Changing tile improvements exploit

[Trik]

I don't know if this is well known or not.

Building another tile improvement over your current improved luxury tile, it will cancel any trade you have with that luxury. Kinda like the pillage your own tile exploit except it take longer to get the luxury resource back online.

You could even go for pyramids too make this even more powerful.

Does anybody do this or do you consider to be too cheap?

 

[Misterboy]

Sounds very cheap to me. I would never do it. But that's just me.

 

[RealHuhn]

I already feel dirty when I resell a luxury after a barb destroyed it

It's a bad mechanic although your suggestion would take it to a whole new level

 

[TheMeInTeam]

This move never passes the "is there any strategic depth to this action" test. In PvP games, it doesn't matter; people will police this sort of thing outright. In single player, it's strong enough that you want to do it at every opportunity as the opportunity cost is generally just what the worker could have been doing instead of reworking the same tile (or possibly the cost of an extra worker for this); the returns are consistently enormous enough to dominate that.

Of course, if you want that edge in SP knock yourself out. It's kind of like world-buildering or any other such move but there's nothing wrong with messing around in SP anyway.

Incidentally, it escapes my why firaxis doesn't just drop you a "negative resource"; IE the deal continues anyway and you just have a negative value in the column for that resource until either 1) the deal expires naturally or 2) you hook it up again. THAT would permanently eliminate this "self or barb pillage to break deals" nonsense.

 

[joncnunn]

Actually just change it to where AI would never pay up front cash for per turn items would remove all the resource exploits cleaner.

 

[LostInTime]

I already feel dirty when I resell a luxury after a barb destroyed it

I don't go out of my way to abuse stuff like this myself, but when it happens like this or an AI does it I'll sell it again. Else some stupid AI will just come demand it from me for free.

 

[Steelyglint]

I'm surprised it works. Never really looked but I don't think I noticed that my workers COULD build anything on top of any sort of resource other than that resource's improvement.

 

[Strategist83]

Does anybody do this or do you consider to be too cheap?

Why would you cheat against the AI? Actually, why would you cheat at all?

 

[KrikkitTwo]

This move never passes the "is there any strategic depth to this action" test. In PvP games, it doesn't matter; people will police this sort of thing outright. In single player, it's strong enough that you want to do it at every opportunity as the opportunity cost is generally just what the worker could have been doing instead of reworking the same tile (or possibly the cost of an extra worker for this); the returns are consistently enormous enough to dominate that.

Of course, if you want that edge in SP knock yourself out. It's kind of like world-buildering or any other such move but there's nothing wrong with messing around in SP anyway.

Incidentally, it escapes my why firaxis doesn't just drop you a "negative resource"; IE the deal continues anyway and you just have a negative value in the column for that resource until either 1) the deal expires naturally or 2) you hook it up again. THAT would permanently eliminate this "self or barb pillage to break deals" nonsense.

Well, the problem is there is no "negative" luxury values [unless - luxury was -4 happiness]

The simpler method would be to eliminate lump sum gold trades, and allow someone to give per turn gold if they had a stockpile.

 

[TheMeInTeam]

Well, the problem is there is no "negative" luxury values [unless - luxury was -4 happiness]

The simpler method would be to eliminate lump sum gold trades, and allow someone to give per turn gold if they had a stockpile.

1. Someone already made this point
2. It's nonsense. Firaxis could easily code such a thing

Or are we going to get into joke discussions about realism in this context? Please don't sink this thread into that muck. I'm sure it wouldn't be hard to rationalize a reason.

Anyway denying lump sum for per turn trades yanks strategic depth out of the game AND alters its strategic balance. Why do we have multiple people jumping on that "solution" when a negative resource value would prevent the exploit without changing gameplay strategy at all? I'm curious.

 

[KrikkitTwo]

1. Someone already made this point
2. It's nonsense. Firaxis could easily code such a thing

Or are we going to get into joke discussions about realism in this context? Please don't sink this thread into that muck. I'm sure it wouldn't be hard to rationalize a reason.

Anyway denying lump sum for per turn trades yanks strategic depth out of the game AND alters its strategic balance. Why do we have multiple people jumping on that "solution" when a negative resource value would prevent the exploit without changing gameplay strategy at all? I'm curious.

lump sum for per turn trades is depth the AI can't handle (of course the AI can't handle luxury itself either)

Having a negative luxury resource value with -4 happiness would make it workable.

However, it would still leave open major other exploits with lump sum for per turn, that I doubt the AI can handle and doesn't add much strategic depth to the game.

 

[TheMeInTeam]

lump sum for per turn trades is depth the AI can't handle (of course the AI can't handle luxury itself either)

It can't handle war. It can't handle building city/tile improvements in proper order. It can't handle other trade options. It can't handle strategic planning.

The logical extension for "AI can't handle the strategy" is that a person should never play single player because the AI can't handle it, following that logic. The argument stance is nonsensical.

Having a negative luxury resource value with -4 happiness would make it workable.

Having a negative luxury resource value with + 417 happiness would make it workable.
Having a negative luxury resource value with -10 happiness would make it workable.

What the above 2 statements have in common with each other, and the quoted line, is that they are all completely arbitrary and pointless changes for having - of a luxury instead of 0 luxury.

The basis for - luxury is simple; pillaging it does nothing to the current deal; the AI keeps getting the resource. It just wastes worker turns. With that simple change, there is suddenly NO incentive to pillage (or allow pillaging of) luxury tiles. You take a slight penalty for no gain whatsoever.

However, it would still leave open major other exploits with lump sum for per turn, that I doubt the AI can handle and doesn't add much strategic depth to the game.

The line defining exploit is made up anyway. In a single player game you just do whatever you feel is most fun. Most of the other "exploits" carry significant cost to them, certainly much more than a few worker turns. Most of them are also similarly AI flaws. However THIS tactic would be completely blocked w/o changing game strategy one iota with my suggestion...

 

[Martin Alvito]

Let's put it this way:

The AI is indeed too stupid to properly evaluate some lump sum for GPT trades. However, removing them from the game is not a unique solution.

There are three types of lump sum for GPT trades right now:

- Resource (luxury/strategic)
- Open Borders
- Loan (its GPT for , or vice versa)

Nobody complains about Open Borders, as selling it to an AI invites a real risk of inviting a DoW that otherwise would not happen. The AI does a pretty poor job of putting a value on loans (it insists on making money on both ends, undervalues early game and doesn't recognize the risk of not getting paid back), but that's fixable if you tweak some scalars and make loans persistent even in a state of war. The latter may not be very realistic, but it's a necessary sop to the fact that AIs are often ill equipped to evaluate credit risk. I spent nearly five years with a major bank a number of years ago, and evaluating credit risk was part of my job. Trust me, I know.

The big problem is luxuries and strategics. Not because they don't confer something of value (more luxuries -> more and faster GAs, and more strategics -> more Deity unit spam), but because the trades can be terminated by the player free of consequence and warning.

One solution to that problem is to ban the trades. Another is to penalize the player rather than the AI for the deal's cessation. Either will work, but the latter is undeniably a more minimalist solution.

 

[TheMeInTeam]

I spent nearly five years with a major bank a number of years ago, and evaluating credit risk was part of my job. Trust me, I know.

I don't think it takes legit finance experience to realize the AI is bad with credit risk . "Hmm, this guy just reneged on deals the last 5 times he made them. What's this? He'll give me lots of gold/turn for a lump sum?! Sold!" Granted, I have some business training myself, but most people could probably figure somethings up with that behavior even without it I'd guess .

Another is to penalize the player rather than the AI for the deal's cessation. Either will work, but the latter is undeniably a more minimalist solution.

While I agree with this, I find it a bit over the top to get to the point of losing happiness beyond the trade itself if you misclick and get your tile pillaged by a barb or something. Locked resource trades would (minimally) penalize the player though...you have to go replace that improvement to use the resource again and receive no benefit for letting the tile get pillaged in return. Lacking any actual incentive for it, players wouldn't do it.

As for gold trades, a fairly simple approach would be for all AI aware of your presence during a broken deal to not make lump sum for PTO deals with the offending player for the rest of the game (including other AI that pull this crap, which they do on occasion ). This would put serious pressure on a player wanting to break such a deal; you can do it, but it will compromise future trade income...however in some contexts that might be worthwhile. Barring that, simply leaving the deal in place would suffice and be much more dynamic than banning that class of deals entirely.

I'm sure we can come up with an excuse for the continued gold loss during a war post-deal break if we want to be creative lol, but the important question is how good this is for gameplay? I think it would go a long way toward helping the AI play better than it does now...

 

[Martin Alvito]

I don't think it takes legit finance experience to realize the AI is bad with credit risk . "Hmm, this guy just reneged on deals the last 5 times he made them. What's this? He'll give me lots of gold/turn for a lump sum?! Sold!" Granted, I have some business training myself, but most people could probably figure somethings up with that behavior even without it I'd guess .

It's obvious with this AI. It's less obvious that the problem is intractable unless you've observed AIs dedicated to the problem in action. They function a lot like the computer on Jeopardy - they're very good for simple problems, but their errors are whoppers.

While I agree with this, I find it a bit over the top to get to the point of losing happiness beyond the trade itself if you misclick and get your tile pillaged by a barb or something.

You don't have to make things additionally punitive. You just have to make the player realize a cost for having the resource pillaged, and not have the AI realize the cost. Your suggestion would do precisely that.

As for gold trades, a fairly simple approach would be for all AI aware of your presence during a broken deal to not make lump sum for PTO deals with the offending player for the rest of the game (including other AI that pull this crap, which they do on occasion ). This would put serious pressure on a player wanting to break such a deal; you can do it, but it will compromise future trade income...however in some contexts that might be worthwhile. Barring that, simply leaving the deal in place would suffice and be much more dynamic than banning that class of deals entirely.

There are certainly simple algorithms that we could use, but the problem is that all of them can be gamed under certain conditions. Leaving the deal in place is the most elegant solution, though the realism is dubious. After all, one principal cause of World War II was Germany's continued efforts to resettle the peace deal at Versailles. In CiV terms, they forked over several cities, all their cash and all their GPT (and then some) for peace, then reneged on the deal after fifteen turns. You can do that in CiV now; if you make deals sticky through war, you can't.

I've been a strong advocate of adding an additional reputation dimension to diplomacy for some time now. The only concern that I have is that it would either be very easy to trash AIs' reputations (by loading them up on loans when you see their army on its way) or make diplomacy both tedious and much more transparent than the devs seem to want (deal refusal = WHEOOHRN, so just test the AI's deal willingness regularly).

 

[KrikkitTwo]

Let's put it this way:

The AI is indeed too stupid to properly evaluate some lump sum for GPT trades. However, removing them from the game is not a unique solution.

There are three types of lump sum for GPT trades right now:

- Resource (luxury/strategic)
- Open Borders
- Loan (its GPT for , or vice versa)

Nobody complains about Open Borders, as selling it to an AI invites a real risk of inviting a DoW that otherwise would not happen. The AI does a pretty poor job of putting a value on loans (it insists on making money on both ends, undervalues early game and doesn't recognize the risk of not getting paid back), but that's fixable if you tweak some scalars and make loans persistent even in a state of war. The latter may not be very realistic, but it's a necessary sop to the fact that AIs are often ill equipped to evaluate credit risk. I spent nearly five years with a major bank a number of years ago, and evaluating credit risk was part of my job. Trust me, I know.

The big problem is luxuries and strategics. Not because they don't confer something of value (more luxuries -> more and faster GAs, and more strategics -> more Deity unit spam), but because the trades can be terminated by the player free of consequence and warning.

One solution to that problem is to ban the trades. Another is to penalize the player rather than the AI for the deal's cessation. Either will work, but the latter is undeniably a more minimalist solution.

Actually, another method would be to make the AI unwilling to give up more than 30 lump sum gold for a 30 turn benefit (luxuries, open borders, etc.) and allow the AI to per turn deficit spend (if it has a treasury >0).


Peace treaties might be an exception since they are effectively indefinite. (they have their own penalty of being declared a warmonger)

The big issue is there is no reason for "lump sum" to be tradable at all.
The only "lump sums" tradable are gold, cities, and RAs.

 

[TheMeInTeam]

It's obvious with this AI. It's less obvious that the problem is intractable unless you've observed AIs dedicated to the problem in action. They function a lot like the computer on Jeopardy - they're very good for simple problems, but their errors are whoppers.

Ah, I gotcha. Well, one can figure there's a reason that AIs aren't used for these evaluations routinely on the professional scene...

There are certainly simple algorithms that we could use, but the problem is that all of them can be gamed under certain conditions. Leaving the deal in place is the most elegant solution, though the realism is dubious. After all, one principal cause of World War II was Germany's continued efforts to resettle the peace deal at Versailles. In CiV terms, they forked over several cities, all their cash and all their GPT (and then some) for peace, then reneged on the deal after fifteen turns. You can do that in CiV now; if you make deals sticky through war, you can't.

These are obviously tougher choices than they seem on the surface. Even so, I feel this would be a useful change to both players and AI functionality. There are other realism concession the game already makes; this one isn't MAJOR in terms of breaking reality.

I think, for MP however, leaving the game how it is for trades is a better idea. If HUMANS fail to evaluate credit risk, it's their own fault .

The big issue is there is no reason for "lump sum" to be tradable at all.
The only "lump sums" tradable are gold, cities, and RAs.

Actually, there are reasons to both trade away and trade for lump sums, depending on the situation. Right now trading lump sums away is a bit less popular for reasons Martin points out; the AI doesn't value them properly against PTO. However, there are still situations where the human has considerable incentive to do it.

 

[KrikkitTwo]

I think, for MP however, leaving the game how it is for trades is a better idea. If HUMANS fail to evaluate credit risk, it's their own fault .

Actually, there are reasons to both trade away and trade for lump sums, depending on the situation. Right now trading lump sums away is a bit less popular for reasons Martin points out; the AI doesn't value them properly against PTO. However, there are still situations where the human has considerable incentive to do it.

Well the only reason you would Give a lump sum (instead of per turn) is if you don't have enough per turn income.
[There a game change would be needed, allow deficit diplomacy, as long as your treasury is still >0.. if you trade yourself into default, its your problem]

The reason to Get a lump sum is for
1. buying units/buildings
2. upgrading
3. CS
4. RAs

all of those could be handled by... Plan ahead.

I agree it would be better to change the AI than changing the game engine (humans can evaluate their own risks). and just making the AI not do "lump sum" deals (except in peace) and letting the AI deficit spend if it still has gold in its treasury. (let the human player do it too... if they want to stop science for 30 turns and lose all their units go ahead)

 

[Randall Turner]

Geez. Is it only me that read this and thought, "boy, Ima gonna use that there".

 

[vexing]

Geez. Is it only me that read this and thought, "boy, Ima gonna use that there".

why not just load up firetuner and give yourself infinite gold?