BiteInTheMark
Deity
With AI only observer games? If yes, maybe this is the problem.
New Version - March 11th (3-11)
- Thread starter Gazebo
- Start date
- Status
Lol *munches on popcorn*...
G I know the end goal is for everything to be as balanced as possible, but I'll ask again. Do you have a list or tiers of any kind for overall civ strength/performance? It matters not, except I'm somewhat curious.
Also, you're able to pick 2 reformation beliefs right? I had built CoSB in a game (2/27 version), and later ended up completing my holy wonder(?), except it didn't allow me to pick another reformation belief and I was bummed. I remember on a previous version being able to pick 2 (I built holy wonder before CoSB that time), so I don't know if something has changed, or if it was just a bug. I'm guessing a bug, but seeing as how I didn't try to replicate it, I'm not quite sure and would just like some clarification before I waste hammers on CoSB in my current game.
Keep up the great work by the way. I'm thoroughly enjoying the latest release! The new puppet changes specifically are great IMO, allowing for much less micro-manage, and more overall fun. I'm staying afloat much better through domination (granted I only play Prince upon testing new versions), thus allowing me to annex outright and have to puppet less in the first place (which you'd mentioned before puppets are intended to be temporary for the most part). I'm just about to grab Imperialism and am excited to see how it all pans out (wonder spam Egypt is ahead of me in policies and has already taken Imperialism after going Tradition/Artistry so that's interesting considering he hasn't been involved in a single war all game on his own continent...).
Deadstarre
Expert
no. building cosb would be a waste
There's no intended tier list, no. I'd say that the range is more of a 'always useful if average traits' versus 'very powerful but niche traits.'
G
That's a good description which makes sense. Thank you.
BiteInTheMark
Deity
If you guys say, the AI is now much better than some time ago, you really should watch the AI more carefully.
Watching them squeezing more and more cities in tiny spots between already established cities, without evaluating the possibility of hostility or spamming cities on every one tile 2 fish islands, leading me to the conviction, the basic AI routine is as stupid as it was one year ago. (In the aspect of settling cities)
The only thing that has changed noticeable is their building logic and anti runaway logic, and maybe some logic to warfare.
And cause creating the right buildings in the right time is the key of civilization, making them strong together with their handicap.
Watching them squeezing more and more cities in tiny spots between already established cities, without evaluating the possibility of hostility or spamming cities on every one tile 2 fish islands, leading me to the conviction, the basic AI routine is as stupid as it was one year ago. (In the aspect of settling cities)
The only thing that has changed noticeable is their building logic and anti runaway logic, and maybe some logic to warfare.
And cause creating the right buildings in the right time is the key of civilization, making them strong together with their handicap.
tu_79
Deity
Were you playing with Byzantium the other time?
Building your founder building after St Basil does not grant another reformation belief. Never did. But maybe you are mistaken if you played Byzantium.
So, you agree that AI is better now. Recap. Following your own sentences: AI is as stupid at settling cities as it was a year ago. AI is still making mistakes at positioning military units (though we all think that it makes fewer mistakes than before). Those aspects are no worse than before. On the other hand AI is better at building order, and AI diplomatic logic is better at handling runaways... Those aspects are better than before. In conclusion, AI is better. (Or you do think that building order logic is not AI?)If you guys say, the AI is now much better than some time ago, you really should watch the AI more carefully.
Watching them squeezing more and more cities in tiny spots between already established cities, without evaluating the possibility of hostility or spamming cities on every one tile 2 fish islands, leading me to the conviction, the basic AI routine is as stupid as it was one year ago. (In the aspect of settling cities)
The only thing that has changed noticeable is their building logic and anti runaway logic, and maybe some logic to warfare.
And cause creating the right buildings in the right time is the key of civilization, making them strong together with their handicap.
It would be nice if settling cities is improved a bit more, but I'm seeing AI delaying to settle when its capital is doing better things, and I've never seen AI settling in bad spot before a good one. So AI is not doing that bad in this regard. Maybe it's not considering properly the risk of settling near a warmonger without a reasonable army to protect new cities, but sometimes, no risk, no gain. AI cannot know in advance whether the human player is going to play peacefully (though maybe it could look at the warmonger threat status, this value that is used for the warmonger penalty).
Last edited:
BiteInTheMark
Deity
I agree, AI is now better. But not as much better as a lot of people say. The main change is the better building logic. This gets exponentially raised by the handicap the AI gets. If you are able to build everything in time it gets unlocked, there is no big need for logic or strategic superiority. For me, all the strenght of the AI comes from the production modifiers. And less from strategy.
The hammers influence your abiliy to create your infrastructure and go to war.
And with the raised effect of population and the introduction of the need buildings, the hammers directly influence your happiness to. And with -5 instead of +5 happiness, it influences your science, gold, culture too.
Iam sorry for this, but this is my view. I like the happiness system, as it is, but it is now too influential on the game, and the hammer handicap of the AI has too much influence on the happiness too.
It is like you spread the handicap modifier not only on production cost, but also on culture or science cost.
If I say, the handicap modifiers are making the AI too difficult, Iam told, I have to go down with difficulty.
But I would prefer to play against AI which make the best, perfect calculated choice but with less extra yields by handicap. Like the AI is playing with deity logic but with king handicap. That would be fine (or something like that).
And I dont think its a good idea to balance a game with warlord difficulty games, while most players are using king/emperor difficulty.
Grabbl
Prince
- Joined
- May 25, 2016
- Messages
- 431
It is only natural that AI makes less errors in building logic, it is much easier to optimize because it has far less degrees of freedom than tactical AI. But if you don't only focus on errors of the tactical AI (of course, they exist, I got so many great generals without sufficient production, and AI is still not opportunistic enough to capture my unprotected first settler, and sometimes fails to see that he needs to launch his nuclear missiles or else I will launch mine), but it is able to many more things now than even one year ago (not speaking of Vanilla even...). Naval invasions can not be automatically defended anymore, AI can set up "unbreakable" defenses with the right terrain, uses interception, pillages important roads and resources... I don't even want to imagine the amount of good work that went in there.
That would be interesting for me as well, however. AI takes the best option from difficulty 7 onwards, does it? And one of the two best options below?
Moi Magnus
Emperor
- Joined
- Mar 1, 2015
- Messages
- 1,867
One of the 3 best at low difficulty (1-3), one of the 2 best at mid difficulty (4-6), the best at high difficulty (7-8).
If I remember right, I think Gazebo has mentioned before that he varies the difficulty setting used in testing, so sometimes it is Warlord and sometimes it is not.
Ahhhh you're right I was using Byzantium! Good call. I remember now it was the version where G updated Theodora, so I gave her a go.
Deadstarre
Expert
i've almost said this to you a thousand times, youre going to be a lot happier once you learn to start making some simple edits. DifficultyMod.xml
Code:
<CityProductionNumOptionsConsidered>1</CityProductionNumOptionsConsidered>
<TechNumOptionsConsidered>1</TechNumOptionsConsidered>
<PolicyNumOptionsConsidered>1</PolicyNumOptionsConsidered>BiteInTheMark
Deity
Thank you.
LukaSlovenia29
Emperor
- Joined
- Mar 13, 2016
- Messages
- 1,500
Deity AI is brutal. With the exception of a few Ottoman games, I've been consistently falling behind and being DoW'ed. I've tried Portugal, India and now one with Korea. The AI really is very good. Ughhh, I need a break from this "self-flagellation" that is Deity VP.
It's about to get slightly harder, too. Whipping up a tasty hotfix for you all.
glider1
Deity
Biggest problem I'm seeing with the tactical AI is it is still tending to throw the occasional single unit suicide style into enemy lines without backup, support or any sensible reason. It just "sees" a path forward and launches itself right into the jaws of the enemy. Can that be fixed by giving them more line of sight with a mod-mod? I know it comes with a performance penalty.
Main problem I'm seeing with the operational AI is that the AI will tend to pull out and retreat from a battle a little too willingly but that is so hard to judge. Is there are a mod variable that controls the propensity to retreat in the operational AI?
Main problem I'm seeing with the operational AI is that the AI will tend to pull out and retreat from a battle a little too willingly but that is so hard to judge. Is there are a mod variable that controls the propensity to retreat in the operational AI?
Deadstarre
Expert
im seeing the retreats as well, even taking their garrisons with them prematurely
- Status
Similar threads
