Real Strategy (AI) 2.3.1

[V. Soma]

I am happy to report that AI now more agressively do conquest!

I gave the new patch and RST a go with my saved file of turn 1
with the tiny map and Hungary, Mongols, Scíthia, (and Maori human)
DV only, no other victory allowed...

And Hungary DID domination victory!

see:


Some changes in AI build:
- Encampment in literally every city
- NO campus, only a few at the end
- entertainmant the same amount as before (OK)
- lots of IZ

 

[alexu4326]

Is ok to put the new patch wihit this version of mod or i shall wait to a update?

 

[Infixo]

I am happy to report that AI now more agressively do conquest!

I gave the new patch and RST a go with my saved file of turn 1
with the tiny map and Hungary, Mongols, Scíthia, (and Maori human)
DV only, no other victory allowed...

And Hungary DID domination victory!

see:
View attachment 521861

Some changes in AI build:
- Encampment in literally every city
- NO campus, only a few at the end
- entertainmant the same amount as before (OK)
- lots of IZ

Some Civs received AggressiveYieldsPreference, and it usually adds to the same preferences that RST uses. So, the effect is multiplied sort of.
Also, Great Generals and Admirals are boosted by crazy 50% and that would make an AI build an encampment in virtually every city.
This is not „balanced”, it’s overkill.

 

[Infixo]

Is ok to put the new patch wihit this version of mod or i shall wait to a update?

It would work. There would be few imbalances (e.g. see the post above).
I will update the mod around the weekend.

 

[aikido_panda]

Some Civs received AggressiveYieldsPreference, and it usually adds to the same preferences that RST uses. So, the effect is multiplied sort of.
Also, Great Generals and Admirals are boosted by crazy 50% and that would make an AI build an encampment in virtually every city.
This is not „balanced”, it’s overkill.

Totally agree. RST with AI focusing on campuses and theatres at least makes scientific/cultural advancement not that easy. Moreover, encampment boost can make AI wars even more defensive, cause it actually doubles the number of fortresses with ranged attack, which AI is still unable to besiege properly...

 

[Infixo]

Infixo updated Real Strategy with a new update entry:

Version 1.4 April patch

Version 1.4
- Updated for April 2019 Patch: flavors, integrated game's Diplomatic victory, updates for new traits added to civs.
- Added: Few missing start biases for Civs with abilities dependent on geo location.

Read the rest of this update entry...

 

[V. Soma]

Running an AI autoplay test (RST 1.4)
I see Hungary doing an attack on a city, with five units... fine...
...those 5 units are ALL catapults... uh, not so fine...

EDIT:
and science is still low priority: ENC - and the COM and IZ are first
(I play the only DV allowed game test)

 

[Infixo]

Running an AI autoplay test (RST 1.4)
I see Hungary doing an attack on a city, with five units... fine...
...those 5 units are ALL catapults... uh, not so fine...

EDIT:
and science is still low priority: ENC - and the COM and IZ are first
(I play the only DV allowed game test)

Yeah, I also noticed that AI are building a bit more siege weapons than usually.
I think this might be a result of the April Patch changes.
As I wrote in an earlier post:
AiTypeDependence - new field in OpTeamRequirements as for now City Attack Force needs UNITTYPE_SIEGE
It’s hard to say how exactly this field works. They added this in a line with UnitType_Siege_Support.

Also, my second remark.
Also, Great Generals and Admirals are boosted by crazy 50% and that would make an AI build an encampment in virtually every city. This is not „balanced”, it’s overkill.
I decreased this params to some reasonable values but when a civ is aggressive and (usually) goes for conquest, they still are quite strong. I don’t know however why only Campuses ere neglected.

 

[Infixo]

@V. Soma Anyway, in 1.4 i have not changed anything in team compositions, so anything which seems weird must be related to the patch and we have to somehow address it.

 

[V. Soma]

Just for tests I started a only SciV test game, and the AI duly builds campuses at start...

 

[TYPEZEROdk]

@Infixo here is the log lua we talked about earlier today on Steam

 

[aikido_panda]

I have found two more issues, don't know whether they may be fixed by modding, but just in case.
1. AI never founds any religion, when he buids Stonehenge. It's certainly a bug, but I don't know - it's probably initial problem of firaxis.
2. AI often offers huge amounts of money in diplomatic deals, especially on high difficulty levels. Seems like an additional exploit for player

 

[Infixo]

Ad. 1. They do found if they have a Holy Site. Probably Prophet cannot recognize Stonehenge as a valid place. Firaxis.
Ad. 2. Have you not seen dozens of posts complaining on broken deals? Firaxis.

 

[Infixo]

Question to all.
Do you have any observations regarding Religious Victory? How would you compare AI behavior in this area to its behavior in the vanilla GS game?

 

[Dëzaël]

Not really able to compare with vanilla since I have only one vanilla GS game under my belt, but what I see with your mod on is consistent with pre RF&GS vanilla Civ. AIs seem to often stall at some point in their religious victory pursuit. Some kind of religious frontline is drawn after some time and it becomes trench warfare. Always the same fights in the same places. You said it is not unusual for an AI to win religious victory on small maps, I play only huge maps. I'm guessing with more civs trying to achieve this victory, combined with religious pressure mechanics and more land to cover, the odds for a stalemale to occur are higher. I recon the religious system doesn't give enough tools for the AI to reliably be able to take the upper hand. Lack of ways to support the fight in far away land due to unit production centers being at home maybe?

Anyway what I see is the AI trying the same thing over and over in a loop without consistent results, and the religious victory is consistently the one lagging behind in my games, progress wise. Only significant shifts I saw were when one or more of the religious civs got weakened by war, and that's more often than not by random expansionism rather than by their religious competitors.

2. AI often offers huge amounts of money in diplomatic deals, especially on high difficulty levels. Seems like an additional exploit for player

Ad. 2. Have you not seen dozens of posts complaining on broken deals? Firaxis.

At least they are consistent with their needs. Civs needing amenities give more for luxuries, culture civs give more for GW, etc... They need to tone it down now...

 

[Infixo]

@Dëzaël Yup, what you describe is pretty much vanilla behavior. And I concur with the diagnosis. When 2 civs with strong religions meet and high faith output, they are just bleeding out. They send waves after waves to each other but doesn’t change much. Just grind. Plus Inquisitors can revert their efforts instantly. There are no tools that would help globally, that is a good observation.

 

[V. Soma]

I played a game with new FXS patch and RST and other mods - and I got beaten by AI on smoother modded Immortal.
Small pangea, all victorí types - USA AI had a BIG army and I got attacked in classical age...
AI was effective... I go back to Emperor

 

[Infixo]

USA AI had a BIG army and I got attacked in classical age

Did they use siege units? Rams, catapults?

 

[V. Soma]

Did they use siege units? Rams, catapults?

No, but I use mod where spears are ranged, and AI used a lot to siege... then captured with sword...

 

[Melkus]

Can also report progress - I got into surprisingly serious trouble in my last smoother immortal game with Cleopatra (no other gameplay mods, standard fractal) First the Arabians wiped out my religion with a steady flow of missionaries. (never happened to me before) Then got completely boxed in after my fourth city, so had to take my chances to expand, took a city the Mapuche had taken from the Inca earlier and brought a 4 civ emergency on myself. (An earlier emergency against the Mapuche, that might have tipped the game in my favor, failed by one vote)

All four AIs contributed at least some coursers and knights (unfortunately France just went through my territory to get to the city and lost their units easily to city fire). But the Mapuche attacked relentlessly, in several waves and with diverse units (catapult fire, swords, knights), while my guys got stretched on other fronts. I held on for about 20 turns, while my Maryannus / XBows gradually lost relevance as knights flooded the field. The Mapuche then teched to their - at this point - almost untouchable Malon Raiders, built them and finally took the city (which had walls and a hastily chopped encampment) back. First forced peace for me after the classical area, as far as I remember. Also the first time I had a genuine reason to put Victor in a city. No golden age btw (so no +10 combat for the Mapuche), and certainly can't afford to get one now with the Malons sitting at my border.

Saw the usual tactical flaws, of course, the AIs lost a lot of units through stupid movements and failed to pillage (my little empire would have collapsed quickly). Still nice to see some serious counterpressure. Just shows the potential of the game - one narrow diplomatic decision can make or break a game when the AI is at least somewhat threatening. That's the kind of chaotic, unpredictable history I hoped to see...