New Beta Version - November 8th

[Coffee Monopoly]

major competitors

How do they define major competitors?

 

[Recursive]

How do they define major competitors?

/// Is this player one of our major competitors?
bool CvDiplomacyAI::IsMajorCompetitor(PlayerTypes ePlayer)
{
   if (!IsPlayerValid(ePlayer) || GET_PLAYER(ePlayer).isMinorCiv())
       return false;
  
   if (GetPlayer()->GetCapitalConqueror() == ePlayer)
       return true;
  
   if (GC.getGame().countMajorCivsAlive() == 2 && !IsNoVictoryCompetition())
       return true;
  
   if (GET_PLAYER(ePlayer).GetDiplomacyAI()->IsCloseToDominationVictory() || GET_PLAYER(ePlayer).GetDiplomacyAI()->IsCloseToDiploVictory() || GET_PLAYER(ePlayer).GetDiplomacyAI()->IsCloseToSSVictory() || GET_PLAYER(ePlayer).GetDiplomacyAI()->IsCloseToCultureVictory())
   {
       if (!IsNoVictoryCompetition())
       {
           return true;
       }
   }
  
#if defined(MOD_DIPLOMACY_CIV4_FEATURES)
   if (MOD_DIPLOMACY_CIV4_FEATURES)
   {
       if (GET_TEAM(GET_PLAYER(ePlayer).getTeam()).IsVassalOfSomeone())
           return false;
   }
#endif

   if (GetWarmongerThreat(ePlayer) >= THREAT_SEVERE)
       return true;
  
   if (GetMajorCivOpinion(ePlayer) == MAJOR_CIV_OPINION_UNFORGIVABLE)
       return true;
  
   if (GetBiggestCompetitor() == ePlayer)
       return true;
  
   if (GetVictoryDisputeLevel(ePlayer) >= DISPUTE_LEVEL_STRONG || GetVictoryBlockLevel(ePlayer) >= BLOCK_LEVEL_STRONG)
       return true;

   return false;
}
#endif

Or in plaintext:
- Must be a major civ
- Must be a player we've met and is alive
- Must meet one of these conditions:
* Player captured our capital
* There are only two major civs left
* They're close to any victory condition

- OR must meet one of these conditions and not be a vassal of someone:
* Warmonger threat of SEVERE or higher
* Opinion of UNFORGIVABLE
* Be the biggest competitor (highest competitor score from DoRelationshipPairing)
* Victory Dispute/Block of STRONG or FIERCE

 

[tu_79]

The only credit I give to this troll is that late game turns are very long. But we did make the choice between long turns and stupid AI, and we wanted smart AI. The only way to ease the processing time then it is to reduce the number of units.
But here we face again a choice, and people decided we wanted so many units.
So, it's either to play smaller maps or tweak values for getting smaller armies.

If I wanted every player in my game to have fewer units, what values should I change?

 

[Recursive]

The only credit I give to this troll is that late game turns are very long. But we did make the choice between long turns and stupid AI, and we wanted smart AI. The only way to ease the processing time then it is to reduce the number of units.
But here we face again a choice, and people decided we wanted so many units.
So, it's either to play smaller maps or tweak values for getting smaller armies.

If I wanted every player in my game to have fewer units, what values should I change?

Easiest would be to reduce the unit supply cap in DifficultyMod.xml. The relevant values are ProductionFreeUnits, ProductionFreeUnitsPerCity, and ProductionFreeUnitsPopulationPercent. You have to edit the values for the handicap you plan on playing on.

For the AI, their unit supply values are under the AI_DEFAULT difficulty level, and they apply regardless of whatever difficulty you are playing on.

They also get a bonus to the total supply cap from AIUnitSupplyPercent and AIPerEraModifier (determined by what difficulty you're playing on).

Keep in mind reductions from tech level and war weariness apply.

https://civ-5-cbp.fandom.com/wiki/Customizing_Difficulty_Levels

 

[General_Drax]

Agree with Recursive.

At this stage of development could be difficult to balance.

Late game turns take very long especially for Domination. Player got massive army and is fielding multiple wars.

Other victory conditions are faster turn wise. In my CVs I had to wait 100+ turns just for the t3 ideology tenets and completing the influential on all Civs. Game was already over. Player still need to micromanage citizens, buildings and specialists but a lot less than domination. Diplomatic Victory is maybe the most interesting: you can lose the requisite for victory just the turn before the vote.

On my current game (Standard Continents (Planet simulator)/Emperor/Epic I think a 25 % supply cap reduction can be fine. Then the balancing will shift on strategic resources but it's a minor concern. Standard continents give already an overwhelming abundance of them. AFAIK map scripts can be easily twitched.

 

[cerk]

A 25% supply cap reduction would be devastating for tradition. That is around 5-6 units that for a 5 city civ.

 

[Ess85]

I recant my previous report of not enough wars happening/AIs being too pacifistic. Must've been an unlucky game for some reason. In my current game the action is kicking off in Classical with Boudicca getting gangbanged by Napoleon, Sejong and me. First Sejong asked me to declare together, and I accepted only to salvage my scout who was stuck between borders (I ******* hate it when that happens), but then the very next turn Napoleon opportunistically joined in on the fun as well. The Celts are sandwiched between France and Korea, I have a feeling she's in for a really rough time. Well, you know what they say about redheads...

Weird.
I started a new game with the hotfix and not a single war has been declared yet.
I even built a city four tiles from an enemys capital and he doesn't care.
The Zulus and The Huns are in the game but they just don't do anything

 

[Mambo668]

I think we should slightly increase supply units for AI. This is becoming easier and easier each version.
On maps like pangea+ when AI builds some ships it needs now more land troops to be a threat on land.

 

[BiteInTheMark]

I think we should slightly increase supply units for AI. This is becoming easier and easier each version.
On maps like pangea+ when AI builds some ships it needs now more land troops to be a threat on land.

Better to change the evaluation how many ships are necessary. Maybe by the ratio of coastal cities compared to landlocked cities?

Maybe:
(NumCoastalCitiesGlobal / NumTotalCitiesGlobal + NumCoastalCitiesOwn / NumTotalCitiesOwn) / 2 = TargetShipSupplyRatio
If (NumCoastalCitiesOwn * 5 > NumCoastalCitiesGlobal * 4)
then (TargetShipSupplyRatio / 3 )
Capped at maximum with 75% and a minimum of 10%

 

[AndreyK]

I think that instant yields from Great People need some consideration - espesially in later eras (you are able to faithbuy them in Indastrial era at least 3 times in a row - they are not expensive). I think that Great Scientists in particular and may be Writers give to much - like 2-3 tecnologies each and you can get 3 of them. I think that in Ancient, Classical and Medieval Eras impact of Great People was truly great, but in later eras it somewhat dwindled and society as whole and environment had more influence (like public education, more information floating about, more international trade and interchange).
So I think we should consider less instant yields starting from Industrial (or Renaissance) and each era after.

 

[ridjack]

There were some issues with tactical AI aggression in the 10/23 version; the 11/8 version and hotfix were intended to address this.

Feedback seems to be that this version has an issue with diplo AI aggression, and the state of tactical AI aggression is unclear.

Does this version have anything about the AI's tendency to squeeze a new city onto some tiny corner of your civilization where there are like 3 tiles available (none particularly valuable), with no hope whatsoever of military reinforcements being able to reach it? It's super annoying to those of us who are OCD about cluttered maps, but I also think I just made the case on why the AI should consider it bad tactics and even a flat-out liability.

I'd really love to see something happen with this. It gets quite silly.

I do like the AI being smart and opportunistic about when to start wars, but it seems it can lead to some cases where different factors bring about the perfect storm of no one daring to stir things up.

This seems perfectly acceptable to me. Echoes the real world Cold War quite well.

@tu_79 @Recursive So I posted a brain-dropping I had recently about AIs being tuned to 'seek out' masters to vassalize them in the late-game, when they're no longer really capable of getting an actual victory; that idea hasn't gained much traction as far as I can tell, but this thread does have me wondering if something along those lines could be used to reduce the strain.

My thinking is something like... if you have, say eight out of twelve Civs "resigning" from the game, so to speak, and lending what power they have to one of the world powers, you could tune them to reduce military to only what's needed for border defense, vastly simplify their AI so that they're basically only reacting as opposed to proactively working towards a victory they can't possibly attain; that sort of thing, if any of that makes any kind of sense.

 

[BiteInTheMark]

I think that instant yields from Great People need some consideration - espesially in later eras (you are able to faithbuy them in Indastrial era at least 3 times in a row - they are not expensive). I think that Great Scientists in particular and may be Writers give to much - like 2-3 tecnologies each and you can get 3 of them. I think that in Ancient, Classical and Medieval Eras impact of Great People was truly great, but in later eras it somewhat dwindled and society as whole and environment had more influence (like public education, more information floating about, more international trade and interchange).
So I think we should consider less instant yields starting from Industrial (or Renaissance) and each era after.

Yes, Iam trying to tell this since some month, but not much response. In later stages of the game, the pure influence by population (citizen/specialists) diminishes extremly and goes under 30-40% of your total empire yield generation. Mostly true for science and culture.

I would even go that far, that I would remove one of the eras, so that yields by era multiplier gets tuned down a bit from a maximum of 7 to 6.

 

[pineappledan]

I think that in Ancient, Classical and Medieval Eras impact of Great People was truly great, but in later eras it somewhat dwindled and society as whole and environment had more influence (like public education, more information floating about, more international trade and interchange).

Oh good lord, that's a can of worms.
GScientists and GWriters don't scale by era, they scale by your total / output over the last X turns. They also scale via number of Academy tiles or GWs created by your civ. Each Academy you build increases the bulbing of subsequent GScientists by 10%. There's also a Rationalism policy that boosts GScientist's bulbing amounts.

So there's 2 problems:

1. Late game buildings have to give big yields or else they simply aren't worth building. This results in a surge of / per turn in the last few dozen turns, enough to significantly affect the bulbing potential of those last few GPs.
2. Human players know they can work processes, and they know when the next GP is going to be born, so they can amp up their output for 5 turns prior to the next GScientist being born so that the GP isn't bulbing for (Math)*(5 turns of output), but for (Math)*(5 turns of output + 5 turns of 0.25*output)

I don't actually know if this is really a problem though. GScientists' massive bulbing potential is basically the only thing that makes Order a viable ideology.

 

[Earf]

I understand there's been a lot of discussion around the skirmisher unit line lately. I don't see the point of building melee mounted units over ranged mounted units anymore. With 5 movement points on mounted ranged units, you can easily go on the offensive and get some hits in on enemy units/cities. If a mounted ranged unit gets close to harm, unless an enemy is able to take out that unit in one turn, you will easily be able to get away. Something seems off about this. I would expect melee mounted units to be the better option when deciding which type of unit to build because melee is forced to take damage in order to deal damage. Mounted melee units also do not get defensive bonuses. There is so much more risk in using mounted melee units over ranged units. The only reason I see for taking mounted melee units now is to have at least one to take a city. Maybe mounted melee needs to be balanced more in line with mounted ranged units?

 

[AndreyK]

GS just give to much, I am playing in epic speed Japan, so China went Rationalism and I had to follow, after we finished the tree, at first I was behind by 2 policies, then because of GG and GA I used GW so finished first (we were at war then). Then like in 12 turns I got 4 GS and 12 technologies (China got 9 or 10) so whole era went by in a blink of eye.

 

[ridjack]

GS just give to much, I am playing in epic speed Japan, so China went Rationalism and I had to follow, after we finished the tree, at first I was behind by 2 policies, then because of GG and GA I used GW so finished first (we were at war then). Then like in 12 turns I got 4 GS and 12 technologies (China got 9 or 10) so whole era went by in a blink of eye.

That sounds likely to be a problem with scaling to epic speed, to be honest.

 

[Recursive]

Okay, for the next version I've boosted diplo AI aggression.

Tentative Changelog:

Fix some more backstabbing bugs

Fix "you forgave them for spying" modifier not being applied

AI will now always declare war if a demand is refused and they're stronger than the other player both militarily and economically, unless a sanity check prevents it.

AI now considers ideological opponents, untrustworthy friends and anyone who captures their Holy City a major competitor.

Increased war likelihood towards major competitors, decreased it slightly for others for prioritization reasons.

AI more likely to use coop or third party war against major competitors (reverted the check requiring the
biggest competitor to be the target, since there's now a separate check to stop any backstabbing).

Greater AI consideration of military & economic strength, ideology when calculating approach.

If there's been a denouncement in either direction, FRIENDLY and DECEPTIVE approach weights are set to 0
(no more obviously fake friends).

AI now more likely to declare war if they have bonuses towards war (authority, imperialism, autocracy, leader
traits, temporary attack bonus).

AI now more likely to declare war against people they don't like on higher difficulties (uses DifficultyBonusBase
and Opinion); combined with the increase in land dispute penalties below, makes the AI smarter, more dangerous, and more aggressive.

Land disputes will matter more in Ancient/Classical and for AIs with leader bonuses towards war, and will
increase war likelihood (should result in more early wars, less passivity).

Increased no contested borders modifier to counteract this a bit; AIs with no contested borders will gain +15-20
opinion rather than +6 in Ancient/Classical, and +10-15 past Classical - should help with early DoFs and bloc formation.

Overall aggression increase.

 

[Coffee Monopoly]

Amazing! Can we have a hotfix release?

 

[Recursive]

Amazing! Can we have a hotfix release?

You'd have to ask Gazebo.

 

[JamesNinelives]

I understand there's been a lot of discussion around the skirmisher unit line lately. I don't see the point of building melee mounted units over ranged mounted units anymore. With 5 movement points on mounted ranged units, you can easily go on the offensive and get some hits in on enemy units/cities. If a mounted ranged unit gets close to harm, unless an enemy is able to take out that unit in one turn, you will easily be able to get away. Something seems off about this. I would expect melee mounted units to be the better option when deciding which type of unit to build because melee is forced to take damage in order to deal damage. Mounted melee units also do not get defensive bonuses. There is so much more risk in using mounted melee units over ranged units. The only reason I see for taking mounted melee units now is to have at least one to take a city. Maybe mounted melee needs to be balanced more in line with mounted ranged units?

How do you find them in rough terrain?