WHEOOHRN Hiding

[DaveMcW]

In CvTeamAI.cpp, move this block

if (getAnyWarPlanCount(true) > 0)
{
	return DENIAL_TOO_MANY_WARS;
}

... to the bottom of the function. After DENIAL_POWER_THEM, DENIAL_NO_GAIN, DENIAL_ATTITUDE, DENIAL_ATTITUDE_THEM, and DENIAL_JOKING.

This will make it harder for humans to see a sneak attack coming.

 

[MadmanAtW]

Good idea.

 

[r_rolo1]

Not sure if everyone will like that ...

 

[MartinHarper]

I think the important thing is that DENIAL_ATTITUDE goes first. If a civ hates me, that's all the information on their war plans that I should get.

 

[phungus420]

This is a gameplay change. It's effects would be huge. It's far too large of a gameplay change to fit BBAI in my humble opinion. I fully support this change going into RevDCM though.

 

[Yakk]

No, it is an AI change. The AI simply tells you something different when you talk to it: it behaves EXACTLY the same, it just gives you less "sensitive" reasons first.

 

[phungus420]

Which is a huge gameplay change. Read the Succession games Yakk, WHEOOHRN is a huge aspect of gameplay; hiding this from the player, except when all other conditions are met will create big changes to how the game is played. No way around it, if you disagree, again check out the succession games forum.

 

[Yakk]

Yes, it will change how players play, because the AI is no longer saying "I'm going to attack you" (or something that can be interpreted as same), when they can give some other excuse.

Heck, the rule that "never, ever communicate that you are in the final stages of planning to go to war with someone to you are planning to go to war with" would be an AI improvement.

Improved AI should result in the human having to treat AIs differently. When AI learns how to smash invading stacks, doesn't fall for siege city-grinding, builds effective SODs, figures out the industrial revolution, picks technologies in order to trade with fellow AIs more effectively, values technology less when it is widely known, etc -- all of these things will change human gameplay, because the AI is no longer being a fool.

In this case, a player can determine immediately after an AI makes a threat with a high success rate if the threat resulted in the AI starting war plans against the player, by comparing the before and after state of the AI's response to that query. And that is a flaw, both smart-play wise, and roleplaying wise, in the AI.

 

[MaxCiv]

I agree with Yakk 110% . Using that trick to know if the AI is getting ready to attack you is very close to cheating IMHO.

 

[Lord Tirian]

...except when all other conditions are met will create big changes to how the game is played.

Better AI changes the game anyway, as you need to adjust to the new AI. Isn't that a gameplay change already? Or the tech diffusion effect jdog5000 implemented recently? BetterAI changes the gameplay already. Besides, it isn't a rule change... it's not changing in the game rules or giving the AI an advantage or anything, just makes the AI not announce their internal workings to the rest of the world.

Cheers, LT.

 

[Munch]

Could someone summarise the various DENIAL_* in terms of how they appear in game? I presume that DENIAL_ATTITUDE is 'we just don't like you enough', and occurs when the AI dislikes you sufficiently, but what are the others?

 

[phungus420]

Hmm, I see your point, and you have convinced me. However I'd still like to see what some other player, specifically those who write strategy articles and play succession games and use Better AI (such as TheMeInTeam) think.

 

[r_rolo1]

Ok, let me give my opinion as a SG player, writer of strategy article(s) ( I only wrote one so far, but have another in the forge ) and a user of BBAI

Well, first of all, the order of the denials as it comes from the box is pretty messed up .... if you assume that you are actually talking with the AI, that is. As I pointed out here, I am not so sure if the denial text were intended to be atributted to the LH speech , and even as it is now, it is a quite a bit of suspension of disbelief to think that you are actually talking with the enemy leader in the diplo screen and not via one or two proxies. The point is that the order of the denials at it is now makes sense if you assume that you are actually talking with one of your ministers and not with the leader and I wonder if that was not the original philosophy behind it ( that is, to put the text that we see in the hovers coming out of a third party, similar to what happened with Civ III diplo advisor in the diplo screen ) ...

Well, gamewise, not having the "hands full" alert so often would make things harder, that is a certain I can definitely say that I rely on it extensively and would probably miss it ... in fact that alone IMHO would be enough for a lot of people to be forced to play a level below of what they do now ( OFC there are ways of recognizing a war prep out of the "hands full", but they are far less reliable ). But it would probably be bearable, barring those cases of a AI sticking in hands full for ages because it has no means of attacking the selected target ( say, closed borders in between or lack of ocean-going vessels .... a situation that I'm quite used to in the LHC games ) ... but in the end those situations shouldn't happen in the first place if the code was minimally right ( hint to jdog )

But, regardless of what my sentiments are regarding this change ( I'm quite ambivalent regarding this, due to the fact that I'm not sure of what was the original coder idea ), I don't think this is a issue to be tackled by this mod. Why? Simple: changing it does not make the AI to play better ... at best it makes the human to play worse, and that is not the philosophy behind this mod ( atleast it is how I see it: the BBAI is a project to make the AI to play better this game, right? ). Just because of that, I say that this change should not be included in this mod.

 

[scu98rkr]

"changing it does not make the AI to play better"

Yes it does it can more easily plan wars against the human/other AI without telling the human player of its intention to attack.

Your statement really depends on what you mean by make the AI play better. I mean as it currently no matter what clever programming the better AI team does the AI wont be able to plan wars without a careful human noticing.

What could they do to make the AI better in this respect ? Randomly say WHEOOHRN when they dont mean it to confuse the human ?

Really it would be better if the AI just gave no info out.

 

[r_rolo1]

Ok, let me rephrase it: the proposed change will make 0 for making the AI moves to be more well planned ( like not go in farm->cottage->farm cicles or to choose better the targets in war do )... it would only make the other players ( remember, just because the AI currently is not coded to take ilations of the denials does not mean that it can't do that with a diferent code ) have less acess to their plans. Said in other words, the only effect that the proposed change has is to make the life of a player ( human or not ) that can read between the lines harder... and that is not make the AI to play better, is to make the player that uses this feature ( human or not ) to play worse. As I understand it, the objective of this mod is not to make the game more chalenging ( we have a handicap file just for that in the stock game ), is to make a AI that plays this game better. So, IMHO, in spite of agreeing or not with the change, I think this mod is not the place to implement it .

Now on my opinion on what could be done.... if you read the thread I linked on post #3 ( where this thread OP already had suggested this change BTW ), you would see that, if I had made the code of the game, I would not had made the "hands full" denial as a separate entity, but would had mixed it with other denial and call it "We are not interested in your proposal" ( probably joining it with either ( or both ) DENIAL_JOKING or DENIAL_NO_GAIN ) ... it would be vague enough to not be easy to pinpoint what was really on their "mind", but would give space to know if the AI was preparing a war combined with other intel ( say, espionage given city screens ). But again, IMHO, the better AI mod forum is not the place for a change like this to be discussed or implemented

On the AI bluffing ... well, it would be refreshing to see a AI to pull a decent bluff... but I'm not seeing the AI doing that without some serious developments ( like a robust threat level function )

 

[Fimbulvetr]

Actually you are wrong,, BBAI definitely aims at making the game more challanging for the human player, without having to use ******ed handicaps.
AIs giving out info about their war plans is both unfair and stupid - or does the human tell the AI when he's going to war?
If you want to know an AI's war plans, use the cheatcode in the ini, and ingame press shift+alt and hover over that leader name. But know that when you do that, you are cheating.

Of course, the AI is cheating too here and there (like knowing where the next good city location is without scouting that area first) but I really think giving the human the ability to cheat against the AI like this was definitely not intentional.

It's probably a matter of opinion. One thing I think is a fact though: As a part of the AI, this can be changed for this mod.
Whether or not it 'changes gameplay': If it changes your own gameplay, that of course depends on if or if not you have used this method to predict AI wars against yourself. I didn't know about it, I didn't use it, don't plan to, and therefore get surprised by AIs attacks sometimes. As I should, in my not so humble opinion.
So yeah, I guess I am biased towards implementing this change. Just as everyone who has used that information leak successfully would probably be biased against.
Anyway, I don't see how this changes any game rules, or 'gameplay' per se. It changes how a human has to play because the threat from the AIs become less predictable? How is that a bad thing in the BetterAI mod?

 

[r_rolo1]

Actually you are wrong,, BBAI definitely aims at making the game more challanging for the human player, without having to use ******ed handicaps.
AIs giving out info about their war plans is both unfair and stupid - or does the human tell the AI when he's going to war?
If you want to know an AI's war plans, use the cheatcode in the ini, and ingame press shift+alt and hover over that leader name. But know that when you do that, you are cheating.

If you have that position, you have to be ready to take it all the way... for a example, why not simply remove the hover texts in all the denials? There is any reason to have the AI to give ANY intel at all about their reasons for any of the denials in the diplo screen? Would that not make "the game more challanging for the human player, without having to use ******ed handicaps" as well ? Isn't that " unfair and stupid - or does the human tell the AI" the reason for any of their denials ? If you think so, well and dandy, but the simple fact that the coders introduced them indicates that they would not agree with you . If you don't think that way, it would be a nice idea to tell where you draw the line and ( more importantly ) why the "hands full" hover is diferent of the other hovers you consider that deserve to be maintained. But anyway, that is not the position of the OP and it was not about that that I was talking about...

P.S To make it more clear, there is a infinite number of ways of making the game more chalenging to the human player that don't pass for making the AI players to play smarter or by the human handicaps. Stock BtS already had some introduced in the various layers of the code ( intended or not ) since it got out , like the diplo win rule change, the change on the hammer overflow rules or the change of the culture needed to get a legendary city in No Espionage. I strongly doubt that you would consider this to be improvements to the AI just because it makes harder for the human to win ( and I also doubt that you would suggest any of the above quoted as changes for a Better AI before the time they were implemented ). IMHO the change the OP suggests files in the same cathegory ( already explained why above ) ... if you don't think so , please state your reasons.

 

[Lord Tirian]

Something that BetterAI should ultimately strive for is, I think, to be closer to a multiplayer experience, so to speak... a bit as if a human was playing the other leader (human with certain personality, that is). WHEOOHRN fails the "act human"-idea hard and is something that's specific to the AI only - you can't apply it to human players. Hence, I dislike WHEOOHRN.

On the hiding diplo-hovers in general... not so much, a "human" would give you a response as well (saying nothing is equally bad), he just wouldn't give you such a telling and revealing response.

Cheers, LT.

 

[r_rolo1]

Well, you are assuming that MP players act inteligently , I assume

On acting as a human.... regardless of disagreeing with you or not, I must point the fact that there is a LOT of stuff that humans can do and AI can't and vice versa . Humans can't be vassals, AI can't bribe AI into war, you know, stuff like that ... My point has been always that I would love to see a Civ IV AI passing the equivalent of a Civ IV Turing test ( as a quick search on my alias + "turing" will show ), but the fact that ( for a example ) humans can/can't vassalize has 0 to do with AI smartness: a AI will not get magically a better player because it can/can't vassal a human ( or by being able/unable to bribe other AI to war or any other thing in that cathegory of "things the AI can/can't do and that the human can't/can" ). IMHO the denials, "hands full" included, fall in there as well: they could definitely be worked in the way of looking more human ( as I also sugest in post #15 of this thread for the "hands full" denial ), but that will not add by it self a picogram of smartness to the AI player. If it doesn't make the AI smarter, why discuss it in here? That has been my point in here ... it doesn't mean I like the things as they are now ( I don't ) , but just that I think that it is not a thing to be discussed in this forum as a feature for this mod, unless it comes bundled with some feature that actually makes the AI play smarter.

P.S I just wonder how diferent would be the positions on this if BBAI already had AI capable of taking ilations of the denials of the other AI and act accordingly with a minimum of efficiency

 

[BobTheTerrible]

I agree it should be moved. This seems like a general bugfix, something for the UP - which would carry over to BBAI anyway