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

Thoughts on A/B/C bonuses

Discussion in 'General Balance' started by Milae, May 22, 2020.

?

What do you think of A/B/C bonuses?

  1. Like them.

    8 vote(s)
    22.9%
  2. No strong opinion.

    4 vote(s)
    11.4%
  3. Don't like them.

    12 vote(s)
    34.3%
  4. Don't like the production part.

    16 vote(s)
    45.7%
Multiple votes are allowed.
  1. Rhys DeAnno

    Rhys DeAnno Prince

    Joined:
    Feb 22, 2011
    Messages:
    358
    Gender:
    Male
    Assuming that the production bonus isn't turned off in the next patch, how possible is it to modmod it off, or to modmod the magnitude of the various A/B/C bonuses in general? I think when I last looked at that part of the code there was stuff to control all the A/B/Cs at once but it wasn't separated into different bonus types.
     
  2. Recursive

    Recursive Emperor

    Joined:
    Dec 19, 2017
    Messages:
    1,737
    Gender:
    Male
    You can control the four main variables for all bonuses but there's no mechanism to control each one individually - that'd require four memory values for each yield.

    I think G has removed the Production bonus from a certain trigger for next version, though...
     
    CrazyG likes this.
  3. Rhys DeAnno

    Rhys DeAnno Prince

    Joined:
    Feb 22, 2011
    Messages:
    358
    Gender:
    Male
    Would it be possible to do it with multiplication, and just use one extra memory value per yield, multiplying with the general ABC structure?
     
    CrazyG likes this.
  4. Recursive

    Recursive Emperor

    Joined:
    Dec 19, 2017
    Messages:
    1,737
    Gender:
    Male
    Yes, actually. That's a good idea. But it'd need to be added to the DLL. I'll add that when I get back from my vacation - if you want it done sooner you'd need to persuade one of the other devs.
     
    CrazyG likes this.
  5. Stalker0

    Stalker0 Baller Magnus

    Joined:
    Dec 31, 2005
    Messages:
    6,752
    HOOOOOOOLD on a minute! We still haven't even decided if we are going to use this yet. Memory is such a precious commodity in the mod now a days (the damn missing yield memory issue was gotten worse and worse over time). I would ask lets not add extra memory use to the mod until we are confident we are going to use it.
     
  6. Recursive

    Recursive Emperor

    Joined:
    Dec 19, 2017
    Messages:
    1,737
    Gender:
    Male
    It'd be about 160 bytes. I deleted at least 6 KB of consumed memory for next version by removing useless diplomacy variables - and I deleted several times that in a recent version.

    More deletions of unused memory are coming, too.

    In any case, the memory footprint is in GB, and most of that is consumed by the UI.
     
    Last edited: Aug 2, 2020
    CppMaster likes this.
  7. Stalker0

    Stalker0 Baller Magnus

    Joined:
    Dec 31, 2005
    Messages:
    6,752
    That's great to hear, but I would say my point still stands. I know I'm going to sound really picky (it is only 160 bytes as you said which is almost nothing), but considering the game state is technically in a "bugged form"....where the mod pushes the memory so hard that we have disappearing yields... we should celebrate every byte that is freed....and harshly fight every byte that is consumed.

    If my assessment is wrong that even a 1 MB of space either way will make "no difference" to the user experience than I'll withdraw my concern, but I was also under the impression that "every byte helps" reduce the memory issues.
     
  8. Recursive

    Recursive Emperor

    Joined:
    Dec 19, 2017
    Messages:
    1,737
    Gender:
    Male
    Running out of memory tends to happen when modmods are used, particularly in the lategame. Eventually once enough units, cities, population, etc. are on the map it's going to run out. The impact of a few MB are negligible when the footprint is in GB, let alone 1/7,500th of a MB which is my proposition here.

    ilteroi has also made some significant cleanups in memory for next version, worth noting that.
     
    Last edited: Aug 2, 2020
    telelavoro and CppMaster like this.
  9. CppMaster

    CppMaster Emperor

    Joined:
    Feb 13, 2018
    Messages:
    1,197
    Gender:
    Male
    Location:
    Poland
    I agree. Let's not be paranoid about every MB.
     
  10. ICanHasName?

    ICanHasName? Warlord

    Joined:
    Feb 24, 2009
    Messages:
    257
    Why are these bonuses a separate system from the difficulty bonuses to yields? Why not just increase % bonuses instead of giving lump sums at irregular intervals? It's hard to quantify, and seems to cause weird imbalances from cities instantly gaining tiles, walls or shrines when they should have to build up gradually.
     
  11. Vhozite

    Vhozite Chieftain

    Joined:
    Dec 18, 2019
    Messages:
    92
    Gender:
    Male
    The more I think about this the more I think that the bonus hammers are my only issue with the bonuses.

    The impossibly fast shrines after settling messing up the religious race, walls built in a turn or two after the enemy forward settles you, and the AI dumping an impossible amount of hammers into WC projects. I feel like all those things are from the extra hammers. An entire overhaul of the system isn’t necessary imo but the hammers need to be addressed imo.
     
  12. Rhys DeAnno

    Rhys DeAnno Prince

    Joined:
    Feb 22, 2011
    Messages:
    358
    Gender:
    Male
    Getting rid of the ABC hammers and giving (stronger) cost breaks instead elegantly solves a lot of issues. It makes the Authority hammers more valuable to the AI (I think those hammers being meaningless is why AI Authority sucks now), it prevents a burst of new stuff in a new forward settle, and it means you can exclude certain things from the cost break, like Shrines/Wonders/Projects.
     
    CrazyG, a3kov, CppMaster and 2 others like this.
  13. tu_79

    tu_79 Deity

    Joined:
    Feb 11, 2016
    Messages:
    7,165
    Location:
    Malaga (Spain)
    Let me refresh things for the people who are new to this old debate.

    The handicap bonuses have three aspects:
    1. The trigger. This is the action which deserves to be awarded. Gazebo decided long ago that it should only reward AI who achieve anything, so the ones that are performing slightly better become stronger and threatening. That makes a few lucky AI to snowball and gives the human a sense of fighting against some superpowers.
    2. The yields that are rewarded. Culture has been banned for most handicaps, and some triggers get just some gold (world wonders I think).
    3. The scaling of the yields. When we say A/B/C boni, ABC represents the way these boni scales. It's a quadratic function that allows AI to be competitive against the human in all eras.

    So far, the quadratic scaling worked well. We had some issues with the kind of yields and those got solved. For example, it no longer scales on the era of the AI, but on the average era. So now the only controversial debate is what should be the triggers. @BiteInTheMark and I have been asking for adding a timed trigger, something like 'trigger every 20 turns', or 'trigger every 5 worldwide techs', along with reducing the amount of the handicap. This will make every AI slightly stronger, and some AI will still be runaways but the difference shall not be as high as now. Note that this IS NOT a rubber-banding mechanic, it is just an increase on the global difficulty that scales at a human pace and a reduction on the snowballing. A defeated civ is not going to come back because of this handicap trigger, since all AI are getting it.

    The counter argument is that all AI are very similar in intelligence, so if some AI are not getting these big rewards compared to the other AI, they can't beat their neighboring AI. In other words, our idea may result in very few AI being able to expand. (Now that diplomacy is getting more advanced, this should not be a huge issue).
     
    BiteInTheMark likes this.
  14. CrazyG

    CrazyG Deity

    Joined:
    Oct 14, 2016
    Messages:
    5,332
    Location:
    Beijing
    In my next game I'm going to try modifying the AI decision process. For those who don't know, when the AI assesses it's available choices, on most difficulties the AI chooses randomnly between it's two best options. On Deity it always chooses the "best" one. This leads to a big lack in variety, often AI with the same social policies take identical tech paths.

    The AI is good at evaluating but not good enough that number 1 and number 2 are consistently that far apart in quality, and randomness would help them a lot. If everyone else is doing A, it often makes B a better strategy. For example multiple players have noticed the Leaning Tower of Pisa is a better wonder to pursue than Chichen Itza, just because of the speed the AI builds these wonders.
     
    Last edited: Aug 6, 2020
    tothePAIN and Rhys DeAnno like this.
  15. Rhys DeAnno

    Rhys DeAnno Prince

    Joined:
    Feb 22, 2011
    Messages:
    358
    Gender:
    Male
    I predict this makes Deity way harder just because Terracotta gets built now. I guess maybe there are less tradition ai on average though? And more progress?
     
  16. azum4roll

    azum4roll King

    Joined:
    Jul 17, 2018
    Messages:
    992
    Gender:
    Male
    I wonder about the weighting of the random choices. Are rank 1 and rank 2 equally weighted, weighted linearly to the policy/tech/building score, or something else?

    I think a quadratic score weighting would be the best.
     
  17. Drakle

    Drakle Warlord

    Joined:
    Jun 28, 2014
    Messages:
    259
    The AI is ridiculous. I'm Korea, and I'm over 10 techs ahead of Greece. The recent production per citizen chart had me at over twice his. I have more citizens. I have more cities.

    Than Treasure Fleet. The whole thing lasted 2 turns. The first turn completed 90% on Standard. Nearly all my cities were working on it. Then next turn it completes. Nearly all civ got bronze and silver.

    But Greece got gold. Despite being one of the bottom civs, and still working its way through Statecraft, while I and a few other AI are in their third tree.

    And I was second last on small map size in terms of production contributed. Despite being the tech leader, insanely productive cities, and nearly all my cities working on it.
     
    Last edited: Aug 7, 2020
    Vhozite likes this.
  18. CppMaster

    CppMaster Emperor

    Joined:
    Feb 13, 2018
    Messages:
    1,197
    Gender:
    Male
    Location:
    Poland
    That's because of instant production yields. Many wants them to be gone
     
  19. tu_79

    tu_79 Deity

    Joined:
    Feb 11, 2016
    Messages:
    7,165
    Location:
    Malaga (Spain)
    Iirc, the production bonus is applied to every city of the player. I can't remember if it was a simple value that is split among the cities or if it scaled on the number of cities.
     
  20. Recursive

    Recursive Emperor

    Joined:
    Dec 19, 2017
    Messages:
    1,737
    Gender:
    Male
    The full yield bonus is applied as a Production yield in every AI city.
     

Share This Page