New Beta Version - August 16th (8/16)

[Kim Dong Un]

Here's what you get when you rush Equality now. This is in King difficulty, and I'm a little behind in techs having 9 cities. My neighbours, in comparison, have 3, 3 and 2 cities.

Spoiler :

View attachment 566495

On turn 112 with 7 cities, I took Equality and went from 52% up to 70%; I'm happy to see it perform.

I later annexed a Capital while experiencing some weariness throughout a lengthy war on multiple fronts, and I had to peace out before vassaling due to dropping under 35% for a couple turns. It felt in a good spot in terms of how much war/conquering I was able to get away with as progress.

 

[Deljade]

Seems to be a consensus that the AI is slow. And apparantly it wasn't obvious in the test games so it's only slow compared to the player.
Was the initial boost so essential for it that it's completely lobotomised without it? I am mostly looking for a somewhat equal religion race and if other boosts are necessary to keep it competent so be it.
Though I would let the games play out a bit to see how things unfold and if the AI puts up a good fight. Especially considering some players, looking at tu, seem way overextended and, even, just settling for the heck of it, to test new Progress policies (should be done ofc)
I'm also wondering if there's a bug at play here, like that growth bug.

Since I spend 1-2 weeks on a game its a tough call to start a new one on fresh patch but I may just run some test games tomorrow..

 

[tothePAIN]

There's something off with the AI policy selection. The AI are universally choosing authority.

I'm playing a deity game with the following civs. Before this patch, I would have predicted them as follows:

1. Ottomans - Tradition
2. Carthage - Progress
3. Polynesia - Progress
4. Indonesia - Progress
5. France - Authority
6. Iroquois - Authority
7. Shoshone - Authority (but should be Progress)


This game, every AI went authority. Every single AI.

 

[Pathologic]

There's something off with the AI policy selection. The AI are universally choosing authority.

I'm also seeing this. All 4 of the AI near me have been full Authority, even with Random Personalities enabled. Austria, Brazil, France, and Japan.

 

[tothePAIN]

Seems to be a consensus that the AI is slow. And apparantly it wasn't obvious in the test games so it's only slow compared to the player.
Was the initial boost so essential for it that it's completely lobotomised without it? I am mostly looking for a somewhat equal religion race and if other boosts are necessary to keep it competent so be it.

It might be necessary to move the AI bonus from founding the capital to when the AI gets a pantheon. But it looks like there are a few bugs that still need to be worked out, such as this policy bug.

Edit: it might be this hotfix - it wasn't this way earlier.

 

[Stalker0]

There are likely a few bugs, but make no mistake, the AI should be easier now, that’s the consequence of removing the capital founding bonuses. Same thing happened last time we did it.

now there may be some behavior issues to explain why the AI isn’t expanding, but it should absolutely be slower

 

[InkAxis]

Seems to be a consensus that the AI is slow. And apparantly it wasn't obvious in the test games so it's only slow compared to the player

Is there consensus? I see people reporting that the AI is expanding slowly, that's likely a different thing. I'm not saying that it's not, but that we should make sure before adjusting anything.

Either way, if we do find out we have to change it I would prefer it just didn't apply to shrines/wonders/WC projects. The system worked fine except for those things where it was bad. So just make it not apply to those.

 

[DaniSciB]

Some more feedback:

The AI seems to be very "okay" with you trying to enact your World Religion, and I mean: They are Grateful. lol.

Some Civs are going with strange opening policies:

Examples.
Maya = Authority
Brazil = Authority
Aztecs = Progress...

Generally speaking, more civs are willing to go Authority now.

Here's what you get when you rush Equality now. This is in King difficulty, and I'm a little behind in techs having 9 cities. My neighbours, in comparison, have 3, 3 and 2 cities.

Spoiler :

View attachment 566495

Yeah, Equality is surprisingly good now.

 

[Gazebo]

There are likely a few bugs, but make no mistake, the AI should be easier now, that’s the consequence of removing the capital founding bonuses. Same thing happened last time we did it.

now there may be some behavior issues to explain why the AI isn’t expanding, but it should absolutely be slower

I wouldn't even call it a bug. It's more like, "there are a million knobs here to tinker with, let's figure out which knobs need to be tinkered with."

 

[azum4roll]

Gandhi will start throwing nukes before everyone else reaches Industrial if you let him (and only him) get bonuses on turn 0.

On map issues, you'll start seeing oasis horses and marsh bananas and jungle sugar. That's intended.

 

[Maxxim69]

I suspect at least part of the reason the AI is not expanding in this version as much as it used to might be due to recent changes to civilian unit logic by @ilteroi. The AI now seems more reckless with its settlers and workers. In general I don’t tend to see many civilians captured by barbs, but in my first 8/16 game I saw a worker and a settler in the first several dozen turns.

 

[JamesNinelives]

I suspect at least part of the reason the AI is not expanding in this version as much as it used to might be due to recent changes to civilian unit logic by @ilteroi. The AI now seems more reckless with its settlers and workers. In general I don’t tend to see many civilians captured by barbs, but in my first 8/16 game I saw a worker and a settler in the first several dozen turns.

I'm happy to hear this. The AI loosing their initial settler (or their 2nd settler) to barbs because they weren't escorting properly did feel bad.

On map issues, you'll start seeing oasis horses and marsh bananas and jungle sugar. That's intended.

Jungle sugar already existed. I'm not sure how I feel about oasis horses though? That seems kinda OP. Marsh bananas are fine I guess lol.

 

[tu_79]

reminder: if you get deals that are way outside the norm, please share with me on GitHub. Thanks!

I don't consider this a bug. It was a completely reasonable deal.

 

[azum4roll]

I'm happy to hear this. The AI loosing their initial settler (or their 2nd settler) to barbs because they weren't escorting properly did feel bad.


Jungle sugar already existed. I'm not sure how I feel about oasis horses though? That seems kinda OP. Marsh bananas are fine I guess lol.

Horses AROUND oasis (basically fresh water desert tiles), not ON oasis. So no, they're not OP at all.

 

[DeAnno]

I wonder if giving the AI bonus settlers at various timings on higher difficulties might work smoother than bonus production. It would help prevent their capitals getting stuck at low population too. OG Deity got a bonus settler right? We definitely wouldn't want to give them at the start, but maybe X turns in?

 

[LifeOfBrian]

Can anyone please tell me how to change the modfiles (or alternatively if someone could post the files if they have changed them for themselves) so the AI would again get all the bonuses from settling etc. that it used to got before the latest changes? This version's AI is way too slow at the start compared to before. Thank you!

 

[JamesNinelives]

I just played a game up to Turn 170 on Epic speed. I'm playing one difficulty higher than the previous version and the AI is still extremely slow at founding religions. I founded about 20 turns later than usual and there were still 2 (of 7) religions available - normally there would be none. They weren't doing too bad in other areas (most civs were ahead of me as they should be given my poor start), but somehow their religious game seem very uncompetitive.

Spoiler Religion :

Horses AROUND oasis (basically fresh water desert tiles), not ON oasis. So no, they're not OP at all.

OK, there's no need to shout. I like that resources turn up in different areas so far, makes snow tiles less terrible when they have a chance of spawning stone. And it's nice to find resources on marshes sometimes.

Why are resources so concentrated in certain areas though? There were 8 resources directly adjacent to each other in one case (1 perfume, 1 iron, 3 bison, 2 stone, 2 fish). I've never seen that before and all my settings are the same. It feels unbalanced.

 

[azum4roll]

I assume you mean bonus resources. There are a few possibilities.

The script places bonus resources in several ways.
1. It adds random bonus resources to food-poor start locations and stone to production-poor start locations, in order to balance each start location.
2. It adds one bonus resource (deer, banana, wheat, sheep, cattle) on the 3rd ring of each start location depending on region type. So if you start in a jungle region it adds a banana 3 tiles from your start.
3. It adds even more bonus resources to hills regions, randomly scattered, type depending on terrain (deer, banana, wheat, sheep, cattle).
4. It places fish randomly on coast tiles.
5. It randomly scatters the majority of bonus resources (minus fish) on land tiles, with different frequency for different plot types/terrains/features.

Steps 3-5 introduce an impact radius for each placement so that no more bonus resource can be placed within the radius, mostly ranging 0-3 (again, randomly) so there could still be adjacent bonus resources.

Communitu_79a changes the place fish function to do it according to nearby fertility instead. For each small island, there's also a chance that a stone tile is added to it.

Removing all 0 impact radius can (mostly) force no bonus resource adjacency, but that feels unnatural and... forced.

 

[Zuizgond]

Played a 12 civs pangea game and half the AI took a very long time before founding their second city. They built their first settler normally but then it stayed near the capital for ages. I suspect they are overly cautious when there is a barb camp nearby.

 

[JamesNinelives]

1. It adds random bonus resources to food-poor start locations and stone to production-poor start locations, in order to balance each start location.
2. It adds one bonus resource (deer, banana, wheat, sheep, cattle) on the 3rd ring of each start location depending on region type. So if you start in a jungle region it adds a banana 3 tiles from your start.
3. It adds even more bonus resources to hills regions, randomly scattered, type depending on terrain (deer, banana, wheat, sheep, cattle).

It wasn't next to a capital, a food-poor location, or a hills region - it was flat grassland. You can see the edge of it at the top of the screenshot I posted, around Nanjing - 1 perfume, 1 iron, 3 bison, 2 stone, 2 fish. I suppose it's possible that China's settler started near there. IDK. It's just one example, it might be a fluke. The capital locations all seem fine, not much different from before. I am seeing more bonus resources than usual though. How many 'start locations' are there per map?

A couple more early observations: pantheon choice seems more diverse than it was in the previous patch which is nice. Haven't met any civ that lost their initial settler to barbs yet so hopefully the more cautious AI is helping with that at least.

Edit: Authority is popular, but it's not literally everyone: of eight civs met so far, 5 are authority, 2 took progress, 1 has tradition. Fewer tradition civs might be part of the reason religions are taking longer to found.

Edit 2: I'm also seeing different events from usual which is nice. I wonder how the AI is handling them though.