for no benefit, in order to get it to pay you more. Intent is irrelevant: what matters is that you're giving the AI gpt. The definition is clear, imho. If you happen to be giving gpt to the AI for some actual reason, and you get a benefit where it gives you more gpt in return, that's not a problem.

The other problem with that rule: Benefit is too subjective and open to interpretation. Rules need objective measures in order to have meaning. I could argue that you are, by the nature of the exploit, benefitting greatly though your gift to the AI. The benefit is: The AI has more money on hand. You run into the problem of having to define what it means to "benefit" from a trade.

Now, you could say that the exploit occurs when the only effect of an individual trade is an increase in the GPT available to the AI. However, there is always the (incidental, in the case of the explot) additional effect of increased diplomatic points. Now, we could possibly outlaw the use of GPT gifts to the AI. I, personally, have never come across an occasion where it would be useful, and wouldn't complain were that rule to be enacted. You could also outlaw gifting GPT to the AI the same turn you create a deal involving the receipt of GPT. That nearly shuts the window on the possiblilty for exploit, making an attempt very risky. In this case, the time relationship is serving as the substitute for intent. The assumption is: If they occur close enough in time, the intent must be exploitation. It could still outlaw some otherwise legal behavior, but those situations are so few and far between as to be virtually non-existent.

I'm also not entirely sold on the efficacy of the exploit. I've used it to some effect in a couple of games. The only time it had a noticable benefit was when the game's outcome was already a foregone conclusion. But then again, maybe I should move up to Deity and give it a shot there. Maybe then it could actually have some impact on the outcome of the game. I have certainly been unable to cripple an AI anytime prior to controlling well over 30% of the available land (and resources). It did, in the case of the most recent game of the month, speed my victory by *maybe* half a dozen turns. If that is grounds for banning the exploit, then so be it, but there are many other exploits (chopping bug) that have a much greater impact.

 

[Eqqman]

I'm also not entirely sold on the efficacy of the exploit. I've used it to some effect in a couple of games. The only time it had a noticable benefit was when the game's outcome was already a foregone conclusion. But then again, maybe I should move up to Deity and give it a shot there. Maybe then it could actually have some impact on the outcome of the game. I have certainly been unable to cripple an AI anytime prior to controlling well over 30% of the available land (and resources). It did, in the case of the most recent game of the month, speed my victory by *maybe* half a dozen turns. If that is grounds for banning the exploit, then so be it, but there are many other exploits (chopping bug) that have a much greater impact.

Thank you for mentioning this. This is the bottom-line issue we need to be discussing.

 

[DaviddesJ]

The other problem with that rule: Benefit is too subjective and open to interpretation.

I didn't think it was ambiguous. You get a benefit from a trade when you get something in return. Giving gpt to the AI (either for nothing, or as part of a deal where you offer more gpt than the minimum that it would accept) is not a benefit. I guess you could claim that you're giving gpt to the AI for the diplomatic benefits, but I've never seen or heard of anyone doing that; I don't think it's a serious argument.

 

[DaviddesJ]

As I just said (maybe not directly enough), I don't think this particular technique removes diverse strategies, but creates them. The primary benefit is to put more cash in my pocket, which I can then use to employ in the strategy of my choice.

If there were a special key sequence you could hit to gain 100 gold, that would also put cash in your pocket that you could use to exercise more strategic options. It wouldn't be good, though.

As an example, I'm still unable to win the game at Monarch level.

So play at Prince level. Why is it more fun to you to take advantage of this loophole, and then give the AI a higher handicap (Monarch instead of Prince) to compensate? Fewer loopholes and less handicap is clearly better, imho, if it results in the same level of challenge.

How many games have you played where you actually used this technique and tried to squeeze every coin out of it? What was the demonstrable impact on the AI? What was the demonstrable impact on you? Did you win faster than you normaly do? Was it significantly faster?

I don't see why it matters. If it's very little help, then your arguments are that much weaker, because you say you don't want it banned so that it can help you win. If it doesn't help you much, then what's the strength of that argument?

In my opinion ban-worthy techniques are usually ones that can be done by players of any skill level at nearly any difficulty setting with nearly any map settings. If it takes expert-level play to really put the screw on, then less reason to ban it.

I think this is exactly wrong. If the technique is primarily useful to very expert players (and some of the Civ3 exploits fell into precisely this category), that's all the more reason to ban it. The ban doesn't affect the large majority of players, and the expert players would all (well, almost all) rather compete with each other on an equal basis without using major loopholes.

Banning techniques that are accessible and valuable to average or novice players is more problematic, because there's a larger number of people who are affected, and they are generally less knowledgeable about the game and find it more difficult to keep track of what they are and aren't allowed to do. One of the main arguments against adding banned exploits for the GOTM is precisely that it makes it harder to draw in new (inexperienced) players.

 

[Eqqman]

I don't see why it matters. If it's very little help, then your arguments are that much weaker, because you say you don't want it banned so that it can help you win. If it doesn't help you much, then what's the strength of that argument?

That is the argument. With a low impact there's no point in banning it. Some critics are claiming that the impact is big; in particular, that the AI is somehow being crippled, but nothing is coming out to back up these claims.

 

[Ribannah]

I don't see how the fact that you need it to win at Monarch makes forcing everyone to do as you do a good idea.

Start playing the game instead of the exploits, and you will be able to win at Monarch soon enough. And if some level of difficulty stays too hard for you: that's why we have them in the first place, to give everyone the opportunity to enjoy the game at their level of preference.

 

[Eqqman]

I don't see how the fact that you need it to win at Monarch makes forcing everyone to do as you do a good idea.

Start playing the game instead of the exploits, and you will be able to win at Monarch soon enough. And if some level of difficulty stays too hard for you: that's why we have them in the first place, to give everyone the opportunity to enjoy the game at their level of preference.

This comment seems inappropriate. Since I've stated that I haven't won, the fact that I 'need it to win' is obviously not true. I also don't follow how NOT banning the technique forces anybody to play the way I do. I don't see how what you've said here has any bearing on whether or not this technique should be prohibited in GoTM.

 

[DaviddesJ]

With a low impact there's no point in banning it.

I disagree. As I said, if the benefit is low (most of the time) then that works in favor of banning it (if the other factors support that). Some previously banned exploits fall into precisely this category: they are only occasionally useful at all, and that makes them easy to ban because most of the time no one is affected at all (and if you are in the case that would be affected, then you definitely know it).

 

[kingjoshi]

@malekithe: I'm not so concerned about enforcing the rules. Following the rules should be about integrity. I can have two computers, have cIV running on one and anytime I am curious about something, make a save. Then run the save in the other computer so it never shows up that I've replayed the turns. There are many ways to "cheat" that are completely undetectable. Even real laws have loopholes. We have to entrust the players to follow the spirit of the rules the community has placed. Of course there will always be some that break the rules, but that's not reason enough to not have them.

@Eqqman: When I say helpless, I don't mean you automatically cripple them. I hate to use another football analogy, but that's what I know. Anyway, some teams/coaches script their first 15 plays of the game. Now, if the quarterback didn't have any option of audibling the play and the defense somehow found the list, you would say the offense is helpless. Of course, if the offense is an NFL team and the defense is a high school team, it wouldn't make a difference. But it's still a type of exploit (to steal signals, plays, etc).

The outmanuever is not on the field of play. It's coming up with some trick outside the common bounds of competition for that edge.

@malekithe, Eqqman: Programmers make AI all the time that they can't beat. Whether it's a simple game like Chess or more complicated. But using a constraint on the AI (not being able to break deals even after your gold drops or not recognizing that disparity of income and trade) is definitely finding a loophole that has little to do with finding a weakness in strategy or otherwise.

It's not critical to me whether it's banned or not. Just stating my position. All subject to change with new information of course.

 

[Ribannah]

I've heard this argument from other players of other games, and it disproves itself. I'm not sure what to call it that doesn't give offense or imply that anybody using it is a weak player, but I'll call it the 'I want to be in a competition where I don't have to compete' school of thought.

Let's say we agree to play on the same map, but for whatever reason it is a fact that I am a better player than you are because of 'X'. Maybe it's micromanagement skills, or stategic planning, or tactics. 'X' means that the odds are I am going to outperform you in average circumstances. We'll assume to keep things simple that every game we play is under such circumstances. You're concerned that my use of certain techniques (I'll elaborate more on what it might be proper to call these later*) is going to force you to apply the same techniques and make our games the same. However, our games will not be the same since I still have 'X' in my corner.

No, that is not my concern at all. My concern is that I cannot compare my strategy and skill to alternative strategies and skills if the alternatives are tainted. My more important concerns are that because of that, the interesting players will stop competing, and a whole generation of new players will not be able to enjoy the game to the fullest.
Furthermore, I am NOT concerned that you might in any way be a better player than I am (if that mattered at all to me, which it doesn't). Because you rely on exploits, you will simply never have an 'X'. (By the way, the difference between a good and a poor player is never an 'X'. It is always the attitude.)

... It's amusing that this argument never goes away and is probably as old as time. It comes up in almost EVERY GAME EVER MADE.

That is correct. There are always exploit seekers, and therefore we will always need rules to prevent them spoiling the game for the rest of us.

It is relevant because the designers intended for a game that is balanced and offers many different strategies.

You may be thinking that if you're 'forced' to follow what the expert does, there is less strategy. I don't have a hard and fast answer to this.

Not the expert, but the exploiter.
Not just less strategy, but also fewer tactics, poorer immersement, less replay value, etc.
There is no rebuttal to this, period. This is what it all comes down to.

I've written at length in the other thread why it's fruitless to argue over how the AI should or should not be managing its trades, and how I don't see how the programmers could have reasonably been expected to come up with anything better

I know that you don't see that, and you probably never will see things like that if you keep your present approach to gaming. So, you may just need to trust the game programmers / designers and the experienced players on this.

Since I've stated that I haven't won, the fact that I 'need it to win' is obviously not true.

It is, it just isn't enough yet. You'll need more exploits. Now instead of that, if you didn't spend so much effort on exploits but instead played the game as intended, you could probably have mastered Monarch level already. It's not THAT hard.

I also don't follow how NOT banning the technique forces anybody to play the way I do.

Yes, I noticed that.

 

[Eqqman]

Furthermore, I am NOT concerned that you might in any way be a better player than I am. Because you rely on exploits, you will simply never have an 'X'.

If I gave offense for implying I genuinely thought I was better than you or anybody else I apologize. I was merely creating an example, in the future I'll cast myself in the role of the weaker player.

It's presumptuous of you to assume anything on how I play. Just because I enjoy fooling around with subsidies doesn't prevent me from trying to improve my play across the board. Being unsuccessful at my current goals doesn't mean I won't be in the future whether I continue to use this or not. Saying I 'rely' on anything is an untrue assumption on your part. The truth is I don't rely on much since I seldom successfully get strategies explained by experts to work for me on a consistent basis- my games almost never start out the same way twice since I'm always trying different things to find something that works for me.

There is no rebuttal to this, period.

The rebuttal is that making this statement doesn't make it true. This is just like other issues such as the existence of God or the validity of the Iraq war. People on either side will go to their graves convinced they are right. If you want to go on telling me I'm absolutely wrong and you're absolutely right I won't tell you to do otherwise, a courtesy some folks on this forum don't seem to think I deserve. And naturally, I don't deserve it since I'm absolutely wrong and they are absolutely right!

 

[Ribannah]

It's presumptuous of you to assume anything on how I play.

You have been telling us how you play all along. I don't have to assume anything.

The rebuttal is that making this statement doesn't make it true.

It doesn't have to, it was already true before I told you. I have also told you how these things work, and how you can improve your own experience and enjoyment. More that that I cannot do for you, and in the end it is up to the people running this competition to protect it.

 

[pigswill]

As I understand the trading strategy its about getting the AI to pay their maximum for every resource you trade to them; you can argue its an exploit or you can argue that its simply a way of maximising your benefits from trade.
If the designers are that bothered about it then at some point they could add diplo penalties for 'you're ripping us off' in the same way you currently get diplo bonusses for the ai ripping you off. If the Gotm team think its borderline they'll ban it from Gotm.
Until such time its down to individual players to decide wether they want to use this strategy in their games in the same way that worker stealing is not compulsory (and neither are stacks of doom).