I don't think so because the Kuriotates' cities are just as bad as everyone else's, when they should be the ones going to the greatest lengths to settle on freshwater.
Community Feedback Needed: Biggest AI issues
- Thread starter Kael
- Start date
I don't think so because the Kuriotates' cities are just as bad as everyone else's, when they should be the ones going to the greatest lengths to settle on freshwater.
1.in my last game hyborem would not heal himself
he stood in his city of dis for many turn without regaining his hp
he had 0.1 of them btw
2.for some reason after one of the updates (cant say which one) the fellowship of leaves and kilmorph religions started dominating the map in every game i start and because of that everyone is neutral and the game pretty boring. i play huge 20 civs monarch.
thanks alot great game
kael i hope you and your crew are getting payed
he stood in his city of dis for many turn without regaining his hp
he had 0.1 of them btw
2.for some reason after one of the updates (cant say which one) the fellowship of leaves and kilmorph religions started dominating the map in every game i start and because of that everyone is neutral and the game pretty boring. i play huge 20 civs monarch.
thanks alot great game
kael i hope you and your crew are getting payed
Oh yeah, the AI needs to prioritize early military techs a lot more. I think the new tech tree confused it. If they don't have bronze or horses, they should research archery and use archers for defence. Cities should not still be defended by warriors on turn 200 at immortal/normal speed.
The AI is not allowed to / don't know how to use the crew promotions. This combined with trireme having a cargo capacity of one causes the AI to send out settlers without defenders, which is a really bad idea. (especially when sending them to a barb infested island!)
Nor'easter
Deity
- Joined
- Feb 25, 2007
- Messages
- 2,036
This is more of a late-game issue, I guess, but I don't think I've ever seen the AI do a ritual. I don't know if it's possible to program so that circumstances will affect the percentage chance of a given AI starting a given ritual. Some possibilities:
RITES OF OGHMA
You could have the percentage chance that an AI civ would start this ritual increase with the size of their territory. The larger your territory, the greater the likelihood that you'll benefit from this ritual.
GENESIS
The percentage chance of carrying out this ritual could maybe tie in with the percentage of upgradeable (i.e., non-grassland) tiles in the AI civ's territory. Neutral civs would have a very low chance of trying Genesis, since they can terraform with Druids.
GLORY EVERLASTING
The likelihood of starting this should probably be related to alignment in some way, and you'd need a super-production city for the AI to start this.
NATURE'S REVOLT
I don't know what should affect the percentages here.
Anyway, just something to think about.
RITES OF OGHMA
You could have the percentage chance that an AI civ would start this ritual increase with the size of their territory. The larger your territory, the greater the likelihood that you'll benefit from this ritual.
GENESIS
The percentage chance of carrying out this ritual could maybe tie in with the percentage of upgradeable (i.e., non-grassland) tiles in the AI civ's territory. Neutral civs would have a very low chance of trying Genesis, since they can terraform with Druids.
GLORY EVERLASTING
The likelihood of starting this should probably be related to alignment in some way, and you'd need a super-production city for the AI to start this.
NATURE'S REVOLT
I don't know what should affect the percentages here.
Anyway, just something to think about.
this is because the khazad and the ljos have a hardcoded beeline for RoK and FoL. even though this is a viable strat, imho it's wrong to force them into specific religions. Tbh, every civ should get about 3 different tech priorities, eighter assigned by a RNG or a smarter AI script evaluating starting terrain.
On a side note, i also believe tying the hydrassil tree to FoL is wrong, it should rather be on a recon tech.
Slvynn
Duke Vector fon Pixel
- Joined
- Apr 3, 2005
- Messages
- 2,073
Also, Kurios and Malakim have one advantage - they have 2 usefull manas from palace for divination tower. Its so good that its really should be a stratgy and both player and both ai, and should be used as civ major advantage. Varn espessially should try to try and build Divination tower asap and get best tech with it. (lets say Theo, or righteousness at least, or warhorses if he uses a horses). Also that really suck when some good civ build entropy/chaos/death node. Ai should build nodes according civs favor and synrgy with starting palace mana.
Fenham Dave
Chieftain
My AI issue I would love to see sorted is with religion. I love the way that you can still follow a religion without actually founding it due to the free acolyte unit you get when you discover a religious tech. The AI however doesnt seem to use this to its advantage.
As an example in a recent game of mine using the creation map script, the Sheaim were beaten to Corruption of Spirit by someone else who didnt spread AV and didnt convert to it. I'm not sure what happened next (whether they gave up on Corruption of Spirit or just misused the free Savant) but the Sheaim where somewhat isolated and as a result of no religion spreading to their lands they remained without a religion for most of the game. Obviously a human player would have carried on, converted to AV, summoned Hyborem and created general havoc for all.
I have no idea how difficult this would be to implement but giving the AI the option to use religions they didn't necessarily found would in my opinion be a great addition to the AI and make for more diverse religious wars and hopefully better games.
I'm sure other AI issues are more pressing but if this made it onto the list one of your many long term (and usually quiet) fans would be very happy.
As an example in a recent game of mine using the creation map script, the Sheaim were beaten to Corruption of Spirit by someone else who didnt spread AV and didnt convert to it. I'm not sure what happened next (whether they gave up on Corruption of Spirit or just misused the free Savant) but the Sheaim where somewhat isolated and as a result of no religion spreading to their lands they remained without a religion for most of the game. Obviously a human player would have carried on, converted to AV, summoned Hyborem and created general havoc for all.
I have no idea how difficult this would be to implement but giving the AI the option to use religions they didn't necessarily found would in my opinion be a great addition to the AI and make for more diverse religious wars and hopefully better games.
I'm sure other AI issues are more pressing but if this made it onto the list one of your many long term (and usually quiet) fans would be very happy.
The problem here is if you are playing with Religious Victories active, it is a bad idea to join and spread a religion you didn't found, as you are only furthering your opponents victory.
But I think we've established that the AI doesn't play to the victory conditions.
Ecofarm
Deity
If you are not careful, it can still happen. I've had perp do it to me before... just when I was ready to conquer the world
![[NWO]_Valis](https://civfanatics-data.community.forum/avatars/m/110/110973.jpg?1475248361){kind=avatar}
[NWO]_Valis
A bad kitty!
Are you sure that they did not have CoE as their state religion? I think AI is using their free missionaries to establish the religion in their empire but I wont put my money on it.
TheGreatSteve
Warlord
- Joined
- Jan 20, 2006
- Messages
- 145
The AI's, especially the peacemongers and the Khazad, need to be smarter about early expansion and not getting blocked in. To often I see them use their first settler to found a city right next to their last in an area with a fair amount of resources, instead of putting it in a less hospitable spot farther away. It seems great, except it means it's the last city they'll make because I can now wall them off from most of the continent and they won't do anything about it.
The more aggressive civs don't have a problem here, they'll just use their nice second city to pump guys and take my lousy second city before I can expand everywhere and start pulling ahead, but the nice guys are content to just sit there with 7% of the landmass and watch as they become less relevant turn by turn.
The Khazad are of course a bit different, they're not all that nice, but their slow expansion makes things go the same way. Give them a little room, and they'll be happy. By the time they realize they can't get any bigger without going through you you've got the power to keep them locked in.
The AI is way to eager to jump on the first religion that spreads to them. On a religionless continent of (Good) civs, they shouldn't switch to AV just because I spread it in one of their border cities. Because the 15 people of Cowtown, Nowheresville started worshipping demons, Capria decides to end hundreds of years of good relations with her neighbors and become buddies with a guy on the other side of the world? Not a good strategy.
I think someone earlier mentioned forest starts and how they weaken the AI's early game. I know techs can make workers do stuff faster, can a tech increase the speed at which specific improvements are made? If it can, you could make chopping available at 1/3rd speed at Exploration or Crafting, then make Bronze Working triple it to where it is now. AI loves early workers, giving them something to do on those bad starts would help alleviate matters. They couldn't clearcut the land, but getting a few spots for farms and mines open early would help. Could do the same with jungles too, make it available at BW and regular speed at Sanitation.
The more aggressive civs don't have a problem here, they'll just use their nice second city to pump guys and take my lousy second city before I can expand everywhere and start pulling ahead, but the nice guys are content to just sit there with 7% of the landmass and watch as they become less relevant turn by turn.
The Khazad are of course a bit different, they're not all that nice, but their slow expansion makes things go the same way. Give them a little room, and they'll be happy. By the time they realize they can't get any bigger without going through you you've got the power to keep them locked in.
The AI is way to eager to jump on the first religion that spreads to them. On a religionless continent of (Good) civs, they shouldn't switch to AV just because I spread it in one of their border cities. Because the 15 people of Cowtown, Nowheresville started worshipping demons, Capria decides to end hundreds of years of good relations with her neighbors and become buddies with a guy on the other side of the world? Not a good strategy.
I think someone earlier mentioned forest starts and how they weaken the AI's early game. I know techs can make workers do stuff faster, can a tech increase the speed at which specific improvements are made? If it can, you could make chopping available at 1/3rd speed at Exploration or Crafting, then make Bronze Working triple it to where it is now. AI loves early workers, giving them something to do on those bad starts would help alleviate matters. They couldn't clearcut the land, but getting a few spots for farms and mines open early would help. Could do the same with jungles too, make it available at BW and regular speed at Sanitation.
slowcar
King
most was already mentioned, anyhow:
not enough focus on essential starting techs and the things they enable like markets, city states, education, sage
non-elf civs should beeline bronze working if there are many trees around
AI is over-expanding a lot. on noble it can afford some but it could do a lot better without spamming cities everywhere without having access to city states, courthouses, forbidden palace, summer residency, markets and kilmorph temples. it should be manageable to calculate the maintenance cost of a new city (including the heightened cost for all other cities) and decreasing the value of settlers accordingly.
no flexible defense. if the AI stacks units in it important cities you can take them out one by one instad of being overwhelmed by a combined assault.
war declarations with too small stacks. in my last game (submitted for ecofarms immortal thingy) i got early war declarations with only 5 warriors/scouts. they moved back and forth until i killed them as they had no chance attacking my city (at least they did not attack)
moving in stacks is not flexible, a single sun2 spell (binding light?) can immobilize 20+ units until you massaged them down with some air-magic
promotions are not handled well. at emperor and below a single shock2 warrior can eleminate whole civs (with the captured cities supplying).
as there are less paper/scissors/stone units (like spearmen, axes etc in vanilla) in FFH the AI should fill this gap using promotions.
having like shock2, formation and cover units to protect a stack would make it a lot stronger in defense, not speaking from attack.
(many a human player is not able to do this, too
)
AI does not manage healing. stacks should be supported by disciple units that are not used in combat - except maybe in 90%+ situations
heroes are not covered enough, you can often wear them down.
orthus should get some company and wait till 10xp at least. should go for combat1, shock, combat3, march, shock2.
strong barb units should not sit in cities defending them. especially horsemen. and orthus. and...
collateral units should only attack if there are sufficient other units available. bombard or withdraw else.
not enough focus on essential starting techs and the things they enable like markets, city states, education, sage
non-elf civs should beeline bronze working if there are many trees around
AI is over-expanding a lot. on noble it can afford some but it could do a lot better without spamming cities everywhere without having access to city states, courthouses, forbidden palace, summer residency, markets and kilmorph temples. it should be manageable to calculate the maintenance cost of a new city (including the heightened cost for all other cities) and decreasing the value of settlers accordingly.
no flexible defense. if the AI stacks units in it important cities you can take them out one by one instad of being overwhelmed by a combined assault.
war declarations with too small stacks. in my last game (submitted for ecofarms immortal thingy) i got early war declarations with only 5 warriors/scouts. they moved back and forth until i killed them as they had no chance attacking my city (at least they did not attack)
moving in stacks is not flexible, a single sun2 spell (binding light?) can immobilize 20+ units until you massaged them down with some air-magic
promotions are not handled well. at emperor and below a single shock2 warrior can eleminate whole civs (with the captured cities supplying).
as there are less paper/scissors/stone units (like spearmen, axes etc in vanilla) in FFH the AI should fill this gap using promotions.
having like shock2, formation and cover units to protect a stack would make it a lot stronger in defense, not speaking from attack.
(many a human player is not able to do this, too
)AI does not manage healing. stacks should be supported by disciple units that are not used in combat - except maybe in 90%+ situations
heroes are not covered enough, you can often wear them down.
orthus should get some company and wait till 10xp at least. should go for combat1, shock, combat3, march, shock2.
strong barb units should not sit in cities defending them. especially horsemen. and orthus. and...
collateral units should only attack if there are sufficient other units available. bombard or withdraw else.
.Pyr0mancer
Prince
- Joined
- Jan 18, 2008
- Messages
- 542
AI should actually seek to upgrade its adepts when they hit level 4, and not just keep giving them level 1 spell promotions, if they have Sorcery.
Most recent game, Dominated an adept from the Sheaim(Tebryn Arbandi) that was level 9(!!!), had combat IV + Extension II + some spheres. This was very late-game, Sheaim already had Pass through the Ether, Strength of Will, and Omniscience. I suspect the problem is a lack of gold due to general AI economy mismanagement though.
Also, would it be possible to make AIs actually target mage/archmage-containing stacks with their assassins? Same game, I had 4-5 assassins suicide themselves on a stack of a dozen or so champions, but ignore the stack of 3-4 mages, 2-3 Confessors, and 3-4 champions that was right next to it.
Most recent game, Dominated an adept from the Sheaim(Tebryn Arbandi) that was level 9(!!!), had combat IV + Extension II + some spheres. This was very late-game, Sheaim already had Pass through the Ether, Strength of Will, and Omniscience. I suspect the problem is a lack of gold due to general AI economy mismanagement though.
Also, would it be possible to make AIs actually target mage/archmage-containing stacks with their assassins? Same game, I had 4-5 assassins suicide themselves on a stack of a dozen or so champions, but ignore the stack of 3-4 mages, 2-3 Confessors, and 3-4 champions that was right next to it.
On the subject of AI expansion, it's closely related to the poor city placement. BTS changed the civ AI to build its cities closer together which is a good strategy for BTS (where happiness is quite difficult to come by so packing cities in more tightly increases your total number of worked tiles) but a rather poor one in FFH where resource specials are more common and you can have much larger cities in the midgame with brewery, carnival, religion, gambling houses, consumption etc.
I think the FFH AI would perform better if it was tuned to build less, more widely spaced, better positioned cities, and spent the production/economy it saved on settlers and city upkeep on getting archers for city defence more quickly.
I think the FFH AI would perform better if it was tuned to build less, more widely spaced, better positioned cities, and spent the production/economy it saved on settlers and city upkeep on getting archers for city defence more quickly.
When AI is happy with the number of units defending a city, it doesn't care what they are. Late game, there are still warriors guarding cities. If you go in with a shadow and kill these warriors, they are quickly replaced with higher tier units. The AI should try and upgrade city defenders, or replace them with better defenders.
While at war, all AI mages should be casting spells every turn. If at the start of the turn, they automatically cast their fireball, maelstorm, or summon, then they may use the fireball appropriately and can judge their attacks better. At the end of the turn, they should all cast blinding light, treetop defense, or other defensive spells. These spells will never cause war and the only drawbacks I can see to doing it automatically is unit maintanance cost for summons and computer slowdowns.
Summons should have a tag that stops the AI from tracking them down to attack. Unless they are in a stack with other things or are blocking a path, there is no reason for the AI to attack them (I will often attack summons for xp, but wouldn't want the AI to think this way)
Likewise, summons should have a tag that identifies them as throwaway units that can be used to attack even with a low success rate.
While at war, all AI mages should be casting spells every turn. If at the start of the turn, they automatically cast their fireball, maelstorm, or summon, then they may use the fireball appropriately and can judge their attacks better. At the end of the turn, they should all cast blinding light, treetop defense, or other defensive spells. These spells will never cause war and the only drawbacks I can see to doing it automatically is unit maintanance cost for summons and computer slowdowns.
Summons should have a tag that stops the AI from tracking them down to attack. Unless they are in a stack with other things or are blocking a path, there is no reason for the AI to attack them (I will often attack summons for xp, but wouldn't want the AI to think this way)
Likewise, summons should have a tag that identifies them as throwaway units that can be used to attack even with a low success rate.
Bad Player
Deity
Plus the defensive advantage you can sometimes get by placing a city adjacent to a river - it means that often enemies will either have to cross the river to attack the city or move more turns to go onto another square (which may be more vulnerable etc anyway and also this can slow down their attack).
Similar threads
