What do you consider an exploit?

[KevinMiles90000]

The game exploits me, I exploit the game.

No apologies.

Gangsta.

 

[Blue Oyster]

... I'm curious what the CFC members think...I'm not judging those who do such things, but it's not for me.

What's exploitative to you?

Seems like a fair question. It is not a call to ban something, just a request for information. Clearly, opinions vary on what is an exploit. That is the point of the question. People should play in anyway so the game is fun for them. My conjecture: Few people will think it is fun if they use a maneuver that they actually believe is an exploit.

I wrote a learning Tic-Tac-Toe program in 1976, took three Computer Science degrees in Artificial Intelligence, and my career is in its 4th decade. The definition of what is "AI" has changed over the years, and I hesitate to apply it to most, perhaps all, computer games. (Chess a notable exception.) But that is my opinion, and doesn't have to be yours. Most AI is childlike. I think it is useful to examine what is "exploitive" from that perspective.

Example 1
Remember the board game Candyland? Any 4 or 5 year old could play it. I knew a guy who played with his child. He would stack the deck of cards so the child gained a big lead early, then slowly decreased, only to lose on the last turn of the game. He thought that was fun. I thought it was pathetic and, yes, a type of exploit.

Example 2
Ask a child of the right age which they would rather have, a nickel or a dime, and they will select the nickel. Apparently, at that age, bigger is perceived to be better. If you are dividing up money with a child, you can easily take all the dimes and let them have all the nickels. Some people would call that "fair" or a "strategy." To me, it's an exploit, and it isn't fun.

Example 3
Create a tax structure that make it appear a nonresident alien or foreign entity is the owner of assets and income, when in fact and substance, true ownership remains with a U.S. taxpayer. The IRS says this is an EXPLOIT of secrecy laws of offshore jurisdictions in an attempt to conceal assets and income subject to tax by the United States. I call this an accountant earning his salary. (Only kidding.)

The difference to me is in the first two cases you can pull the same trick over and over and the victim (OK, that's a pejorative term) doesn't stand a chance of learning how to defend themselves against it. That isn't fun, for me. And it's a trick I myself wouldn't fall for. OK, not more than 2-3 times. In the third example, you have lots of smart people with lots of resources trying to catch you, and the consequences are dire if you are nabbed.

So, if it's fun, do it, and why do you care what others think? Until, of course, the game learns how to defend itself, keeps track of your behavior from game to game, and warns all the other Civ installations in the world through Steam to watch out for you, and possibly other "AIs" refuse to play with you.

 

[bronzeager]

So, if it's fun, do it, and why do you care what others think? Until, of course, the game learns how to defend itself, keeps track of your behavior from game to game, and warns all the other Civ installations in the world through Steam to watch out for you, and possibly other "AIs" refuse to play with you.

excellent points, and this last bit, while obviously a joke, is exactly what the developers should be thinking of.

if you consistently break deals with the AI, so that you are essentially stealing their gold, the AI should be smart enough to no longer make deals with you. you should also take a relations hit with every other AI in the game. either that, or make it so DOWs are not possible until all trade deals have run their course.

 

[Louis XXIV]

You could also make it so if you sell a luxury to an AI for up front gold, you lose the luxury until the deal has run its course regardless of whether you go to war. I don't think that's ideal, but it's one way to address the problem.

The biggest problem is the AI isn't capable of fully appreciating your strategy and whether or not you're trying to bankrupt them in order to make war easier.

 

[Lyoncet]

You could also make it so if you sell a luxury to an AI for up front gold, you lose the luxury until the deal has run its course regardless of whether you go to war. I don't think that's ideal, but it's one way to address the problem.

I think this used to happen, but it happened on both ends. You'd still not have access to the luxury for the whole duration of the deal, and the AI would keep it. Some people complained that was unrealistic, but I thought it could be justified just fine in that you could just think of it as sending the AI 30 "turns" worth of luxuries at the beginning of the arrangement, while they distribute 1/30th of it per turn. (Makes less sense with GPT deals breaking.)

 

[TheMeInTeam]

Example 1
Remember the board game Candyland? Any 4 or 5 year old could play it. I knew a guy who played with his child. He would stack the deck of cards so the child gained a big lead early, then slowly decreased, only to lose on the last turn of the game. He thought that was fun. I thought it was pathetic and, yes, a type of exploit.

This is not an example of an exploit.

This is an example of cheating.

Granted, candyland sucks anyway because it is entirely luck-based (why more and more luck based outcomes are a trend in modern gaming escapes me), but it's still cheating to stack a deck of supposedly shuffled cards to get a certain outcome. Change "candyland" to "poker" and "child" to "serious individual with a gambling problem", and you're probably going to see some bullets .

Example 2
Ask a child of the right age which they would rather have, a nickel or a dime, and they will select the nickel. Apparently, at that age, bigger is perceived to be better. If you are dividing up money with a child, you can easily take all the dimes and let them have all the nickels. Some people would call that "fair" or a "strategy." To me, it's an exploit, and it isn't fun.

How this is relevant is anybody's guess. First of all, what is the marginal utility of either thing to a child too young to understand monetary value anyway? My nephew is 5 and he's understood a dime is worth more than a nickel since he was 4. What is the "right age"...an age where either of these objects is a choking hazard ?

Even so, by this logic everything is an exploit, including using military units and placing national wonders in the proper cities. After all, the AI lacks the knowledge necessary to make these decisions, so you are EXPLOITING it by beating it at war using tactics. I better not see you reporting that you've played the game and EXPLOITED the AI via war .

On the flip side, you could beat a child at virtually ANY competition...unless it was given severe handicaps. But if it were given those handicaps (as a way to compensate its lack of knowledge), this entire picture changes - presumably these handicaps were set to give him a fair chance against your ability. Are you saying that playing low difficulties is an exploit?

Example 3
Create a tax structure that make it appear a nonresident alien or foreign entity is the owner of assets and income, when in fact and substance, true ownership remains with a U.S. taxpayer. The IRS says this is an EXPLOIT of secrecy laws of offshore jurisdictions in an attempt to conceal assets and income subject to tax by the United States. I call this an accountant earning his salary.

Tax avoidance is legal. Tax evasion is not legal. There are clear rules on these matters, however, so again this is a bad example because it is an example of breaking rules, which is not an exploit. Breaking explicit rules is cheating. There is a difference you're missing in the formulation of these examples.

The AI has been given bonuses to compensate its lack of playing ability. Who is anyone to judge which actions these bonuses cover, and which they do not? There are no exploits by the logic framed above, only pretend rules for variant play.

 

[slobberinbear]

I appreciate all the replies. I find it interesting to see the community's perspectives and approaches to the game. We all play for our own reasons.

"Exploit" is a loaded term, so perhaps my initial question was bound to get some heated responses. Perhaps I should have couched it in gentler terms ... as in, "do you restrict your actions with respect to the AI, or is it anything goes?"

 

[TheMeInTeam]

I appreciate all the replies. I find it interesting to see the community's perspectives and approaches to the game. We all play for our own reasons.

"Exploit" is a loaded term, so perhaps my initial question was bound to get some heated responses. Perhaps I should have couched it in gentler terms ... as in, "do you restrict your actions with respect to the AI, or is it anything goes?"

Actually, I'm glad you used the term "exploit", because HoF uses it too and has used it to create rules without any basis that are in direct conflict with its standards (and are guaranteed to harm that competition). They're ALREADY banning things that for example MadDjinn has proven that the AI does. They're banning things and ignoring arguments about opportunity cost and relative value, it's a mess. It's refreshing to see more people that don't fall into the non-logic trap and as this forum is more well-traveled than that one, to get some fresh blood in the argument.

 

[Architect]

Well with regards to the HoF game don't the people that spend the time organizing it and looking at all the submissions have the right to define the rules for those games however they want?

If you want to run your own HoF type game each month that has a different set of rules I think you could do that here and it would probably be welcomed. I guess I don't understand why people are frustrated its not like HoF games are some sanctioned CiV competition that defines all CiV competition.

In other news I've changed my opinion on breaking deals with the AI through declaring war. When I first saw MadDjinn do that to china in one of his YouTube videos I thought "that's cheating!" Now after considering all these arguments and thinking about it myself its just part of the intended gameplay for official CiV no matter what the AI can or can't do about it IMO.

 

[Kevin J]

Well with regards to the HoF game don't the people that spend the time organizing it and looking at all the submissions have the right to define the rules for those games however they want?

If you want to run your own HoF type game each month that has a different set of rules I think you could do that here and it would probably be welcomed. I guess I don't understand why people are frustrated its not like HoF games are some sanctioned CiV competition that defines all CiV competition.

In other news I've changed my opinion on breaking deals with the AI through declaring war. When I first saw MadDjinn do that to china in one of his YouTube videos I thought "that's cheating!" Now after considering all these arguments and thinking about it myself its just part of the intended gameplay for official CiV no matter what the AI can or can't do about it IMO.

The problem is twofold.

1. The rules were changed on the fly. I had dozens of hours of work in on marathon time victories - one with Songhai and several with India at different difficulty levels. It's exceedingly frustrating to lose dozens of hours of work, especially when you only have a couple hours of gaming time a day. This so soured me on HoF & CiV in general, I've completely quit playing except for a quick game I spun up with the new DLC.

2. The rules amount to a ban on normal gameplay - especially on marathon speeds. Essentially, declaring war is banned unless you intentionally gimp your gameplay and don't trade.

Furthermore, any suggestions fall on deaf ears and result in locked threads. I've been coming to CFC since 1999 (despite my current username only showing a few years). The current HoF format makes me strongly resent one of my all-time favourite web-sites - I think it's an issue.

 

[KevinMiles90000]

I think a simple fix for the "selling all your luxuries/gpt to the AI then declaring war" is to have some kind of diplomatic penalty for doing so.

 

[Martin Alvito]

Exploit = Anything that reduces the complexity of the game so much that players would rather ban it than mindlessly abuse it.

This (from the HOF thread) is on the right track. The key point is that it's preference-based. An exploit requires three components. First, the action is a dominant strategy under certain conditions, such that you have to use it to be competitive with other humans if permitted. Second, the action shouldn't be possible, either because the action violates other game rules as stated or the AI reacts nonsensically to it. Third, the effects on the game are detrimental, in that permitting the action collapses the strategy space and forces the game down a narrow line of play.

A fourth component that isn't necessary for it to be an exploit, but is necessary for it to be actionable in a competitive format, is that the exploit is verifiable with reasonable labor input on the part of an enforcement staff. If we make rules that limit the player's ability to exploit the AI's inability to make war, then enforcement requires somebody to check every save of every war to make sure that the player didn't break the rules. That's too costly, so we don't try to make rules against the behavior. (We ask the devs for better AI, which amounts to making rules against the behavior!)

DaveMcV's quoted statement captures the essence of the third component of exploits. The location of that line is necessarily preference-based for any individual player. One player's detrimental strategy is another player's optimal play.

The result is that you all can argue until you're blue in the face on this, but you're never going to reach agreement. I like Sour Patch Kids, and I don't particularly like malt balls, but if you like malt balls and don't like Sour Patch Kids there is no way that I can argue you into liking Sour Patch Kids and disliking malt balls.

The HOF's staff's preferences get honored, and those of other players do not, because they run the competition. If you don't like those rules, you have three alternatives: play, start (or threaten to start) a competing competition and siphon away enough players to get some bargaining leverage, or don't play. It's that simple.

They're banning things and ignoring arguments about opportunity cost and relative value, it's a mess.

Your arguments aren't being ignored. You're just wrong. This is easy to demonstrate. Suppose that f(t) is the opportunity cost of declaring war on a civ, measured in diplomatic relations hit with other civs, foregone trading opportunities and the risk of loss in war. Since the AI will achieve win conditions on higher difficulties before the timer expires, let's assume that some t = 0 exists at which a win condition is achieved. Then t = -1 is the turn prior to the winning turn, and so forth.

It is obvious that as t -> 0 from the left, lim f(t) = 0. There are no more trading opportunities, the AI cannot engage in reprisals, and your diplomatic relations with other civs don't matter. It is also obvious that as long as you don't have enemy units within range or an RA that is about to conclude with a given civ, f(t) = 0 when t = -1.

The result is that the claim that war always has meaningful opportunity costs is false. Unless you have some combination of pending RAs and neighbors that never obtains in practice, there will exist some t < 0 such that f(t) = 0 for at least one civ. Under those conditions, players can and should declare war to break and remake luxury deals with distant civs that are not RA partners, and should sell excess (non-capital, non-parts) cities to those civs, reconquer and resell before declaring war if they have sufficient units and cities on hand to do so.

More importantly, this is going to be meaningful in Diplo and Culture games. You rarely can achieve sufficient capital population to pop the UN in a single turn with a GE, and Utopia can't be rushed, so rush buying production buildings with the proceeds from the DoWs is going to matter.

 

[Kevin J]

Your arguments aren't being ignored. You're just wrong. This is easy to demonstrate. Suppose that f(t) is the opportunity cost of declaring war on a civ, measured in diplomatic relations hit with other civs, foregone trading opportunities and the risk of loss in war. Since the AI will achieve win conditions on higher difficulties before the timer expires, let's assume that some t = 0 exists at which a win condition is achieved. Then t = -1 is the turn prior to the winning turn, and so forth.

It is obvious that as t -> 0 from the left, lim f(t) = 0. There are no more trading opportunities, the AI cannot engage in reprisals, and your diplomatic relations with other civs don't matter. It is also obvious that as long as you don't have enemy units within range or an RA that is about to conclude with a given civ, f(t) = 0 when t = -1.

The result is that the claim that war always has meaningful opportunity costs is false. Unless you have some combination of pending RAs and neighbors that never obtains in practice, there will exist some t < 0 such that f(t) = 0 for at least one civ. Under those conditions, players can and should declare war to break and remake luxury deals with distant civs that are not RA partners, and should sell excess (non-capital, non-parts) cities to those civs, reconquer and resell before declaring war if they have sufficient units and cities on hand to do so.

Martin, you're taking his argument to the extreme (i.e. end-game) to show where it breaks. Obviously, the turn before your game ends, there is no penalty for taking a diplo hit. However, this does not mean his argument is wrong.

There is a declining opportunity cost as the game proceeds to breaking deals via DoW, eventually reaching zero in the trivial case of the end-game.

Your line of reasoning would argue for banning trades in the last 30 turns of the game - a course of action that was decided *against* in the HoF - frankly, due to it being ridiculously inconvenient. You yourself argued against such a ban. It does not, however, make a strong case against banning all lump sum trades.

Yet the staff took the path of banning all DoWs when lump sum trades have been made - a rule that bans very normal, very frequent, and frankly, a quite often unavoidable course of action.

 

[slobberinbear]

The solution to the lump-sum trade argument seems pretty straightforward to me ... there should be a significant diplo hit (or at least an unwillingness to trade) against any party that does the lump sum/DoW trick.

I suppose that's what feels "exploitative" about it ... there are not currently enough costs associated with the manuever, IMO.

 

[Martin Alvito]

Martin, you're taking his argument to the extreme (i.e. end-game) to show where it breaks. Obviously, the turn before your game ends, there is no penalty for taking a diplo hit. However, this does not mean his argument is wrong.

The claim of the other side is that war always has meaningful opportunity cost. The point I'm making is that the claim is false. The existence proof above demonstrates that.

It doesn't necessarily follow that we should make a rule against the behavior. If we're going to make a rule against something, the following questions need a "yes" answer:

- Is it an exploit? (Yes, the AI has the information available to avoid getting suckered (era jumps, diplo screens, CS ally status, Apollo notification), but doesn't use it.)
- Is it enforceable? (Yes, via deal history)
- Does it have a meaningful, detrimental impact on gameplay? (Yes. If you reason backward through the game, acquiring excess cities in Diplo/Culture becomes a necessary component of a viable strategy. Luck factors are also enhanced, because the amount of cash you can fleece out of AIs matters, and that varies sharply and cannot be predicted.)

Again, notice the preference component of the third term. I say it has a meaningful impact that is detrimental, but you may not. I don't have a good answer to how we should resolve that social choice problem. Functionally, we live in a dictatorship because the staff's decisions are final. That's not necessarily such a bad thing; Kenneth Arrow won a Nobel Prize largely for proving the counterintuitive result that the only system of government that can maximize social welfare while meeting basic democratic standards in solving a social choice problem is a dictatorship. (Of course, that proof doesn't guarantee that it will do so.)

Your line of reasoning would argue for banning trades in the last 30 turns of the game - a course of action that was decided *against* in the HoF - frankly, due to it being ridiculously inconvenient. You yourself argued against such a ban. It does not, however, make a strong case against banning all lump sum trades.

How is that the unique solution to the problem? The issue I posed is this: players can get functionally free gold at the end of the game by making and breaking deals, then reselling. Banning all trades in the last 30 turns is unnecessarily restrictive. Breaking the deals is the problem. I see no problem with exploiting private information about the impending end of the game, since you can do that with humans too. It's the violation of an agreement of something now (gold) for something later (luxury, the / stream from a city) that's the problem. No reasonable human would agree to those terms.

Yet the staff took the path of banning all DoWs when lump sum trades have been made - a rule that bans very normal, very frequent, and frankly, a quite often unavoidable course of action.

What alternative do you propose that solves both the endgame problem and the flexibility problem?

The solution to the lump-sum trade argument seems pretty straightforward to me ... there should be a significant diplo hit (or at least an unwillingness to trade) against any party that does the lump sum/DoW trick.

Firaxis can do that, but AFAIK we can't. If this can be modded, we should do so. To be honest, it seems to me that a lot of the current HOF problems could probably be solved via mod rather than ban. For instance, you could probably adjust the values the AI is willing to pay as you advance eras or otherwise close in on win conditions. In the meantime, we have to figure out how to solve this problem amongst ourselves.

 

[vexing]

Yet the staff took the path of banning all DoWs when lump sum trades have been made - a rule that bans very normal, very frequent, and frankly, a quite often unavoidable course of action.

this isn't true. no such ban exists.

 

[Moss]

Moderator Action: The topic of this thread is exploits in the game, not a discussion about HOF game rules. Those discussions belong in the appropriate forum. Please cease this discussion and get back on topic.

 

[TheMeInTeam]

The result is that the claim that war always has meaningful opportunity costs is false.

The claim of the other side is that war always has meaningful opportunity cost. The point I'm making is that the claim is false. The existence proof above demonstrates that.

Utterly ridiculous. By approaching t = 0, EVERY TACTIC IN THE GAME approaches 0 opportunity cost, until it reaches 0 at t = 0. That is not a valid basis for refuting the claim that DoW consistently has opportunity cost. That a tactic is *sometimes* the dominant option is not a valid basis for ruling something bad for competition. It's also invalid to claim t = 0 is known until one is very close to it. You can't make a serious case that something is an "exploit" on this basis.

- Is it an exploit? (Yes, the AI has the information available to avoid getting suckered (era jumps, diplo screens, CS ally status, Apollo notification), but doesn't use it.)

Suddenly, every possible tactic that takes advantage of bad AI is an exploit. That's...not very useful

Or: Since this is an example of something the AI can do, who are you to say it "doesn't use it" as opposed to "doesn't use it well consistently"?

Yes. If you reason backward through the game, acquiring excess cities in Diplo/Culture becomes a necessary component of a viable strategy.

Yes, your fake rules add some minor strategy, but it also removes strategic usage of the "exploit" based on situation. In insisting on this t ---> 0 nonsense, you are still ignoring the point that war consistently has opportunity cost. That it shrinks along with all other costs near the end of the game is not relevant because it mirrors the tendency of most decisions. Where is the exploit?

No reasonable human would agree to those terms.

I hope you don't sell open borders and other crap to the AI . The AI plays poorly. It gets bonuses to compensate that; you're pulling a line from nowhere and pretending you have basis for it, which you've yet to show.

What alternative do you propose that solves both the endgame problem and the flexibility problem?

Problem? Play the game. The burden of proof lies on proving there is a problem. You haven't yet, doubly so if opportunity cost approaches 0 near end game is the best you can do.

Firaxis can do that, but AFAIK we can't.

Haha! The AI doesn't get mad when you do this.....

...

Except it does. Every DoW made by the player has an adverse impact on AI opinion of the player (setting him closer to being considered a warmonger). It's in the code. People have posted the code. You can only do this a finite # of times before the AI all hate you. You can say you don't like the balance and think the #times should be reduced, but you didn't make the game...and hardly have any more right to decide what's an "exploit" than the next person based on *balance* issues!

Nevermind situations where:

- AI breaks the deal
- AI asks player to go to war
- Going to war has considerable alternate incentives
- AI won't end war w/o receiving something
- AI can execute your "exploit" itself!

 

[NukeEm]

Historically, in gaming, exploits are either bugs or mechanics that are constantly used to gain a large advantage over an opponent and are usually mechanics that the developers did not intend.

Personally, I could care less what is or isn't an exploit in a single player game. When things get competitive, then either work within the predefined rules or walk away from it. The way I see it, is that a good player should be able to "win" within the intended design model.

On the "AI" side of things.. AI's should always be allowed to exploit, cheat and whatever else. Because artificial intelligence is laughable in gaming. It's either, predictable, bad or just plain overpowered. It's why Civ on increased difficulties is STILL bad... because AI in general is bad, and throwing bonuses on bad AI is like giving a gambler a million dollars... Sure, there's more to play with... but let's face it.. it's just more to lose.

So, for me, exploit = anything that uses game mechanics/bugs outside what the developers intended... outside the "spirit" of play.

 

[TheMeInTeam]

Historically, in gaming, exploits are either bugs or mechanics that are constantly used to gain a large advantage over an opponent and are usually mechanics that the developers did not intend.

This is more an urban dictionary definition though. The actual use of exploit as a term for tactics in gaming is confusing; generally exploiting your opposition is a good thing by the actual definition of the word.

anything that uses game mechanics/bugs outside what the developers intended... outside the "spirit" of play.

That sounds all nice and hugs and kisses, except that nobody can actually gauge developer intent on game mechanics (bugs/glitches are a bit different...). Using the "game mechanics outside dev intent" to determine actions in competitive settings is asinine; it will invariably lead to arguments/made up rules over what can be done. Join a MP lobby and you'll see what I mean...people will be all over the place on what is an exploit, completely ignoring the actual, allowed game mechanics.

If it's a competition, play the game, or make up pretend rules and call it something other than civ V.