I hope they fix the Ai city placement in BNW.

H

Hakuoh

Warlord
Joined
Apr 1, 2013
Messages
115
We all know this right?
Ai building Citys on snow surrounded by ice and maybe 1 fish near by..
I know they never have happiness problems but i find it kinda odd.

Even every 1 tile iseland on the ocean is settled with no resource.
Its like no place on the map i save for a new city even if it does nothing than 2-3 pop.
 
Assyria is gonna love those crappy city though ^^

I agree it's strange and slightly enerving, but I prefer that to the pre-patch weak expansion.
 
still would not hurt to place some directives like (don´t settle if 80% ice+trunda unles civ lack resource X, Y or Z and its available there).

half a dozen directives like that would fix the most horrible cases.
 
I agree. Some civs (taking Egypt from my game for ex) settled on a 1 hex tile. I captured it and it was making a monument in 167 turns...

Oh yeah, and it was way down south near the ice
 
I would really like to see the coding tweaked so that building lots of crappy cities costs you more than it is worth (so more like Civ 4) but I have no idea how realistic or large a project that would be.

Between G&K's release and the fall patch the AI went from being incredibly reluctant to expand to pumping out cities on every spare bit of land it could find. I'd like them to bring this back to the middle ground, or at least keep the expansiveness but tell the AI to stop when it runs out of decent places!)
 
I place my cities on 1 tile snowy isles with no resources. So long as you do not go for culture victory and you have enough happiness (which you do with religion), I do not see why it is a bad idea to have more cities. However, happiness does stop spread, when you do not have the right beliefs.

In my opinion the problem is infinite happiness itself, if the AI was smarter it would not need these extreme boosts to happiness, gold and what not. If it needed resources, it would start placing cities in (more?) reasonable locations. What I mean is... make AI smarter, reduce its boosted resources, problem should fix itself. I suppose telling AI not to settle on these 1 tile isles would contribute to making it smarter... Just improve the god damned AI :mad:!
 
I'd agree if the AI was settling these spots in preference to decent locations, as it would both hamper the AI and give you few useful conquests, however in my experience post-G&K this doesn't happen. The AI won't always assign an appropriately high priority to the best spots, but it's actually fairly rare for me to look at the map, think "a city would be good there", and have the AI not settle in the exact spot or next to it if I don't get there first.
 
The main problem is the AI's the calculation does not apply a discount for tiles within 3 hexes of another city that aren't yet within the other cities cultural border combined with only a 50% discount of the value of the tiles within another city (or even another civs) cultural border.

I have noticed that if you go to global defines and increase it by one, that the AI's placement greatly improves. It does however carry with a major downside on the large island script; because of how bad the city state founding logic is.
(On "Tiny Island" such a change would have little affect as it doesn't apply to cities on nearby landmasses and there doesn't appear to be a way in XML file to set one. The "earth" maps have trouble as is; but this increase distance mod should be okay on the other map types.)
 
I certainly hope they fix AI city placement. This is the worst it has ever been for me:
Click to expand...

This kind of city is not really bad if you are willing to spend some gold and later pay with the gold they get from the sea.

Don´t know now in bravenew world.... surely this kind of city now will need something else. Be a mechanic, a building or resource.
 
sterrius said:
still would not hurt to place some directives like (don´t settle if 80% ice+trunda unles civ lack resource X, Y or Z and its available there).

half a dozen directives like that would fix the most horrible cases.
Click to expand...

Completely agree. It is so incredibly stupid and wipes out what little suspension of disbelief one may have going, when they do that crap. Needs fixed.
 
I remember when I played Soviet Union in Civ III and made like 30 iceball cities in Siberia just to stop barbarians from spawning and to have nice-looking borders. They were all size 1-3 and named 'Snow-filler #xx' in Finnish. Good times... :D

OT, this is just one of the things that needs to be fixed with the AI and pretty minor in the scheme of things imo. If they fixed everything with the AI and made it present any kind of tactical challenge in combat, I would pay 200 € for the expansion, no lies. There's never been a Civ game with a decent AI and I doubt I'll see it in my lifetime unless those scienc-y folks are right about the technological singularity.
 
a Civ game with a decent AI and I doubt I'll see it in my lifetime unless those scienc-y folks are right about the technological singularity.
Click to expand...

Well. To be fair the AI evolved a lot in the last 15 years of gaming (civ included) and is going to keep evolving as long they bring more power to the PC.

but more important, a AI really need is time and people that can implement strategy directives to the AI so she can discern all possibilities. Its not impossible, just so time consuming the window to release the game will get years before you can finish it.

A interesting idea for civ6 would be to use more or less the same engine (its still a very beautiful game and if you go from ground 0 you can improve more), that would make possible to use civ5 AI and work from there, instead of working from ground 0.

Unless the engine have limitations.

Another option some games use is to Hide some directives to give the idea that they are learning. (example: AI will use option 1 to 10, but keeps losing using those in the late game part, so he change to strategys 15 to 20 that focus on lategame), but this kind of idea also needs time.
 
I'm pretty tired of the AI deciding that the best possible location for a second city is 20+ tiles away from their capital, behind two other civs. It doesn't happen every game, but it happens enough to be a nuisance.

There doesn't seem to be much reason behind it, either; the Aztecs once settled their second city 25-30 tiles away from their capital to grab one tile of citrus. In my last game, India just went behind my (Greece's) territory and into Celtic lands, just to settle their second city next to one tile of dyes. In each case, non-duplicate luxuries were available in good city locations much closer to home. I understand REXing and map denial, but surely it can't be acceptable for the AI to prioritize one luxury highly enough to handicap itself with such distant city placement.
 
My favourite one with the AI is when it steals city locations... except it builds its city maybe one hex off from the ideal spot, so when you take the city you have to raze it and build a new one. I don't mind it stealing city locations if it knew how to do it properly!
 
You must log in or register to reply here.

Similar threads

ThermobaricPunch
Multiplayer desyncs (loading screen between turns) and weird bugs
Replies
1
Views
791
coolhandluke11
C
clif9710
Enjoying the changes with BNW vs G+K
Replies
0
Views
852
clif9710
clif9710
M
What's going on here with the AI? (Size 1 capital city not growing)
Replies
13
Views
810
reststillunwritten
R
E
Last notes 4.4.2
Replies
9
Views
920
greenfieldpark
G
S
AI Discussion: Mid Game Expansion
Replies
2
Views
419
Stalker0
S
Top Bottom