Deity Toku

[Dirk1302]

@Noogai,
No it doesn't work that way. Ais know all power ratings just for this reason.

 

[Noogai]

500 AD

Spoiler :

I built elepult army and attacked Monte. I got his capital, but then he vassaled to Zara!
I don't have enough anything to bribe anyone against anyone and everyone is in peace with everyone right now, so...
At least I got the Oracle now

Spoiler :

I guess my only change is to attack alien face since he is the hated jew, and hope that someone will join me instead of backstabbing. Or is there any better ideas?
My stack and land:

Spoiler :

Tech situation is just sad..Can't do much with 3 cities in deity.

Spoiler :

 

[Dirk1302]

^

Spoiler :

I guess Zara was at war with Monty too at the time or Zara would be at war with you now. I don't know the code for vassals but i have a feeling this vassalization becomes an issue much sooner if someone else is at war with your target as well. Also taking their capital speeds up vassalization i believe. Zara is ridiculously small to have a vassal in the first place if i look at the map, isn't Monty still bigger ? Well there was no way to bribe Zara to peace with Monty given the tech situation.

Situation looks bleak, you have a decent stack and i think it must be used right now. Attack Sury but don't expect too much of it, he'll be strong for sure. My first thought was maybe ride it out with Monty/Zara. Zara is advanced but he looks really small and he doesn't border you so you don't have to worry about your own cities immed. Then again you'll lose Monty's capital and basically you'll start over again with a good chance of being backstabbed in the process. Why didn't you take on Sury in the first place?

 

[Noogai]

^

Spoiler :

I guess Zara was at war with Monty too at the time or Zara would be at war with you now. I don't know the code for vassals but i have a feeling this vassalization becomes an issue much sooner if someone else is at war with your target as well. Also taking their capital speeds up vassalization i believe. Zara is ridiculously small to have a vassal in the first place if i look at the map, isn't Monty still bigger ? Well there was no way to bribe Zara to peace with Monty given the tech situation.

Situation looks bleak, you have a decent stack and i think it must be used right now. Attack Sury but don't expect too much of it, he'll be strong for sure. My first thought was maybe ride it out with Monty/Zara. Zara is advanced but he looks really small and he doesn't border you so you don't have to worry about your own cities immed. Then again you'll lose Monty's capital and basically you'll start over again with a good chance of being backstabbed in the process. Why didn't you take on Sury in the first place?

Spoiler :

Yes Zara was in war with Monty, and he is small..

Actually I thought about wether I should attack Monty or Sury (I don't think I have a save left before attacking Monty, unless there is an autosave). I chose Monty because he was weaker and closer. I was also afraid of having cities in the middle of the whole continent. Mongolians, ethiopians and zulus were in war with Sury at the time I attacked Monty, so I thought that Sury would be crushed soon and I would get Monty (Only Zara was in war with him). And btw Zara wanted the gold I stole from aztecs, so I had to give it..

I guess I will attack Sury with my stack and try getting some techs along with couple good cities. Then if I get lucky, AIs will declare on each other and let me vassal someone weak(end) before its too late.

 

[AbsoluteZero]

Spoiler :

Can you post a save just before Monty goes into wheoohrn.

I didn't keep any of those saves and I'm playing a game now where he didn't backstab.

Distance is not a factor in most declaration issues.

BS. I don't know where you got this info but logistics is the #1 factor in my experience. The AI does not march out of it's way to declare except in rare instances. If he's cautious with you and annoyed with someone across the map then he's going for you 90% of the time.

It's interesting to note that both Sury and Zara would not declare on Monte because "they had nothing to gain". Probably meaning Monte was not close enough.

In all those games where I was backstabbed either Zara or Justin was Monte's worst enemy.

 

[amanbe]

I believe power rating is more important than distance, then again I'm not a code-diver. It seems to me military techs, barracks in all cities, etc. make a big difference....

Spoiler :

Early archery is your friend on this map...

 

[Mylene]

I agree with @AZ,
Distance, strength of your army (if you have one ), your road network/distance between cities aka how fast they think you can put up a defense.
These are the deciding factors, unless i completely missed out on something during the years

 

[kossin]

Distance from target is only considered after all valid targets have been chosen.

http://forums.civfanatics.com/showthread.php?p=7190899#post7190899

 

[AbsoluteZero]

Distance from target is only considered after all valid targets have been chosen.

http://forums.civfanatics.com/showthread.php?p=7190899#post7190899

I don't care how it's supposed to be calculated. Based on many many MANY games my observation has been the AI will not march out of it's way to DOW it's worst enemy, instead it will go for a closer target. Of course being at pleased can change that, but at cautious you're in trouble if you're closer then an AI's worst enemy.

 

[Dirk1302]

Distance from target is only considered after all valid targets have been chosen.

http://forums.civfanatics.com/showthread.php?p=7190899#post7190899

True


But.....

Spoiler :

AZ is right, distance is more important than i remembered from investigations back in the days. Basically other things being equal (no land target >7 tiles shared) it comes down to:

1. a random number between 0 and 100 that is rolled. Peace prop for Monty at cautious = 30% so if a number = rolled under 30 you won't be declared on if it's higher you will be considered as a possible victim.

2. Once you're considered as a possible victim there are a few important factors:

a) How many border tiles do we share
b) How close are our capitals
c) Some general closeness value, don't feel like investigating that function to deeply now, you get the picture from a) and b) above

a), b) and c) are added, then....
d) this value gets multiplied by pleased:2, cautious:4, annoyed:8.

The player with the highest value is declared on

Now i thought that this last factor was in most cases the deciding factor but it's not that simple

I did a simulation in my own game 1800 BC so altering the Civ4 code (yes you can do that if you have the SDK and know C++ ) that each ai is forced to declare this turn. It turns out that Justin is so far away that he doesn't really come into the picture for Monty. In my game Monty is cautious with Zara and me, i'm a bit closer so i'm chosen, however if Monty would have been annoyed with Zara at that time Zara would have been chosen.

Values from simulation:

Dirk 403
Zara 356 would have been 712 if annoyed with Zara
Justin 136

Power rating doesn't figure at all in this. There is a power check before all this but there is no way a human player can have a strong enough power rating to fend of ai declaration. It's hard for ais as well as basically (with some xml modifications) the possible target must have 1.5 * power rating of the Ai that is considering a declaration.

Info is in my sig: Declaration mechanics DanF, very detailed, he doesn't go into the distance calculations too much so i decided to check how it is this game with this simulation. It's the same as Kossin's link.

 

[amanbe]

I stand corrected

 

[Dirk1302]

^You're in good company. Rusten one of the best players i've seen here was convinced power rating and so early barracks boosting power was important,and he was certainly not the only one, before DanF did that research. And that's how it should have been coded but unfortunately it isn't .

 

[Mylene]

Such things should be random imo anyway, the player is free to backstab at will, why shouldn't the AIs do the same.

 

[Dirk1302]

There must be some logic i think. Just because the human player can backstab at will doesn't mean he will. He will only do that if it's to his own advantage. Ai declaration mechanics is built to mimic this process to some extent, of course you go for someone who's closer. However you also favor those that are really weak above those who are only somewhat weak, closeness being approximately the same, and this is missing in the current declaration mechanics.

 

[TheMeInTeam]

^You're in good company. Rusten one of the best players i've seen here was convinced power rating and so early barracks boosting power was important,and he was certainly not the only one, before DanF did that research. And that's how it should have been coded but unfortunately it isn't .

The real question is whether those barracks and axes increase your likelihood of winning the game. Basically:

1. You can hold off and still win the game

or

2. You can hold off but will be crippled beyond repair

or

3. You can't hold off the attack even if you spam axes/rax

In the case of #1 axes/rax makes sense, in the others you're best off not bothering with defenses at all and hoping you get lucky. The real issue is how to judge this in each situation.

 

[Mylene]

There is no logic in begging 10 gold and getting peace for 10 turns.
No logic in building peacefully with 7 warriors guarding the cities and nothing else, while AIs that have traits made for war sit and watch.

The system is very forgiving for us anyway, a player would always take out a builder if he gets the chance and sees he is weak, or just if he has something like Jumbos while the AI has to run around with axes, swords and spears.

 

[Dirk1302]

Imo the 10 gold beg for 10 turns peace shouldn't be there. The begging is ok, the 10 turns peace isn't.

And indeed i think power rating should be included into the declaration mechanics. If the human player doesn't build any defence the chance to be declared on should be heightened significantly but it isn't.

@TMIT, i think it would be good if ,apart from being able to defend yourself, building barracks and axes would lessen the chance of a declaration on you. It would add depth since you really have to weight this factor risking a war against you if you only build some city police warriors.

Now as soon as the barbs are gone you can basically throw all thoughts about defense out of the window especially if the diplo situation is safe. If as Mylene sketches above city defense is very light Ais should be able to punish this regardless of diplo. A concept like this is not that easy to implement though. But at least weighing the power rating in choosing the target alongside closeness and attitude would have been something.

 

[AbsoluteZero]

@Dirk

Spoiler :

You wanted a game with the Monte backstab and it didn't take me long to get one. First try actually.



I was going to backstab him myself when I saw his stack.

How many games does that make? Probably like 6 by now. Nothing you can do when he starts planning at 2000 BC.

Save: View attachment MonteBackstab.CivBeyondSwordSave

 

[TheMeInTeam]

@TMIT, i think it would be good if ,apart from being able to defend yourself, building barracks and axes would lessen the chance of a declaration on you. It would add depth since you really have to weight this factor risking a war against you if you only build some city police warriors.

Now as soon as the barbs are gone you can basically throw all thoughts about defense out of the window especially if the diplo situation is safe. If as Mylene sketches above city defense is very light Ais should be able to punish this regardless of diplo. A concept like this is not that easy to implement though. But at least weighing the power rating in choosing the target alongside closeness and attitude would have been something.

I do try to keep my anger w/ firaxis to a moderate level on this supbforum, but it would be nice to have a LOT of things from this game. Control hotkeys that work. A consistently honest UI. Units that don't move on their own into dangers that are visible at turn start. A properly functional and non-broken AP. Bugs like "self-close" and overflow fixed. etc. If the AI could play better it wouldn't need the level of bonuses it has now to threaten good players. My point is similar however: as it stands right now, the majority of cases seem to dictate that cutting military is the optimal approach on deity, which IS sad.

Could relative power be a factor in DoW decisions? I have a feeling firaxis specifically avoided that because on emp+ it would virtually always target the human (especially on deity). Is that the right decision? Is power in its current form even reliable? This aspect of the AI is one of the more difficult; they can't even give it decent tile micro!

 

[Dirk1302]

@AZ

Spoiler :

I would need the save just before he went wheoohrn. But it doesn't matter, with equal attitude it's highly likely you're closest and this is certainly the most important factor with equal attitude. And no this save doesn't look too refreshing , too hard when he comes at you so early i agree.