Leader Personalities

[Thalassicus]

I've seen that problem too, but there's nothing we can do about it with the tools available to us. I think the main problem is the AIs are somehow forced to build cities as close as possible -- typically with 3 tiles of spacing between cities. They will break this rule only in very rare circumstances. I think the "stuck at 1 city" situations are when they have no good city spots with a 3-tile spacing from existing cities.

The only thing I've found which significantly affects this is how much they value tiles already containing their own culture. I fiddled with some settings to devalue self-culture within the first 2 rings around a city site, which is why they place cities 3 tiles apart (1 tile of no-culture between). The modding community had a big investigation into AI city placement a few months ago. No one conclusively found any xml variables that control AIs willingness to place cities at further distances to grab specific resources.

The code to control this is most likely contained within the c++, which we don't have access to.

 

[SlightlyMad]

That's strange though, because I'll still occasionally see an AI throw down their second city halfway across the map when they had to walk past a few better spots to even get there. I figured it was a vanilla problem though; hopefully in the next next next next patch ...

 

[Thalassicus]

Has anyone seen an AI conquer a citystate since the June patch? I've tried setting the AI's minor civ conquest values to all sorts of values, yet ever since the June patch, they appear to never even consider it. I'm starting to think this is the main reason the "Conqueror" personality AIs aren't doing well lately.

 

[albie_123]

I haven't at all, in fact. Played two games so far, and the usual suspects, eg. Monte, Isabella, and - most importantly - Genghis, haven't even tried.

The AI seems to be getting worse at embarkation as well, in that, even though it would be infinitely more tactical to just walk around a lake, they send their units straight across it, letting me pick them off with archers and seige units easily.

 

[Thalassicus]

Well that's frustrating. I got those personalities working after lots of trial and error... only to have it blocked somehow by the game core only Firaxis has access to. I've been doing lots of experimentation tonight. I can't get the AI to attack a CS anymore, even if they have a massive army on the CS's doorstep, and are already at war with the ally of the CS! It seems to completely ignore that strategic option now.

 

[Txurce]

Has anyone seen an AI conquer a citystate since the June patch? I've tried setting the AI's minor civ conquest values to all sorts of values, yet ever since the June patch, they appear to never even consider it. I'm starting to think this is the main reason the "Conqueror" personality AIs aren't doing well lately.

Yes, I have. It's the usual suspects, although in my last game an unusually aggressive Iroquois attacked a CS very early in the game.

 

[Jorlem]

I saw a city state conquer another city state once, a week or two ago. That was pretty neat.

 

[Sneaks]

For the first time ever, I saw a CS take a city from Genghis, then lose its capital. Now Brussels has a capital of Old Sarai.

 

[Txurce]

For the first time ever, I saw a CS take a city from Genghis, then lose its capital. Now Brussels has a capital of Old Sarai.

The game is definitely getting weird.

 

[Zaldron]

Has anyone seen an AI conquer a citystate since the June patch? I've tried setting the AI's minor civ conquest values to all sorts of values, yet ever since the June patch, they appear to never even consider it. I'm starting to think this is the main reason the "Conqueror" personality AIs aren't doing well lately.

I have absolutely seen this. Both Greece and Egypt took at least one CS from me in a recent game just before the Aug patch.

 

[Thalassicus]

I've been thinking about the AI's initial rush. What I'd really like is to get the AI to attack nearby citystates instead of major civs. With the dozen-or-so units they usually rush with, it'd be trival to conquer a citystate... but they always go for the stronger major civ instead! I spent ten hours or so running through permutations of the variables yesterday, but couldn't seem to find a combination that could reliably get this result in my test games.

What was more frustrating, the AI never attacked any citystates in the first 150 turns of the autoplay games... not sure why.

 

[Txurce]

What was more frustrating, the AI never attacked any citystates in the first 150 turns of the autoplay games... not sure why.

Because they've been hardcoded not to, in order to sustain the CS aspect of the game?

 

[Thalassicus]

I'm not really sure. They used to attack CSs in autoplays. From what yall have said, they still do so, and I just didn't see it in the games I've tried... for unknown reasons.

 

[Txurce]

I'm not really sure. They used to attack CSs in autoplays. From what yall have said, they still do so, and I just didn't see it in the games I've tried... for unknown reasons.

To be clear, AI definitely attack CS in my games - but I wasn't certain about the first 150 turns. Now that I think about it, the Iroquois of all people did it recently very early in the game.

 

[Txurce]

I played two standard tall games aiming at a Science win, using v9.6. I stuck to the same game plan as much as possible: going for Tradition and then Rationalism. Because of the bugs with Angkor Wat and the Hagia Sophia, I focused less on Patronage, and both civs picked up multiple free cultural buildings. Sticking to 4 cities led both civs to sell off cities captured in tech-fueled counter-attacks. Both benefited financially from this. Korea fought much earlier, and thus had more reparations earlier. In both cases, the focus was on cranking out specialists, although Korea didn't go as heavily for GM's and GA's as it could have.

War timing aside, here were the main differences:

Korea started with three luxuries including marble, in a strong hammer location. It built the GL, NC, Pyramids and HG early on, and never looked back. Later it built the Porcelain Tower. It only had moderate DoF science contributions throughout.

Babylon started in a food-rich location with two luxuries, missed on the GL and Temple of Artemis, but got the Pyramids. It built a frontier city to stake out borders, and slowed down its NC build until it had three cities. It also missed on the Porcelain Tower, but had more DoF's.

Babylon had a higher population, Korea more hammers.

Babylon played second, and seemed to be doing about as well as Korea's specialist tsunami, thanks to the number of GS's it cranks out. But given Korea's legendary start, I was still a bit stunned to find them both finishing dead even at 285 turns.

I will test them again with the presumably bug-free v9.9, but so far would conclude that there is no need to nerf Korea any more than Babylon.

Finally, a relevant side note: in two games playing against Korea, Sejong had his head handed to him right away by the other AI.

 

[Thalassicus]

To clarify a bit, I don't want to nerf Korea. I feel the trait outshadows the turtle ship, and want to shift power from the trait to ship to balance importance of the two.

 

[Txurce]

To clarify a bit, I don't want to nerf Korea. I feel the trait outshadows the turtle ship, and want to shift power from the trait to ship to balance importance of the two.

That's good to know, in terms of focus. I would argue that nerfing the trait makes Korea less fun, because like Babylon and Germany, it's a trait-heavy civ. Speaking more generally, I think you lose more than you gain by "evening" the three traits. As discussed in the Revamping Leaders thread, synergizing traits makes for more enjoyable game play, in my opinion. And right now, Hermit Kingdom synergy is definitely there.

For me the issue with the Turtle Ship isn't its strength - there are lots of underwhelming UU's - but its irrelevance. The Turtle Ship's value could be erased with one more tech, when Korea could build a Ship-of-the-Line in VEM. I would rather address that, since I think the notion of it is historically interesting.

The first thing I would do is move the Frigate out of Navigation. That's too close of an upgrade for a Caravel. There are no other adjacent promotions in the entire game. Put it in Archaelogy instead. Based on this change, we could either weaken the UU, and allow it to upgrade to Frigate with Archaelogy; or leave it as is, not upgrading until Combustion, using the Ship of the Line to explore.

This entire approach may or may not appeal to anyone besides me. What I would like to stress is my sense that the Turtle Ship is synergistic and balanced within the civ, but is unsatisfying due to its technological irrelevance in VEM. Addressing that while maintaining a rough balance with the existing civ - leaving the UA as is - should be the ideal.

 

[Thalassicus]

I don't want trait-heavy civs at all. One of my long-term goals has been for the 3 characteristics of each leader to be approximately equal in importance. There's just so much stuff to do I haven't had time to adjust every leader.

If we look at the light ships:

• Ancient - Trireme
• Classical -
• Medieval -
• Renaissance - Caravel and Frigate
• Industrial - Destroyer

I've been thinking about moving the Caravel to Compass to spread things out a bit. If we move the Frigate to another Renaissance tech they're still in the same era.

 

[Txurce]

1. I don't want trait-heavy civs at all. One of my long-term goals has been for the 3 characteristics of each leader to be approximately equal in importance. There's just so much stuff to do I haven't had time to adjust every leader.

2. I've been thinking about moving the Caravel to Compass to spread things out a bit. If we move the Frigate to another Renaissance tech they're still in the same era.

1. I disagree, but for totally subjective reasons, so I won't argue this (and will help work on it). What surprised me was the focus on a day-old civ (Korea) when there have been prime candidates available for revision for about a year!

2. There's no perfect place for Caravels, but moving it to Compass makes the most sense. I think it not only spreads things out, but makes for a more interesting game. This will buff Siam and Greece as well as England, but that can be taken into account.

 

[Txurce]

I am late in another Korea game, this time with 10.0. I got off to a slow start due to being rushed by China and Songhai at about 50 turns in, after having thought 10 turns earlier that the AI sure start out loaded on Emperor these days. I barely survived with two archers and walls in my capital. (Usually the archers are key. This time it was the walls.) The upshot is that I'm now well ahead of my pace in my last Korean game despite a much slower start and no deep-water ships, thanks to plenty of DoF's.

My feelings as I play them are that 1) I enjoyed the uniqueness of not having intercontinental contact, and 2) it's not that bad, since both the AI and a lot of CS eventually make contact. My point here is that no matter how you balance Korea, there is something special about the deep-water handicap that would be perversely nice to keep.