The problem with AI sneak attacks is that this is the only way the AI declares war. They never outright declare.
I'm not talking about refused threats, or having spy activity caught. These are provocations. The AI never declares war without such a provocation except by sneak attack.
This whole aspect of the game is broken. IMHO.
With a broken game element, there is no happy solution. Since the AI is specifically programmed to keep its war plans secret until it has at least one unit in attack position, I consider it fair game. The AI will HAPPILY cheat the Hell out of you if you happen to sign any goods-for-payments deals with it. I got burned by enough of those, in the early days, that I will never be satisfied that I've obtained enough revenge until such time as this broken game element is fixed. While it remains broken, the AI is designed to lie, cheat and steal, in regard to signing deals while it has ALREADY GONE TO WAR WITH YOU, you just don't know it yet because the sneak attack is in progress.
Can this be exploited? Yes, it can. Is this exploitation good for the game? I don't know. I do consider it "fair" to the AI.
Making a deal with a dying civ is similar, and both come with risks.
Look what happened with my German deal. Germany did not die out quickly, as I expected. With just two nothing towns left on an island, I figured the Mongols would unload some ships up there and it would be over quickly. That didn't happen.
We STILL got a good deal on Nationalism. The price from any other trading partner was more than 270gpt. I paid 127gpt plus Industrialization. I made sure not to trade Germany any lux or goods as part of the deal.
There's an equal risk with the "Cheat the Cheater" move of buying tech for gpt from invaders. What if they don't attack? What if some other civ pulls their puppet strings? I have had it happen to me! I've paid big gpt for techs and had to eat the whole deal, because I misread the incoming attack. That doesn't happen often, but it can happen, especially on Pangaea and larger maps, with lots of AI's running hither and thither.
Bottom line for me remains the game balance. Should we let the AI have its sneak attacks with no means to fight back? I don't think so. I LIKE that it sneak attacks, but I DO NOT like that it does nothing else. That's pathetic. If it were less predictable, it would be less exploitable. In the mean time, I can and will play without swindling the cowardly swindler as a tactical option, in any games where the host wants to take it off the table. If it's not off the table, I have no qualms about using it because I don't consider it wrong. The AI is deliberately set up to function in such a slimy way. I say let it reap what it sows.
In my view of game balance, I don't see it as appropriate to remove an option from the player that the AI is programmed to use. That's why I consider it fair to be allowed to scroll ahead to change build orders. (Ever seen four wonders complete ON THE SAME TURN? I have. That's the AI scrolling ahead to swap its wonder to a different one after the one it was building was finished by another civ). This is also why I consider RoP abuse UNfair: the AI doesn't do any mass RoP abuse. It only does single-unit RoP abuse. So citing its single-unit RoP abuse as grounds to line up your entire army in attack position outside its cities is unbalanced and unreasonable.
Well, the AI will sign YOU to deals while it already knows it's coming to attack you. Something here ought to be fixed. Either the AI should refuse your deals (giving away that it is on a sneak attack, but that's another story), or it should RECONSIDER its attack plan after signing such a deal. If it were changes to be less predictable, less single-mindedly stupid, the game would be more fun.
Right now, it's not really fun, or fair, either way. If you don't set up the lying, backstabbing sneak attacker to pay a price for their sliminess, then you let them get away with betraying you. It becomes a cheat or be cheated scenario, with no fair outcome, only someone on top and someone on the bottom, and in that case, I consider fighting back diplomatically to be self-defense. The player should NOT be required to bend over and take it gladly. That is not what "honor" is really about, you know.
It's broken... and I hope someday it will get fixed. Either way they fix it is OK with me. Until then... you know where I stand.
As for maps... the AI's trade maps ALL THE TIME. They trade pennies back and forth and keep each other informed. If you get in on the action by trading maps yourself, you pay for your own map info by reselling it to others. The whole thing brings in pennies, only pennies.
The only factor there that CAN be exploited is that there is a minimum 1g cost for any map info at all. So you could explore fog with one movement, sell that one new tile of info (which is rightfully only worth a fraction of one gold) to all the AI's for 1g, move your unit again and bust another tile of fog, sell for 1g to all, repeat and repeat, whereas if you just moved the whole movement at once, and then sold, you'd only get 1g from each total. THAT can be exploitive in situations where every penny counts. You turn .05g map info into 1g cash, over and over, you pull something unfair out of the system, something for nothing. HOWEVER, if you selling a bona fide 1g worth of map info for 1g, I don't see a problem.
Short of that, you're worrying over nothing. The AI's trade their maps every single turn in later turns when they have mature borders and lots of units roaming. (Each time their units pass by each other, is a chance for them to open negotionations, so the more units near other AI units, the more often they offer each other deals).
Map trades are important FOR THE MAP INFORMATION. I guess I'm in the minority in considering current intelligence on the map important. But hey, what do I know?
- Sirian