New Beta Version - January 14th (1-14)

[ElliotS]

Sorry, but this sounds more like a buff to warmonger play. More units to kill = more yields by authority or Abilities.

If you really think there is no amount of units or unit production that can make war impossible, you're crazy. If we're agreed that there is some amount that makes war impossible, it should be a logical conclusion that there is a balanced middle ground where war is difficult, but possible.

Obviously there are other factors like free XP and bonus XP, as well as more general bonuses that can make AIs harder to deal with.

 

[intelli]

I play King normally, but just tried a game on Prince where I took Carthage/progress and had marble with a nice jungle coastline for what I had hoped to be rapid expansion and a walk in the park. Instead I had Russia with authority somehow jam six cities down my throat from above along with a DoW and lines of enemies by turn 90 (don't remember the exact turn, but it was before turn 100 which is absurd for Prince)

Why take Progress when all your benefits are from city settling? Get Pottery, build one settler, beeline the left side of Authority for the free settler, then run for Pyramids so your capital gets to keep growing, once you've got monument/shrine in a city which shouldn't take more than a handful of turns with Carthage's gold bonuses move to building settlers there too. I'm playing a game right now and managed to get 12 cities out before I had to move to warmongering to expand further. Once you've taken at least the left side of Authority you can start taking Progress policies with a view to eventually maxing them both out, which ones you take first may depend on how many people you've annoyed with your rampant settling.

 

[Txurce]

I posted this observation around version 1-3 but it is still there, so I will repeat. Around turn 150 many AI cities have already built all possible buildings. Basically AI is able to build buildings as fast as techs are discovered. Along with keeping an army of the maximum size (up to supply limit). As a result, many cities just run processes. I’ve seen many arts and science.
Beside obvious effect of snowballing potential, this adds to happiness problems. AI with complete infrastructure pulls the median up, and humans simply may not keep up.

This would also help to explain a recent surge in AI culture. But the main point is that, while it's okay for the AI to build everything (well, almost everything) on higher levels, it shouldn't have such a negative effect on human happiness. There's nothing you can do about it.

 

[Moi Magnus]

Why take Progress when all your benefits are from city settling?

Authority help you to have a lot of cities, Progress gives you bonuses if you have a lot of cities.
If you are able to settle/capture cities without the help of authority, then progress is probably better because it gives you tons of yields in the long run. So it is not unreasonable to choose progress.

 

[tu_79]

Have you considered lowering AI handicap bonuses? They are there to be played with.

 

[Moi Magnus]

Have you considered lowering AI handicap bonuses? They are there to be played with.

True. If AI improvement now lead to the AI having too much production, then we can maybe reduce their bonuses to production.

 

[Infixo]

This would also help to explain a recent surge in AI culture. But the main point is that, while it's okay for the AI to build everything (well, almost everything) on higher levels, it shouldn't have such a negative effect on human happiness. There's nothing you can do about it.

This can be balance-tweaked maybe? There is a parameter that says which percentile is this "median" (I think it's 40 now, so it is not an actual median). But I don't know if this algorithm is entirely symmetrical i.e. if there are any bonuses for AI.

Even so. I don't think hapiness is the actual problem. I am absolutely not convinced that AI should be able to build all buildings. This makes the game predictable, all fancy algorithms for strategy, build orders, etc. are almost for nothing, it is come-back to "catch me if you can" style of vanilla civ. Most advanced cities - well, probably yes. But not like most of them. I think the normal situation should be for AI to struggle a bit with infrastructure, to really choose.

 

[Infixo]

Have you considered lowering AI handicap bonuses? They are there to be played with.

It is not handicaps, imho. Turn 150-ish, they are not so big yet.

 

[tu_79]

This can be balance-tweaked maybe? There is a parameter that says which percentile is this "median" (I think it's 40 now, so it is not an actual median). But I don't know if this algorithm is entirely symmetrical i.e. if there are any bonuses for AI.

Even so. I don't think hapiness is the actual problem. I am absolutely not convinced that AI should be able to build all buildings. This makes the game predictable, all fancy algorithms for strategy, build orders, etc. are almost for nothing, it is come-back to "catch me if you can" style of vanilla civ. Most advanced cities - well, probably yes. But not like most of them. I think the normal situation should be for AI to struggle a bit with infrastructure, to really choose.

Again, why can AI build almost everything, but the players are not able to? Too many bonuses to AI?

 

[Txurce]

This can be balance-tweaked maybe? There is a parameter that says which percentile is this "median" (I think it's 40 now, so it is not an actual median). But I don't know if this algorithm is entirely symmetrical i.e. if there are any bonuses for AI.

Even so. I don't think hapiness is the actual problem. I am absolutely not convinced that AI should be able to build all buildings. This makes the game predictable, all fancy algorithms for strategy, build orders, etc. are almost for nothing, it is come-back to "catch me if you can" style of vanilla civ. Most advanced cities - well, probably yes. But not like most of them. I think the normal situation should be for AI to struggle a bit with infrastructure, to really choose.

You're right.

What do you think led to this situation, which wasn't as apparent before?

 

[Moi Magnus]

Again, why can AI build almost everything, but the players are not able to? Too many bonuses to AI?

Without counting the instant yields from the bonuses, the Deity AI has a discount of 20% to (almost) every cost, including population growth.

 

[Gazebo]

There’s a lot of misinformation here.

First, Prince AI has very few bonuses. If you’re getting stomped by them, drop difficulty further.

Second, the AI’s production bonuses are much less potent early game than late overall. If the AI is doing well by turn 150 it is because of our AI work, not cheatz or handicaps.

Third, I’m keeping an eye on median happiness. I will reiterate though that being beaten by the AI at difficulties where you used to win is the point of VP- the AI is not cheating more, it is simply playing better. In fact AI bonuses are lower now than ever before on an overall evaluation.

G

 

[Infixo]

Again, why can AI build almost everything, but the players are not able to? Too many bonuses to AI?

Well, probably...? But don't ask me which ones, they all add up.
Just out of curiosity I've just checked my current test game. Emperor, turn 220, Polynesia. They've just researched Acoustics, so not counting Musicians' Guild and Opera House, their 9 main cities (out of 13) have all buildings but 1 (in 4 cases missing Garden or Windmill). 3 of them run Science process. 4 remaining cities are settled after the 9 and are much less developed.

 

[Moi Magnus]

Second, the AI’s production bonuses are much less potent early game than late overall. If the AI is doing well by turn 150 it is because of our AI work, not cheatz or handicaps.

To be sure that I'm not the one making disinformation, here is the building cost I've understand from the difficulty.xml :

Settler : Base 100% ; Per Era -0% so 100/100/100/100/100/100/100/100
Prince : Base 90% ; Per Era -6% so 90/85/80/74/68/63/58/53
Deity : Base 70% ; Per Era -10% so 70/63/56/49/42/35/28/21

Formula used : Base * (100+(Per Era*(Era-1))) / 100
(I first though it was "Base+(Per Era*(Era-1))" but it gives 0 for Deity in end game, so it was false)

 

[Gazebo]

To be sure that I'm not the one making disinformation, here is the building cost I've understand from the difficulty.xml :

Settler : Base 100% ; Per Era -0% so 100/100/100/100/100/100/100/100
Prince : Base 90% ; Per Era -6% so 90/85/80/74/68/63/58/53
Deity : Base 70% ; Per Era -10% so 70/63/56/49/42/35/28/21

Formula used : Base * (100+(Per Era*(Era-1))) / 100
(I first though it was "Base+(Per Era*(Era-1))" but it gives 0 for Deity in end game, so it was false)

Yep. But that’s not for wonders or National wonders, and the AI has unhappiness/supply penalties like anyone else.

Edit: and compare them to vanilla.

G

 

[tu_79]

Ok, ok. AI is playing better. Do we want the game to be harder? If not, then AI bonuses have to cool down a bit.

The question more likely is: Are deity players bored so they need further challenge? Because I'm still struggling with Emperor. Just when I thought I could move up...

 

[ElliotS]

Ok, ok. AI is playing better. Do we want the game to be harder? If not, then AI bonuses have to cool down a bit.

The question more likely is: Are deity players bored so they need further challenge? Because I'm still struggling with Emperor. Just when I thought I could move up...

Literally the only reason to make it easier would be to give an arbitrary ego boost to people over being able to 'beat' a 'harder' difficulty.

Deity seems correctly hard at my current handicap values currently, lowering it would massively reduce my enjoyment.

If anything we should consider moving all difficulties back one and adding a new deity for more challenge. (So current Deity is new Immortal.)

 

[Txurce]

Ok, ok. AI is playing better. Do we want the game to be harder? If not, then AI bonuses have to cool down a bit.

The question more likely is: Are deity players bored so they need further challenge? Because I'm still struggling with Emperor. Just when I thought I could move up...

I'm fine moving up and down and up (done all that in the last year) as long as my adjustments yield results. When I become chronically unhappy for no reason other than relativity, I don't think "I have to up my game." I think the game is imbalanced -- which of course is going to happen here and there, given how it changes every week.

 

[Infixo]

Deity seems correctly hard at my current handicap values currently, lowering it would massively reduce my enjoyment.)

And what are they?

 

[ElliotS]

And what are they?

Something like 100 70 100. Let me check when I get home.