Madfizzicks Madfizzicks

Major Diplomacy War/Peace Cheese at Intelligent levels

Major Diplomacy War/Peace Cheese at Intelligent levels

Ok, last night I'm playing a large map with all intelligent AIs, and I find myself militarily superior to the minor races and altarans (who I am slowly conquering), but second militarily (by a little) to the Drengin. The Drengin are, of course, at war with everyone (except me.. I went pure evil and traded with the lil' bastards to keep them friendly ;) )

So here's what I do: The Drengin have *tons* of stuff that I don't, but that I really want. So, I offer to attack the Alexians and Carinoids in return for a trade good plus cash. They go for it. Then, on the very same turn, I go to the Alexians and the Carinoids and extort a tech and cash for peace. I let a turn go by, then return to the Drengin. I offer to attack the Alexians and Carinoids for a trade good and cash, then go offer peace for cash and tech. Then I do it again. I even start getting the Altarans into the mix, since they have some techs that I don't. This continues for some time. I end up with 4 new trade goods, 10 new techs, and many thousands of bcs before the AI decides that it will no longer accept peace treaties. The Drengin never get tired of accepting the same attack offers for big payouts. On consecutive turns. At least once, I did the cycle without clicking off a turn. Still worked.

To the AIs credit, they do catch on eventually. The minor races wouldn't accept a peace offer for less than my entire empire. But it took five or six consecutive cycles. IMHO, that's way, waaaay too much.

Side note: Funniest GalCiv moment yet: The Altarans finally catch on and when I open the diplomacy window, I'm greeted with the message: "Let me guess. You're here to offer peace. Then you'll build up your forces along our borders, attack without warning, and offer peace again? Does that sound about right?" I just about died when I saw that! LOL

Another thing that seemed goofy was that in order to get the Gravity Accelerators from the Altarans, they wanted a planet in addition to the peace treaty. So, I went back to one of the minor systems I had just taken from them, removed the planetary defenses, parked a transport directly adjacent to the system, went back, offered them that planet, got the accelerators, re-invaded the planet, and put the garrison ships back. All in the same turn. It's a subtle AI point, but I wouldn't expect them to accept a planet from a race that is exterminating them if said race has an armada right on top of said planet. ;)

Anyway, I'm not saying you shouldn't be able to fight wars for profit, or extort big dough for peace. I just think that whatever AI safeguards there are against being repeatedly betrayed kick in way, way too late. And there doesn't seem to be any safeguard against continuously paying the human to go to war against the same races over and over again.
1,727 views 45 replies
Reply #26 Top
WOK, on the latest (sorta private) build:

Clearly Brad likes to let me go walk the plank and then call me back aboard, because this *appears* to have my #2 issue mentioned above (the more important one to me) fixed.

What I'm seeing is that certain attack offers don't even come up. As long as this is a relative weighting as to who can/cannot come up and not based on my political standing with the poor sodded victim (I reserve the right to turncoat on anyone strong enough to be a challenge in the game), then problem solved.

Is something like this what is different in the build?

Now, it's cold out here on the plank so high over the water... any chance of getting back into my cabin, sir?

~SDC~
Reply #27 Top
What does WOK mean you might ask? It means... -ahem- "Well, OK:" I'm just too tired to type I think.

~SDC~
Reply #29 Top
I believe so. Though it's some sort of candidate that may or may not have features to be release d along the official line. I don't know.

~SDC~
Reply #30 Top
Oh you can offer to attack enemies races. It just won't value them very much dependin gon the circumstnace. I put this in last week. But I can't update SDC until after 1.02 is put up. So that's where these private builds come in.

~SDC~
Reply #31 Top
Ah... now I'm understanding the system a bit more... Makes sense. Brad's got his little hoarde of goodies poised to pounce on us all here soon. GOOD!

~SDC~
Reply #32 Top
I have a bunch of other stuff ready to go as well, just not checked in.

I've started putting in code that basically goes like this:

if(AILevel > Bright)
{
Greatly increase odds of using tougher algorithms.
}

The only bad side is that it will slow down turns at higher levels. So you'll need to have a faster machine or be willing to wait longer.

~SDC~
Reply #33 Top
Brad: thanks for listening! I agree that this is a cheesy tactic, but it is the surface of a deeper problem. An AI that is too 'trusting'. It shouldn't be too dificult to fix this a little.

If the AI would also consider how the player treats other AI empires that would also have some great effects. This prevents players from using the same tactics against multiple AI players. For example: if i make peace to the torians and receive lots of money but attack again the next turn, the drengin would learn that my peace treaties are rather empty..... And so the player would be punished for his actions because the other AI empires would lose their trust.

Or... maybe i am dreaming to much.

Reply #34 Top
frogboy:

sweet! i guess i'll be holding off on starting a new game. awesome work brad. i can't wait to actually get my arse kicked by the AI.

dikdirk:

great idea! i think civ2 had something like this. a single, global "credibility" value that gets lowered when you abuse peace treaties would be cool.

cheers

h

Reply #35 Top
Frogboy,

I think the reason this severe bug has not been considered important by 99% of customer is simply
because they have not been AWARE of it.

There once was a diplomacy problem with Civilization 3
- witch emerged months after the game had been launched.
After beeing discovered, it suddenly got HUGE attention since it spoiled the game for those aware of it.
Of course the programmers at firaxis promised to fix it right away.
Therefore I cannot but wonder how you can consider this kind of gamebreaker a "low-priority" problem.

Reply #36 Top
I agree. If you break a peace treaty, it should hurt your peace reputation, and make other races unwilling to make peace with you. The more you do it, the more it should hurt your rep. Your rep should slowly return towards the mean.

If you tried to make peace, and the other party said something like "we remember how you broke your peace treaty with the Drengin, we don't trust you", that would be extra cool.

I think it's a big problem that this doesn't happen. It doesn't affect 99% of the gamers because this game has only been out for two weeks, people haven't discovered this yet.

In a realistic diplomacy model, you could get away with breaking peace treaties only once or twice and then the other races wouldn't put up with it anymore.
Reply #37 Top
I agree with Frogboy; mostly. It's not hard to *not* exploit the war/peace cycle once you've become aware of it (I certainly don't do it). But if the situation is subtle enough, like the situation where the AI declares war but will immediately accept peace, what should I do? "Fix" the situation by not asking for peace because the AI shouldn't accept peace, or "fix" the situation by asking for peace because the AI shouldn't have declared war in the first place?

And yeah, I'm using 1.02... but not the latest version of 1.02, since a few posts seem to indicate that new bugs have been introduced.

Anyway, this post is basically moot since I trust Frogboy's prioritization. As long as someone is aware that this stuff happens (which is my only goal when I post on a 'cheese' tactic), I don't really care if they put off fixing it until after they've fixed CTDs and other more severe bugs. I mean, at the end of the day, it's still a much better diplomacy model than anything else out there.
Reply #38 Top
Brad,

EXCELLENT! Waiting a little longer on turns is hardly a price to pay for the INT level AI to be made smarter. Sounds GREAT!

~SDC~
Reply #39 Top
I'm glad to see the cheese is somewhat being rectified.

However, I am bothered by the "Just don't ruin your game by abusing the system" argument. The same could be said about the deficiencies you are deciding to work on. For example, you are working on the strategic AI and its ability to deal with starbases, etc. I could argue that AI is good enough and only a small percentage of the players have noticed the AI's deficiency in this category. I could then say that you should just not take advantage of the AI in this regard. Don't build loads of starbases and don't go kill the enemies's starbases. But what fun is that?

Diplomacy should be no different. It is a tactical option to be used to defeat the AI without war. Deficiencies in the AI's ability to handle diplomacy should be just as important as deficiencies in the AI's ability to handle war tactics.

I hope this does not come across too negative though. I enjoy the game and the continued effort being put into it and I understand the list of issues must be prioritized. Just please remember diplomacy was a major selling point of this game.

~SDC~
Reply #40 Top
It'd be nice if casus belli were implemented. If I declare war on an empire because I ate a bad burrito the night before and woke up grumpy, sure I'll take whatever peace-bribe the saps offer. If I declare war on them because I've already lost a couple of worlds to their insidious marketing droids, it's going to take a lot more than a box of chocolates and some holovids to make me happy. The more CBs an empire has against another, the less likely they are to accept a peace offer.

Other CBs could include declaring war less than x turns after arranging peace, breaking an alliance or adding your personal email address to the Fundie mailing list.

~SDC~
Reply #41 Top
I agree that that the diplomacy screen and the Diplomacy skill were put in the game for you to use, and give you an advantage if you have a high Diplomacy skill. If because of bad AI you have TOO MUCH of an advantage, then this is not the fault of the player, it's a fault of the AI.

I like the comparison to the fighting part of the AI. If you build some military starbases in your home sector, the enemy civs keep throwing their ships at you and they don't have much of a chance because your ships have all these bonuses on account of the starbase(s). Is this cheating too?
Reply #42 Top
My main problem with the peace/war diplomacy is that the AI can "abuse" the system as well.
For example, playing on "normal" AI (i know, a lot of routines are disabled here), I had the two most powerful races attacking me during the early-mid game. I finally pulled even militarily, and once they had some losses, both were ready for peace. Since one was good and one was evil, I decided to play them against each other while getting myself out of the way. I went to the good AI, and got peace and a monetary tribute. I then went to the evil AI, and got peace + an agreement to war against the good AI in exchange for paying the same amount of tribute. However, after only 2 turns of war, they declared peace. The tribute was for 10 turns; I paid for 8 turns for a war that wasn't going on!
It's one thing for me to decide not to "abuse" the peace/war diplomacy. But the AI does it on it's own, and I would like it looked into eventually. Sure, this isn't the highest priority; but it is more than a "cheese tactic".
Personally, I'd prefer there to be an enforced minimum time on any diplomatically acheived peace or war; if there is money changing hands, then the length of the payments would be a nice value to use; the AI would have to be tweaked to know about that, of course.
Reply #43 Top
Has anyone ever programmed an AI? It isn't as simple as lets just add a length value to war declarations and let the AI know about it. Entire new algorithms and methods need to be written for every level of AI intelligence plus how much it would value this new aspect in diplomacy settings. At some time the AI will hit a point of diminishing returns. I am betting that we are already there. Brad could keep on adding more and more stuff to the AI, but the computing power needed to run it increases more than the benefit of the added feature. The minimum requirements are a P3 600mhz. Considering I am running a AMD XP 1900+ and from time to time turns slow down. We can't just ask brad to implement wonderful new features because I can run them. Think from the perspective of this game running on min requirements and try and figure out how much more can be added before the turns get too long and it starts to be considered unplayable. This wasn't meant to attack or offend anyone, but at some point we have to realize that the computer doesn't run as fast or as well as our creativity or imagination. After all it is only an AI and no matter how good it gets we will always find holes and cheese tactics.
Reply #44 Top
But Nibor, I LIKE it when the AI tries to screw me over once in awhile... I've had that happen too, LMAO on that one and lesson learned. I can adapt much easier than the AI. :) besides, I've had them fight longer than the terms of some of my tributes to them. So it washes out.

~SDC~
Reply #45 Top
Not that I have an opinion on these particular tactics, but generally speaking, I'm not sure you should dismiss cheese tactics "preventable".

If the diplomacy system has obvious flaws like these (assuming they exist up the AI line), it undermines the use of the entire system. In effect, if you use diplomacy at all, you're running the risk that you're getting a cheesy victory, not a well-earned one. I really like the idea of being able to do an all-out assault with all the tools at my disposal. If a particular tool is beyond the ken of the AI, then you really should just yank it from the game. It's just no fun to have to second-guess all your actions as to whether they're fair.

~SDC~