Permanent Alliance Guide

Maybe not correct place for asking that.. but if someone can help - is it possible somehow to enable PA for "Play now" game, I mean - or my current game, where I'm already around techs that allow PA, managed to keep 2 great friends.... but than I realise that I play standard settings and there is no PA possible... what a bad luck in my best Emperor level game... :(

And another question - for PA Domination victory - do we need more % together for win?
 
Play Now has all default options so no... try Custom Games :)

You can save the game as a worldbuilder scenario and then edit it, but that resets food, hammers, gpp, beakers ...

Domination % required is affected by the number of teams I believe, but it's very small, so it won't make much difference.
 
It seems to be working like peace vassals. An AI will refuse to become a peace vassal if it puts you through the domination limit. Weird, but logical in a Firaxis way of implementing things.

Yes, logical in the "idiot logic" Firaxian sense. "No, I'd rather not win instantly. I'd rather play in an annoying fashion that instead virtually guarantees my loss. Derp."

A lot of their choices and their choices of priority are so genuinely terrible it amazes me...and it's so consistent! Shoddy controls, completely terribly streamlined AI turns/cpu usage/etc, utterly broken victory conditions, known bugs, ludicrously terrible units with almost no uses, AIs that play to lose on purpose (but only at random), events that can change a game on a dime independent of strategy, completely irrational diplomacy from both a sandbox and a competitive standpoint, etc.

Those issues known, the most important thing is to patch barb galley spawn rate and mess with espionage a bit...

...

:sad:. What went through the minds of the people who made that choice? They're professionals on an AAA title. I'd love to hear their explanation. Maybe I'm wrong and they'd roast me in a debate...but I don't think so :).
 
Yeah it's rally weird that you give the AI a rare chance to acutally win and then it says: No, I'd rather srew you over and have both of us loose. It's the kinda AI that would have made the Cold War hot. It's not how diplomacy should be. How ironic that I would have won by 1800-1850 AD if I had not killed the Native Americans which was a though piece of work. My landmass would have been lower, I could have whipped away my people, Charly would have PAed and settled the Ice belt in the south for our domination. Now it would take me at least until 1950 to rebuild and crush 1 1/2 more civs with Infantery.
 
It seems to be working like peace vassals. An AI will refuse to become a peace vassal if it puts you through the domination limit. Weird, but logical in a Firaxis way of implementing things.
Actually no, but it has another WTF "logic" within:
Code:
DenialTypes CvTeamAI::AI_permanentAllianceTrade(TeamTypes eTeam) const
{
    PROFILE_FUNC();

    AttitudeTypes eAttitude;
    int iI;

    FAssertMsg(eTeam != getID(), "shouldn't call this function on ourselves");

    if (isHuman())
    {
        return NO_DENIAL;
    }

    if (AI_getWorstEnemy() == eTeam)
    {
        return DENIAL_WORST_ENEMY;
    }

    if ((getPower(true) + GET_TEAM(eTeam).getPower(true)) > (GC.getGameINLINE().countTotalCivPower() / 2))
    {
        if (getPower(true) > GET_TEAM(eTeam).getPower(true))
        {
            return DENIAL_POWER_US;
        }
        else
        {
            return DENIAL_POWER_YOU;
        }
    }

    if ((AI_getDefensivePactCounter(eTeam) + AI_getShareWarCounter(eTeam)) < 40)
    {
        return DENIAL_NOT_ALLIED;
    }

    eAttitude = AI_getAttitude(eTeam);

    for (iI = 0; iI < MAX_PLAYERS; iI++)
    {
        if (GET_PLAYER((PlayerTypes)iI).isAlive())
        {
            if (GET_PLAYER((PlayerTypes)iI).getTeam() == getID())
            {
                if (eAttitude <= GC.getLeaderHeadInfo(GET_PLAYER((PlayerTypes)iI).getPersonalityType()).getPermanentAllianceRefuseAttitudeThreshold())
                {
                    return DENIAL_ATTITUDE;
                }
            }
        }
    }

    return NO_DENIAL;
}
Ok, in resume:

- The AI will not PA with it's worst enemy. Logic enough ...

- The AI will not PA with someone that has not atleast get 40 total turns of shared war or defensive pact ( or any combination of those two that gives 40 turns ... like 10 turns of DP and 30 of shared war ). The fact that is speed independent is a WTF in it self, but it is common enough in Civ IV ( I guess they skipped the coding class where their teachers teached that magic numbers are bad :D ), but atleast it reflects a game need: it assures that you will need to show will to protect the other civ and/or to fight alongside it.

- The AI will not PA someone that they diplorate below or equal to their PermanentAllianceRefuseAttitudeThreshold. Fair enough for "ropeplay" issues

- Now the REAL WTF : the AI will refuse to make a PA if their power + your power is bigger than 50% of the sum of the alive civs power, giving DENIAL_POWER_US if they are more powerful than you and DENIAL_POWER_YOU if it is the other way around. Besides this being quite iffy if vassals are around, I can't really understand what the coders wanted here , atleast in terms of a good playing AI: basically it disallows the AI of making a PA with more than 50% of the total world power. I can see why they would like to put something like this for roleplay issues ( it basically forces a far more equal bunch of PA in game , by making hard that the top players make PA with each other ), but even then it is a very crude mechanism ...

P.S It is even worse, because in some cases it might prevent #2 and #3 in power to PA against #1 ... say with 4 alive civs with #1 having 35 % of the world power, #2 with 31%, #3 with 29% and #4 with 5%. #1, #2 and #3 can't do any PA with eachother, meaning that #2 and #3 can't gang up against #1 :/
 
@rolo
don't touch my magic numbers!

I follow very easy rule regarding magic numbers... if I write the number 1st time and I suppose it will be only 1 time use, it's fine... then I come to the 3rd usage and start to think that using magic number wasn't probably that hot.

On 5th usage I start to think that it really should be changed to some form of constant, but who in the hell would search whole project for all ocurences? thus skip it!

On 20th usage you usually lose your mind and try to fix that mess... all say hello to bugs and Q&A department!
 
Firaxis uses magic numbers in another and more creative way : it hardcodes a number that, due to program needs, it has to work within varying enviroments ... they do that a lot with turn numbers. 10 turns for a forced peace deal, no matter what speed, is probably the more known example ...

In other words, using a magic number where a variable ( not even a constant ) is needed :/ I'm sure my 1st semester programming teacher would :cringe: if he saw this :D
 
Yeah it's rally weird that you give the AI a rare chance to acutally win and then it says: No, I'd rather srew you over and have both of us loose.
I think the basic goal of the AI programming is to make you lose, rather than make the AI win. The AI winning is only important if it also makes you lose.
In other words, using a magic number where a variable ( not even a constant ) is needed :/ I'm sure my 1st semester programming teacher would :cringe: if he saw this :D
I was a 1st semester programming teacher for a while, and I randomly swapped reactions among :cringe: :mad: :rolleyes: :eek: :cry: :nono: :thumbsdown: :wallbash: :shake: :aargh: :old: :sniper:
 
I made a quest to unite France :D I chose DeGaulle (I looove charismatic and prefer industrous to Nap's organised) and preselected 2 other french, 3 english and 2 random civs. (monarch, big and small with massive continents)

Unfortunately I started around 10 tiles away from Napoleon (the most uneasy and agressive leader of all preselected), on heavy flooded terrain with no hills and some forest, imposible to chop due to health problems (and I reserched therefore pottery first) - there was no possibility to rush Napi therefore, so the only option was to befriend him. I took huge afford to convert him to confucianism, as most of hist cities were early hindu, finally I got it and we formed a "only-2-confu" block.

I declared on Louis twice, taking all his land, and to unite France I need PA with Napoleon. Now it is ca. 1650-1700 and I can easily get Space Victory in ca. 40 turns of finishing spaceship + 30 turns of flight (marathon speed). I can also easily crash him with my mech infs against rifles or grenadies, but I am to lazy to take his 20 cities in war. I want PA !!! Each of us has a vassal (Nap - our common hatred Elisabeth, me - Willem after joining war at Nap's request)

During history we shared some wars (the only two we didn't was my 2 wars against Louis), so I have PA option allowed. But...

First
Napoleon didnt like me enough at +12 relations (-2 to borders, -2 to declared war on Louis, and rival vassaled -1).
So I switched back from free religion to confu, espionage mission to Napoleon to confu, and took representation as well - a huuuuge production lost to my 1/3 cottaged cities, but representations speeds up my research a lot due to tons of corn and Cereal mils and I work a lot of engineers)
And then...
Second:
He likes me enough but He is doing well on his own!

So please make me sure:
1) These is because he has stronger army ratio? (Actually my is weeakest, ratio 0,7). If so, just a few built tanks could resolve it?
or
2) If I do 1), he wouldnt form a PA because I will be to strong for him?

Is the only option build a mass of units and gift them to Churchill? (Napoleon is preparing now for some molitary campaing, for sure Chirchill or his vassal Victoria, in this case some mech inf gidted to churchill will bring napoleon into troubles and by this making him more willing to PA?)

How is tech advance calculated into possibility of PA? (I have never had such a huge tech lead before compare to other nations - imagine 4th civ in rank, chinese have not even gunpowder yet!, churchill's best units are cavalries and rifles, napoleons - grenadiers)


PS.
I know taking 3-5 his cities will cause him to capitulate, as his core land is close to my core land, but vassal is not "united France" xD)
 
I had always misunderstood the 40 turns of war qualification for the PA. I thought it had to be an ongoing war for 40 turns, continuing when the PA was offered.

So, I was quite surprised in a game where I shared a war with Joao (for exactly 40 turns) and he offered a permanant alliance 29 turns after the war was over once he got communism. He still had the vassal from that war too!

It sounds like others are saying that if you have a Defensive Pact for 35 turns and share a war for five turns, this can get you a PA also. Is this correct?
 
Yup, it is correct. It is 40 turns of total duration of shared wars and DP. It does not need to be continuous and can be any combination between shared wars and DP that lasts atleast 40 turns in total
 

Thank you for clearing that up. Knowing how it works now, it really doesn't make for a good representation of reality in the game but at least I know with whom PA and why it doesn't work.

Does the rule about the AI being #1 in power still stand? Because last game I played, I had 0,7 my PA partners strength after signing the deal. I'm not sure if it factored in my power as partly being his power or something.
 
The thing I like best about PAs is that they decrease the luck factor, much like disabling tribal villages and random events. If I start isolated by some stroke of bad luck, I can cozy up to the most powerful/advanced civ on the map later and sign a PA as soon as possible. This makes it a lot easier to win a game that would have been lost due to bad luck.
 
The thing I like best about PAs is that they decrease the luck factor, much like disabling tribal villages and random events. If I start isolated by some stroke of bad luck, I can cozy up to the most powerful/advanced civ on the map later and sign a PA as soon as possible. This makes it a lot easier to win a game that would have been lost due to bad luck.

Unfortunately, that doesn't spill over into competitive settings. They're a farce in MP and in formats like hall of fame relying on them often makes the game take too long to be competitive (though it IS how I won g major 96, that was a rather exceptional case).
 
I'm just playing with PAs for the first time, and wanted to warn you of something:

You know the AI behaviour to their Vassals? That they gift them all techs one after another without getting anything in return?

This is exactly the same if you have a PA and a vassal, the PA-partner gets all your techs automatically, and then gifts them to the vassal, so no chance to keep the vassal from building certain wonders or even get his techs through trade once he has researched them.

I think, playing with PAs is best when turning of vassal-states.
 
I'm just playing with PAs for the first time, and wanted to warn you of something:

You know the AI behaviour to their Vassals? That they gift them all techs one after another without getting anything in return?

This is exactly the same if you have a PA and a vassal, the PA-partner gets all your techs automatically, and then gifts them to the vassal, so no chance to keep the vassal from building certain wonders or even get his techs through trade once he has researched them.

I think, playing with PAs is best when turning of vassal-states.

Yes, yes, and yes. I got the same thing yesterday when I finally got PA. I just thought vassal will speed up my domination plan with OCC. At first, I think it's ok, maybe a new way to use vassal (AI way).

Then suddenly learn this hard way.

A vassal broke free after got ALL my tech. This should not be a problem if I got my normal empire with 3 monster production city. But with OCC, well, and I got so used to fight cannon against longbow, so..... LOL, Civ never cease to impress me.

My thought, crush everyone except your PA. :D

I realize that best way to got PA is share war. I never sign DP cause it took a detour from my lib-communism. Thanks to share war, religion, civis, I got +22 with my choice #1 and +17 with choice #2. And with low peace weight AI, it is easier cause they have better war bonus and cheaper to bride them to war. I paid around 400 gold for them to start a war (which is super cheap consider all the money you get when play OCC).

P.S: Seraiel, I thought you are in hiatus? I seriously enjoy your write up, all of them, I read every post. U know that you and Wastintime is my hero doing sth I would never image, manage 200++ city?? LOL
BTW, from your game that I learn the power of Orcale Currency, which is so OP in normal game, but totally pointless in OCC. Man, it took my like half hour to decide which I should choose for my Orcale.
 
Hi ScorpionK :)

It's right that I was on hiatus, but atm. I'm learning, that I'm not quite rdy to live completely without games, as that maybe would mean, I'd have to become a polititian? *lol*

I'm flattered by your compliments, find it funny that you experienced excact the same picture in your game, and welcome to these forums, where people actually really help each other, at least sometimes :) .

Wonder on which difficulty you play on, as I heard, Deity OOC can be quite a challenge :rolleyes: :D .

Seraiel
 
Top Bottom