1. We have added a Gift Upgrades feature that allows you to gift an account upgrade to another member, just in time for the holiday season. You can see the gift option when going to the Account Upgrades screen, or on any user profile screen.
    Dismiss Notice

Anti-Wonder Snowballing

Discussion in 'General Balance' started by Gidoza, Mar 6, 2020.

  1. tu_79

    tu_79 Deity

    Joined:
    Feb 11, 2016
    Messages:
    6,916
    Location:
    Malaga (Spain)
    I think that would be a fine thing to do.

    It's always worth to look at the facts, and the fact says that it's not the AI who build most wonders who is too far away, but the AI that reaches new eras in advance. The difference in yields from handicaps is very noticeable even in King difficulty. I can only guess how strong it gets in higher ones.

    I know you want some epicness with some runaway adversaries, so if it flattens too much, you can always give a little more yields on events.
     
    CrazyG likes this.
  2. Stalker0

    Stalker0 Baller Magnus

    Joined:
    Dec 31, 2005
    Messages:
    6,158
    I’m for it as well.

    I think the issue with wonders is the weakness of culture rubber banding. Once a AI becomes culture leader (through the various AI bonuses), it takes a lot to overcome that. And when your the only civ with access to a tier of wonders for 10 turns...your going to get them.

    the simple answer may be to remove culture restrictions on more wonders. Certain wonders remain the purview of culture focused players, but more wonders become competitive in the general pool.
     
  3. tu_79

    tu_79 Deity

    Joined:
    Feb 11, 2016
    Messages:
    6,916
    Location:
    Malaga (Spain)
    Perhaps. Perhaps it would be enough to use average era for scaling, as showed. And it would a behind the curtains change no one would notice, except for the effects on AI distribution. Removing cultural restrictions would be a change in mechanics for all players, and these restrictions were put there for a reason.
     
  4. BiteInTheMark

    BiteInTheMark Emperor

    Joined:
    Aug 23, 2016
    Messages:
    1,812
    Gender:
    Male
    Location:
    Germany
    If the amount is so marginal, why do they even get those yields? Buildinga wonder or finishing a trade route is beneficial in it self, why should successful AI get even more yields, istn this much in favor for runaways?
     
  5. tu_79

    tu_79 Deity

    Joined:
    Feb 11, 2016
    Messages:
    6,916
    Location:
    Malaga (Spain)
    I think the real question is what causes what. Is building wonders what pushes civs ahead by the success of the wonder and the extra yields or is it civs that are already ahead who get to build wonders in the first place?
    Maybe it's a feedback mechanism, in this case, which is the strongest force behind the observed differences? Building wonders or being ahead? Take a look at the AI handicaps log, maybe this will clarify things for you.
     
  6. BiteInTheMark

    BiteInTheMark Emperor

    Joined:
    Aug 23, 2016
    Messages:
    1,812
    Gender:
    Male
    Location:
    Germany
    If the AI worker and build logic is causing the increased difficulty, then its fine, cause we all want an intelligent as possible AI.
    But in this case, every AI would be more powerful, and a lot of people report from extreme runaways.
    Personally, I dont understand, why rubberband mechanics are so hated, but instead, we have a push-runaways mechanic, and it everyone seems ok with it.

    I would call it a flexible difficulty adjustment. If the player does very well, then the bonuses for the AI increase, up to a degree it is half the way to the next difficulty level. If the player do very bad, then it decreases down to half the way of the lower difficulty.
     
    Heinz_Guderian likes this.
  7. tu_79

    tu_79 Deity

    Joined:
    Feb 11, 2016
    Messages:
    6,916
    Location:
    Malaga (Spain)
    I think I know the reason. If the AI adapts tightly to the player skills, then there are no difficulties. You can't know whether you are progressing. You might never get better since you can always get your 50/50 win/loses quota, no matter how you play. There's no incentive to get better, either, since the game will not feel different after you have improved.
    And without the challenge, there goes the fun.
    This is the issue with extreme rubberbanding. Even AI suffer from rubberbanding. If a bad performance is covered by extra yields, then it becomes harder to notice that this AI, or this civ, is performing badly.

    About handicap bonuses, there are two kinds:
    1. Those that are given equally to every AI. They exist to increase difficulty versus the human player, where the AI cunning does not reach.
    2. Those that are given to AI after a successful event. They exist to make one or two civs look like they are the archenemies, so the player has an easy story to tell, of good and evil, the human overcoming the powerhouse runaway civ. They also exists so one civ can get strong enough to overcome the strength of their neighbours and be able to conquer some cities.

    Only that currently the runaway seems overtuned, especially in higher difficulties. Remember that G has done a first round of tuning this down, by removing culture from quite some events.
    You could be right about every AI getting stronger, but consider that the idea is to use the world average era for scaling, so that's more yields to the loser civs and less to the winners. The risk of flattening the bonuses too much is AI never being able to expand, for the bonuses on the losers hold them up.
     
    Heinz_Guderian likes this.
  8. BiteInTheMark

    BiteInTheMark Emperor

    Joined:
    Aug 23, 2016
    Messages:
    1,812
    Gender:
    Male
    Location:
    Germany
    I am not thinking about that huge adjustments you might think. If Emperor AI gets for finishing a wonder 50 yields and an immortal AI 60 yields, then the range of adjusted yields could range from 45 to 55 for emperor AIs and 55 to 65 for immortal AI.
    You might say, "then the difficulty for a bad immortal player is the same as for a good emperor player", but you would leave out the fact, that an emperor and immortal AIs still get different advantages (production discout, improvement rate, supply cap,...) and still have their normal yields.
    I think the method I presented, showded, that there is a middle ground between..... every city gets same amout of yields favoring extremly wide empires .... and .... get a a static global yield which favors extremly tall play ..... cause the same amount of science is worth more for small than wide empires.

    The main question is..... are those yields for events really necessary? The one who collect some wonders first already is in advantage over other AIs, making it more likely, that this AI will collect even more wonders/cities/techs/....
     
    Bromar1 likes this.
  9. Heinz_Guderian

    Heinz_Guderian Chieftain

    Joined:
    Jan 2, 2020
    Messages:
    82
    Location:
    Sweden
    I have to say I agree with you both partly. I think it is weird to say that rubber banding is evil and then actively promote runaways with event based yields. Doesnt even matter if the yields for wonders arent huge, it is symbolically wrong either way. And of course they make some difference. So I agree with Bite there.

    I really agree with Tu about the flexible difficulty (and rubber banding in general) being a very bad idea. It would remove so much flavor from the game, would make balancing somewhat pointless, would make the performance from the player irrelevant etc etc. We want to improve the AI and make the game harder in that way, and imo the levelling out of the inadequacy in terms of AI skill should comprise of flat bonuses for every AI. In that way, balancing would be much easier, and the AI should perform more predictably, which would make difficulties easier to manage.
     
    Recursive likes this.
  10. Recursive

    Recursive Emperor

    Joined:
    Dec 19, 2017
    Messages:
    1,218
    Gender:
    Male
    I've readded the "wonder spammer" modifier (after fixing it, since it was not working as intended), so wonder spamming will be punished with more DoWs. If it's insufficient I can look into additional scaling, i.e. +WAR/HOSTILE for each Wonder above the global average.

    Two of the natural weaknesses of spamming Wonders are that turns spent building Wonders can't be used building military units, and cities with lots of Wonders make attractive conquests. :)
     
    vyyt, Zanteogo and burleigh like this.
  11. Zanteogo

    Zanteogo Prince

    Joined:
    Nov 13, 2015
    Messages:
    483
    Overall I find the average wonder has become weaker over the years of tweaking, other than a few stand outs, wonders aren't has game changing as they used to be. With a few being not worth the time, particularly with the building penalty for the next wonder on top. (I'm looking at the religious wonders who's main effect is giving a free religious building...)

    You can actually be fine not building a single wonder for the whole game a survive. In games where I have built zero, the only real issue is the happiness bonus many wonders have tacked on to them now.

    Though it's probably too late for such a change, I would love to see wonders as a whole increased in power, with the building penalty for further wonders increased a lot. Particularly the "same era" penalty. With the increase in power it will encourage targeted wars to take said wonder. With the increase in penalty it will allow the civ's that are ahead to get "first pick" of a wonder, a wonder that will match their civs play style, but allow other civs to take whats left, or decide to not bother and focus on building military units to perhaps steal a wonder by force.
     
    Last edited: Apr 2, 2020

Share This Page