Community Feedback Needed: Biggest AI issues
- Thread starter Kael
- Start date
The AI does not seem to do (Unit A -> Upgrade to Unit B) when the results are far more powerful than simply building unit B.
The AI will waste a lot of units attacking summons that will disappear before they can do anything; try sticking a decoy earth golem outside one of your cities to see what I mean.
I've never seen the AI place it's units in an invulnerable spot (in fire for orcs, on mountains for druids, on water for water-walking civilization-smashing dwarven druids)
The AI never casts courage (I think, I may just have never fought an AI making use of spirit mana) and this may extend to other permanant buff spells.
The four horsemen sit in cities, since the AI seems to have the strategy of defend with toughest unit/end out other units. Having the horsemen spawn a defensive unit when they cap a city and then moving on would fix this, without the drawbacks of razing end-game cities
On a related note, Orthus gets "stuck" on lairs and bear dens, presumably due to the "toughest unit defends" strategy.
The AI will waste a lot of units attacking summons that will disappear before they can do anything; try sticking a decoy earth golem outside one of your cities to see what I mean.
I've never seen the AI place it's units in an invulnerable spot (in fire for orcs, on mountains for druids, on water for water-walking civilization-smashing dwarven druids)
The AI never casts courage (I think, I may just have never fought an AI making use of spirit mana) and this may extend to other permanant buff spells.
The four horsemen sit in cities, since the AI seems to have the strategy of defend with toughest unit/end out other units. Having the horsemen spawn a defensive unit when they cap a city and then moving on would fix this, without the drawbacks of razing end-game cities
On a related note, Orthus gets "stuck" on lairs and bear dens, presumably due to the "toughest unit defends" strategy.
Bad Player
Deity
Maybe it's a good strategy to join an existing religion so as to get diplomatic bonuses.
Nikis-Knight
Deity
Generally, but imo it makes for a more interesting game if more religions get founded, even if that means the ai is more likely to fight amongst itself.
Besides, diplo bonus is less important for ai, since they can't count on it mattering from the player's perspective (You might kill them even if they are your allies.)
So I'd like trying to get their own religion to be a priority, unless they had heavily invested in their first.
I agree with mailbox as regards religions. I usually turn off RoK & FoL otherwise they dominate the game. They get founded early, and the Ljosalfar & Khazad AIs spam Disciples/Thanes like crazy until whole continents are under their religion.
xienwolf
Deity
Should Bhruic's Tradefix be used, the AI needs to be given an extra line to seek up to 3 of any Mana type. But more importantly (and needed even without Bhruic's fix), they need to assign a value to the extra Mana beyond the first.
If Resources which they already have are all granted a value in trade still, the AI is less likely to demand the overly unfair trades in general. The value should be based off: Buildings with cummulative bonuses (do not believe that we actually have any), Potential Trade Value (is the resource strategic & do I have good relations with someone who does not have the resource), Free Arcane Promotion (do I have 3 of the mana?)
This tends to be a problem for gameplay, though. Any game with the Khazad or Luchuirp seems to result in all of the Evil civs converting to Runes of Kilmorph and becoming Neutral. The incentive to spread religions needs to be toned down, IMO, or at least for RoK. (An alternative would be to make RoK alignment-neutral like FoL.) This doesn't seem to be much of a problem in reverse with OO.
By contrast, Ashen Veil and the Order are almost always the last religions to be founded. There should be stronger incentives for the AI to go for them.
Monkeyfinger
Deity
- Joined
- Oct 24, 2006
- Messages
- 2,002
You're not annoying, the trading glitch is.
And on that note, it needs to work in reverse too. The AI's currently not willing to trade you mana of a type you already have, even though there is a reason to have cumulative mana sources.
And on that note, it needs to work in reverse too. The AI's currently not willing to trade you mana of a type you already have, even though there is a reason to have cumulative mana sources.
Here are the major things I have noticed that don't involve combat magic.
AI city positioning is poor in general. I see far too many cities founded with a 1 tile gap to freshwater/coast putting them at a health disadvantage.
Civ specific: Lanun don't seem to build pirate coves, elven civs with FoL should be using bloom.
AI with significant amounts of desert should get Water I and terraform.
Bears will walk into barbarian cities and sit there as defenders, probably shouldn't be doing this.
Diplomatically hostile AIs make no attempt to stop a religious victory.
AI builds too many scouts early game weakening its performence on both attack and defence.
AI city positioning is poor in general. I see far too many cities founded with a 1 tile gap to freshwater/coast putting them at a health disadvantage.
Civ specific: Lanun don't seem to build pirate coves, elven civs with FoL should be using bloom.
AI with significant amounts of desert should get Water I and terraform.
Bears will walk into barbarian cities and sit there as defenders, probably shouldn't be doing this.
Diplomatically hostile AIs make no attempt to stop a religious victory.
AI builds too many scouts early game weakening its performence on both attack and defence.
Monkeyfinger
Deity
- Joined
- Oct 24, 2006
- Messages
- 2,002
I'm responding to uberfish saying this but this is something that a lot of people have been saying, not just you, don't take my singling you out the wrong way or anything, it's just the most convenient.
Often, the tile configuration you gain from doing this is worth more than 2 extra health, especially in FfH where you have agriculture and sanitation and 3 early health buildings to make health less important. Also, there's no levees in FfH, meaning less reason to seriously regret not building next to that river in the last third of the game. I agree that the AI does this, but I disagree that it's a problem. I tend to find that the AI is pretty smart about realizing when 2 health isn't worth sacrificing a certain fat cross configuration.
Often, the tile configuration you gain from doing this is worth more than 2 extra health, especially in FfH where you have agriculture and sanitation and 3 early health buildings to make health less important. Also, there's no levees in FfH, meaning less reason to seriously regret not building next to that river in the last third of the game. I agree that the AI does this, but I disagree that it's a problem. I tend to find that the AI is pretty smart about realizing when 2 health isn't worth sacrificing a certain fat cross configuration.
MagisterCultuum
Great Sage
Up to Four. Why would you stop at getting the second level of a spell sphere if you could get the third?
Placing cities next to rivers also enables the brewery so it's more than just 2 health. Placing cities next to coast enables harbour/lighthouse and gives you workable coastal tiles. I think you are incorrect that the AI is making good tradeoffs in its positioning, because it will for example often found 3 cities away from water to claim a certain set of resources when it could have placed 2 cities on water and one off.
Starting locations are a big issue, though i dont know if anything can be done about it. I often see that some A.I.s have their starting city in a location sourrounded by ice, mountains and tundra. Maximum population is two and the A.I. will never recover from this bad start. If there is a way that the A.I wont found its first city instantly then let them move untill the find a spot that will give the city at least a max pop of five. (If this kind of coding is possible then it could be applied everytime an A.I. founds a city, this will eleminate all the useless cities the A.I tends to built.)
KingOfLands
Warlord
- Joined
- Dec 21, 2006
- Messages
- 227
Capria is profoundly "aggressive" with missionaries too. It's distinctly possible that the AI is trying for a religious victory but doesn't understand how inquisitions work and can't ever get it.
Slvynn
Duke Vector fon Pixel
- Joined
- Apr 3, 2005
- Messages
- 2,073
Yeah btw disciple spam is really annoying. Ai should favor building combat units more.
Once i was killing about 100 fol amelanchier disciples. Alot of free xp, and alot of waste of time. I just imagine if those hammers were spent, lets say for rangers.
Once i was killing about 100 fol amelanchier disciples. Alot of free xp, and alot of waste of time. I just imagine if those hammers were spent, lets say for rangers.
marioflag
History Addict
In general i have seen a poor use of Heroes by AI. They should use their heroes more carefully than a normal unit.When attacking, the AI should risk its hero at higher percentage of success than the chance of success AI would use normal units. AI should also use heroes in bigger stacks than normal.
When defending, heroes should be the last units attacking enemies which are more than 1 tile away from a city so that it doesn't use its hero to kill a unit and the next turn it is killed by nearby enemy stacks.
About Orthus, it should wait some turns before it goes attacking enemy civs, so that he gets some free XP and become more powerful. he sould get also a small bunch of goblins and orcs to make him less liable to be killed 6-7 turns after he has been spawned.
When defending, heroes should be the last units attacking enemies which are more than 1 tile away from a city so that it doesn't use its hero to kill a unit and the next turn it is killed by nearby enemy stacks.
About Orthus, it should wait some turns before it goes attacking enemy civs, so that he gets some free XP and become more powerful. he sould get also a small bunch of goblins and orcs to make him less liable to be killed 6-7 turns after he has been spawned.
Sofista
Deity
I saw many times entire stacks parked outside Acheron's city, unable to attack, but still standing there even during wartime. Could the AI be taught to take care of the little darling when it's ready to do it for real?
Another thing the AI can't handle are barb cities spawning right beside Barbatos: a crazy, wasteful cycle of capture city-lose it-recapture-lose it again ensues.
Another thing the AI can't handle are barb cities spawning right beside Barbatos: a crazy, wasteful cycle of capture city-lose it-recapture-lose it again ensues.
Ecofarm
Deity
Good point. I often disable Acheron because multiple AI civs will suicide/park armies at it and not move them home even during war.
Similar threads
