Post 0.41h AI feedback needed

In the new patch I saw the Kuriotates found OO, something that I'd never seen before, and then build the Tower of Complacency in their capital which was also the OO holy city. It was the perfect strategy for them. Throughout the game they quickly spiralled out of all proportion, bulding huge amounts of swordsmen and completely repelling my invasions. The other AI's that survived my initial Alexis rush did very well too.

And I was playing on prince so I was pretty impressed.
 
Playing as Clan of Embers, I had Kurios next to me. I decided I wanted to invade and so started spamming axemen. Looking through their borders I saw two very substantial stacks of archers and centaurs wandering about, and they seemed to notice my stack if I brought it near, so when I declared war I moved my stack onto a forested hill next to their city rather than attacking the city itself (my culture was pressing up against it). They proceeded to suicide one entire stack against my army -- didn't kill a single orc (Rantine had woodsman and hill defense). I took their city and noticed they'd brought their other stack close, so I left five axemen in that city to maintain the illusion of threat and sent the rest of my army on to their other cities. Their stack (maybe 6-8 centaurs) proceeded to just sit there. Didn't counterattack, didn't move to defend either of their other cities, didn't even pick off the reinforcements from my territory.

So, analysis: they seem to have the right idea about how many units and of what kind to build, but they aren't using them effectively. Suiciding a stack of archers against 20 axemen and a hero on a forested hill is a genuinely bad idea, especially when they could have moved to fortify in the city instead, and having the cavalry sit there motionless while I conquer their cities one after another is even worse -- they should at very least have been killing my reinforcements.
 
just played a clan game on patch i, huge mountaincoast map, increasing difficulty. I started a very early war against the hippus, i retreated after seeing i couldn't take his capitol,a little later i attacked them with a stack of about 15 axemen + rantine and to my surprise he had like 15 warriors in his capitol and + a stack of 50plus warriors with orthus axe just wandering about, he never attacked with them nor did he defend his capitol so my axemen cleared out his capitol and it was game over for the hippus who never expanded beyond the first city.
Also saw some very good ai behavior ,on turn 220 cassiel ,who was first in score declared war on me with a huge stack of dragon slayer with some very potent heroes and it was game over for me.
 
Really? I remember discussion of having the AI use warrior upgrades as a strategy in the WM thread. Wonder who decided not to use that, Sephi or Kael?
I decided to wait with these kinds of strategies until the AI handicap for upgrade costs can be put to a decent level. Didn't know that time that the Deity AI in FFH has a 90% discount on upgrades (compared to 50% in BTS)
 
I haven't played enough to give specifics, but the AI does seem to handle itself better in general.

It did just leave an undefended Orthus's Axe inside Kuriotate territory again. No one bothered to take it even though Kuriotates had open borders with lots of people and the Axe was inside a city's BFC.
 
I just went through the AIchooseProduction function in cvgameutils.py, which handles AI production now in almost all cases if I understand the code correctly. Sometimes it appears that this function gets bypassed. It would be nice if anyone could explain why this happens.

There are some very huge flaws in that function, most notably:

A lot of stuff is missing, so the AI will never produce the following unless in the very rare cases that the function gets bypassed:
1)All rituals, so the AI will only extremely rarely build the Pact of the Nilhorn or any of the Illian related Auric progression path.

2) Several lategame units: All nationals (Knights, Beastmasters,Crossbowmen), Muskets, Horse archers.
This means the AI will build the lower tier units and upgrade them. Since they get a huge discount on upgrades at all levels (75% on noble, i. e. the AI pays only 1/4th of what a human player pays) this is kind of an exploity strategy.

3) All ships apart from workboats.

4) A few world wonders: Bone palace, the pillar of chains has a value of 0(meaning it never gets built), which is bizarre as it's one of the better wonders.

There is also some code at the beginning that aims to prevent the Sheaim from building Pyre zombies as primary city defenders, because they don't function as military police. This check should be modified to also include warriors if the city can build any other kind of troop, as the AI will simply upgrade all warriors to Pyre zombies anyway.

#1 and #3 are problems for obvious reasons. #2 will be very bad for the Clan of Embers and Lurichep, as neither of them can promote units to their end game versions.
 
Oh man, I wish I hadn't read about the AI's upgrade exploit. Is it possible these huge AI armies are not a better economy, but simply the AI just using an exploit?
 
Oh man, I wish I hadn't read about the AI's upgrade exploit. Is it possible these huge AI armies are not a better economy, but simply the AI just using an exploit?

Its not an exploit so much as it is a large bonus. The AI gets many bonuses to make up for its lack of real intelligence on the higher difficulty levels. This one is particularly pronounced.

And no, they had this bonus before, so any improvements in AI army size and composition since the patch are due to AI improvements not this "exploit."
 
Is there anyone on with a better than rudimentary knowledge of python scripting for FFH2 who could supply me with a bit of code I could add to the AI_chooseProduction function?

I would like to get the AI building some ships in FF2 v2.41-i while I'm testing.
 
The AI does not defend cities it takes with attack groups. It also doesn't send defenders along with attack groups, so the result is that when a city is captured the attackers leave the city undefended until it can build its own defender or one is sent from another city. This leaves a window of opportunity for an enemy to simply walk into the city unopposed. (I think someone else already reported this; I just wanted to confirm that I'm seeing it too, and frequently.)
 
The AI does not defend cities it takes with attack groups. It also doesn't send defenders along with attack groups, so the result is that when a city is captured the attackers leave the city undefended until it can build its own defender or one is sent from another city. This leaves a window of opportunity for an enemy to simply walk into the city unopposed. (I think someone else already reported this; I just wanted to confirm that I'm seeing it too, and frequently.)

This happens in BtS, too, but the window is smaller or nonexistent because the attack stack will wait around for units to heal. By the time they are ready to move on, the AI usually has defenders in place. On the other hand, FFH has more opportunities for aggressors to capture cities without taking any damage because of magic and summons and stuff, so it is more common for an attack group to move on immediately, providing the window of opportunity to recapture.
 
You can allow the AI a small cheat to expand a little bit. In handicapInfos the AI is granted a certain amount of free production in their initial cities, IIRC this is only applied in their first city ever generated, but you can easily grant it to any city they ever have, thus it should also apply to those they capture, allowing them a nice boost toward that first defender (it isn't large enough of a bonus that it can be much of an exploit, but I am sure people will whine about the AI cheating if you do it, but never notice the effects ingame)
 
But cities they capture have revolt time which doesn't get a chance to expire.
 
Hmm... not only does the AI not build ships, it also disbands any ships you give it in the world builder. Very curious. And this is with an AI nation completely trapped on an island (Australia in my game).
 
I'm in the early mid-game (2xx turn at medium speed) and my neighbour Falamar has got 5 cities and still hasn't build any culture, apart from capital (4 out of 5 of his cities' 'fat crosses' are neither fat, nor crosses). Other AIs are performing normally in this respect.
 
The AI also doesn't really seem to know what to do about Barbarians. More than once I've seen a lone barbarian warrior saunter into Decius's territory, pillaging as it goes, and it's not until a good while later that he splats it with one of his axemen stacks (which was within striking distance of the barbarian the whole time).

I think this is due to the AI not being suicidal anymore. They seem to radomly gather into stacks for combined attacks and pillage along the way. They are hard to mount a defense against when we're used to buffing up on their suicides. It has increased the Barbarian difficulty in the early game. :goodjob:
 
The better AI shouldn't really apply to all units. Skeletons for example; they should just beeline the nearest living non-barbarian unit and try to kill it.

I agree. I'm lovin' the new smarter Barbs, but I miss the indescriminate aggression. Skeletons, unless summoned and under the summoner's command, should suicide.
 
The AI does not defend cities it takes with attack groups. It also doesn't send defenders along with attack groups, so the result is that when a city is captured the attackers leave the city undefended until it can build its own defender or one is sent from another city. This leaves a window of opportunity for an enemy to simply walk into the city unopposed. (I think someone else already reported this; I just wanted to confirm that I'm seeing it too, and frequently.)

Yeah. Arturus lost one of his cities to barbs while I didn't have visibility, took it back with a stack of ~20 warriors, then moved all of the units out except for a warrior and a scout. Now the barbs are about to retake the city while the warriors sit one square outside it.
 
Top Bottom