Removing Growth from AI Handicap list

[Owlbebach]

EDIT: added a poll

For the past several months I've been analyzing why my experience regarding many items of the game differs a lot compared to many other people, especially those who play Emperor difficulty and lower.
Why am I worried about this? I want this game to be in a better state and in we are to make a balance change - there should be an agreement in the community on that change.

Problem is that right now there is a huge difference in game balance between difficulties. It is not only harder to play on higher difficulties. The problem is that many things on higher difficulties (Immortal and Deity) do not work the the same as they work on lower difficulties. Balance is different.

Why it is important? It causes disagreement in the community on balance questions. For example i'm sure ALL Deity players will agree that Goddes of Love is unplayable on Deity, but for a King player it might seem okay.

I've identified one thing that really throws balance off when moving up in difficulty level and that is AI growth discount. AI cities on Immortal and Deity grow stupidly fast. There are others for sure, but i know this one and i know how to fix it

Spoiler Several examples :

My Capital has a population of 5. City of Guangzhou has 8 and it is China's 3rd or 4th city.
This is TURN 63, that the city was founded ~ around turn 40, which means it grows 7 population in 20 turns!


Turn 88, Washington has 6 pop on floodplains with Tradition. Beiging has 13 (to the right, under Hiawata)


Late into the game my cities have 17-19 population. AI cities have ~30


Latest game from @chicorbeef nuff said

Now, why it is important? Because there are LOTS of game aspects that are tied with population and births of new citizens. Here are those that i can remember right away:

Pantheons:
God-King, Goddess of Love, Earth Mother, Ancestor Worship. ALL of them are by far the best choices for AI on Immortal+. Love and Ancestor consistently found 15-20 turns before anyone else, God-King consistently snowballs. At the same time they all are absolutely unpplayable for human. They are just awful.
Followers:
Cooperation - the best belief by a huge margin for AI. Showballs out of control in 90% of games, again absolutely useless for player
All beliefs that are "+1 yield per 2 followers"
Enhancers:
Suprisingly, Scripture and Ritual. More population >> more followers >> more pressure, snowballs frequently
Policies:
Progress (receives science and gold when citizen is born). AI with Progress civs consistently outperform others on Deity
Ideologies:
Freedom (AI with universal healthcare ridiculously snowbals in culture)
Specialists and Great People
Everything that is Great Person-related (more citizen >> more specialists >> more great people), Arabia, Glory of God, Ceremonial Burial

Understand me right here. I am not asking to make Deity easier. The problem is that AI with Goddess of Love on Deity and AI with Goddess of Love on King have HUGE difference. On King it is in the middle of the pack, and on Deity it founds religion on turn 50. This means that when i come to the thread and say "goddess of love should be changed" - Deity players agree with me, while everyone else does not because they do not see the problem.

I want us all to have the same impression on balance stuff. This is why i am proposing this change. As far as i understand from DifficultyMod.xml - it is very easy to change that, we only need to change AIGrowthPercent and balance it after (http://civ-5-cbp.wikia.com/wiki/AI_and_Difficulty). Of course i can change it in my own games, but i want the whole community to be on the same page.

@Gazebo @CrazyG @ElliotS @chicorbeef @ashendashin thoughts?

 

[Der_Zorn_gottes]

Not really helpful, but it remindes me of a weird anecdote:

Once, I was playing tradition with a food monopoly. In the modern age, a TC plotted a city on my island. That city had only desert and water in its borders, and not a single food resource. Within just a couple of turns, it has outgrown my captital. It was hilarious

 

[chicorbeef]

The question is what will you replace it with? People will have problems with Culture/Production handicaps as well.

 

[Owlbebach]

Not really helpful, but it remindes me of a weird anecdote:

Once, I was playing tradition with a food monopoly. In the modern age, a TC plotted a city on my island. That city had only desert and water in its borders, and not a single food resource. Within just a couple of turns, it has outgrown my captital. It was hilarious

I couldn't find the screenshot, but once upon a time i had Inca who had a 25-pop city that had ONE non-mountain tile. And it was the time when Inca did not have food on mountains!

 

[Owlbebach]

The question is what will you replace it with? People will have problems with Culture/Production handicaps as well.

I think production handicap will work just fine. Don't see it having those issues. There is not too many things triggered by building stuff. I can thing of a Progress policy only and well buildings are limited. I had many games where AI was sitting on 50 thousands of gold cause it had nothing to build

 

[chicorbeef]

I think production handicap will work just fine. Don't see it having those issues. There is not too many things triggered by building stuff. I can thing of a Progress policy only and well buildings are limited. I had many games where AI was sitting on 50 thousands of gold cause it had nothing to build

I do actually like the idea of removing Growth handicaps but I just fear that the AI will be hurt and we will have to compensate it somehow. Production handicaps can snowball pretty hard, it will make getting a Religion a lot harder, etc.

 

[Owlbebach]

I do actually like the idea of removing Growth handicaps but I just fear that the AI will be hurt and we will have to compensate it somehow. Production handicaps can snowball pretty hard, it will make getting a Religion a lot harder, etc.

Not that much. Keep in mind that AI will have less citizen thus less production early on. I think only a small additional discount will be required because every next citizen is less usefull, especially lategame

 

[ashendashin]

This is definitely a longstanding balance issue and you explained it well, but removing it entirely might be too much. This is AI growth combined with everything else they get so beefing up other areas in its place could easily cause larger imbalances. It's also easy to see how tradition seems so weak when you can't surpass the AI on a per city basis, even against wide civs (though I'm sure the specialist nerfs were a bit much, I bet it mostly made synergies harder).

 

[Owlbebach]

This is definitely a longstanding balance issue and you explained it well, but removing it entirely might be too much. This is AI growth combined with everything else they get so beefing up other areas in its place could easily cause larger imbalances. It's also easy to see how tradition seems so weak when you can't surpass the AI on a per city basis, even against wide civs (though I'm sure the specialist nerfs were a bit much, I bet it mostly made synergies harder).

To me it seem that it will not bring a lot of balance issues. I'm not saying that it can be removed and everyone will be happy, but much-much bigger changes were made successfully to this game. I'm sure it will require everyone to play 2-3 games and report to set right numbers

 

[CrazyG]

Something I don't see mentioned here is different unique abilities and buildings. If I grew like the AI does, I think the Harrapan reservoir or the Floating Gardens would be much less valuable.

 

[Owlbebach]

Something I don't see mentioned here is different unique abilities and buildings. If I grew like the AI does, I think the Harrapan reservoir or the Floating Gardens would be much less valuable.

Why is that? Human's population will remain the same and actually AI will still be ahead of human cause it build everything much faster, thus will be able ti build Granaries/Groceries etc much earlier than human. Did you meant more valuable? Or did you mean that this will make Aztec/India AI weaker compared to others?

 

[CrazyG]

Why is that? Human's population will remain the same and actually AI will still be ahead of human cause it build everything much faster, thus will be able ti build Granaries/Groceries etc much earlier than human. Did you meant more valuable? Or did you mean that this will make Aztec/India AI weaker compared to others?

I mean +1 food on river tiles is very, very good for human players. For AI its meh. They already have so much growth. The extra growth on AI makes warmonger better as well. Capturing their cities gives you a higher population that settling your own city does.

With that said, I think their bonus growth is a fundamental part of the increasing difficulty. Perhaps it could be reduced and another aspect added to compensate, but I think we need a detailed proposal with all thought explained. Being experienced at the game doesn't mean we understand the AI all that well. A while back I did push for a big change to AI's border growth, but border growth is minor compared to pop growth.

 

[Owlbebach]

I mean +1 food on river tiles is very, very good for human players. For AI its meh. They already have so much growth. The extra growth on AI makes warmonger better as well. Capturing their cities gives you a higher population that settling your own city does.

Actually this is also true. Many times i had situation when i capture cities that were higher pop and better infrastructure than my capital

With that said, I think their bonus growth is a fundamental part of the increasing difficulty. Perhaps it could be reduced and another aspect added to compensate, but I think we need a detailed proposal with all thought explained. Being experienced at the game doesn't mean we understand the AI all that well. A while back I did push for a big change to AI's border growth, but border growth is minor compared to pop growth.

Well i have 2 options for detailed proposal.
1) Remove Growth handicap entirely, but compensate with additional building cost reduction (i'd propose 55-60% instead of 70%). Maybe additional tech cost discount can be introduced, but i'm not sure. This will make AI build faster so its population will still be higher than players population, but not THAT higher.
2) In case someone is afraid that it will break the balance - we can set it to 90% for Deity and Immortal and 100% for everyone else and compensate by a smaller building cost discount.

Anyway it is always a matter of testing. I think that recent Distress introduction shakes balance much stronger than what i propose. In fact everyone can try it with any new number and share results. It is really easy to modify it vie DifficultyMod.xml
Think of this that way: in any case this change will affect only Deity and Immortal. If we agree that game is more or less balanced on King and Emperor - this will only make Deity/Immortal experience closer to King/Emperor rom the point of relative balance between aspects of the game (i mean things like Pantheon A vs Pantheon B of Deity compared to A vs B on Emperor). Even theoretically, it simply can't make difference between Emperor and Immortal bigger than it is right now!

 

[tu_79]

Not building discount. It will make progress much better for AI.

I never play over Emperor, but if you want to strengthen AI in some aspect, more culture or faith, or more unit strength seems the right call. Both science and gold yields would break their Goddesses pantheons, the same growth breaks their respective pantheons.
I'm inclined towards more combat bonuses, as it seems that winning wars is so much better for you in the higher levels.

 

[Owlbebach]

Not building discount. It will make progress much better for AI.

Well it is already much better for AI, especially on higher difficulties. Given that AI will have less population i think it will remain about the same. Or at least it is possible to make it about the same after we figure out right nimbers

 

[tu_79]

Well it is already much better for AI, especially on higher difficulties. Given that AI will have less population i think it will remain about the same. Or at least it is possible to make it about the same after we figure out right nimbers

Once you take away growth bonuses, Tradition and any gp related thing will suffer. By enhancing building production you are just increasing the difference.

 

[Owlbebach]

Once you take away growth bonuses, Tradition and any gp related thing will suffer. By enhancing building production you are just increasing the difference.

Not really. In fact the opposite. You thing of this comparing numbers, but you actually should think of this in terms of relative effect. Tradition will suffer, but so will other trees. And Tradition Growth bonuses will become more important. As it is right now - AI has so much Growth that there is no difference between Tradition and Authority in term of Great People because AI is able to fill every specialists slot as soon as it is avaialble. Without Growth handicap Tradition will finally have more population compared to two other trees.

 

[Moi Magnus]

Would you find acceptable to reduce (or get rid of) the growth bonus for the early game, but keep it in mid-late game.

(Like most other AI bonuses)

 

[Owlbebach]

Would you find acceptable to reduce (or get rid of) the growth bonus for the early game, but keep it in mid-late game.

(Like most other AI bonuses)

IMO it will definitely be a step into right direction, however i am quite confident that (1)removing it entirely is better (2)removing it entirely requires modification of one .xml file, thus it is very easy to do compared to your idea

 

[ElliotS]

I think it does cause balance concerns, and think that reducing or removing it would make things better IF we could do it without destroying balance.

My instincal reaction is that it will be harder than you think. I also think SOME bonus is fine for it, just less than now. If AI's Growth Percent went from 70% (as it is now) to 90% I think it would be easier to balance and more fair. The AI does mess up tile choices pretty often, so some bonus is needed. I would go 100-100-100-98-96-94-92-90 for handicaps there.

The problem is that they already get 70% cost for buildings. Buffing that means that they'll be working processes even more often leading to massive culture or science snowball.

I'm in favor of you trying out numbers to see what works and if you find something happy to see G try and implement it, but this seems like a lot of trouble.

I think that recent Distress introduction shakes balance much stronger than what i propose.

I don't. We've had plenty of patches with happiness too easy or too hard. Other than walls line needing changes nothing else really changed compared to normal in a negative way. The mechanic is a big improvement imo.