How is the AI ?

Yeah, it's an opt in.
 
Sorry to bother, how does one do that?

Ah, right click humankind in steam library > properties > betas > drop down select 1.2.116. Will show that in [ ] next to HK in your library.
 
Last edited:
perhaps i'm misremembering but captured flag on botched sorties meant the city was lost... and even if it isn't, how is she triggering another sortie? my siege move that she ignored was 3 turns ago

anyway
 
Sorry to bother, how does one do that?

Ah, right click humankind in steam library > properties > betas > drop down select 1.2.116. Will show that in [ ] next to HK in your library.
Wow thanks so much for this info I'm going to give this a try too!

What happens when they release an official update btw? Do I need to go change the settings again to opt out of this beta (which becomes obsolete) or does Steam do that automatically?
 
Aha this game I had lots of sorties, perhaps the beta update had more aggressive siege defense. I can confirm that you can capture a city during a sortie by killing all of the enemy units. Since you do not have a flag to capture, you cannot capture the city that way, and as you say sounds like defending your flag results in returning to the status quo.

The AI does seem much less inclined to run its defenders out of its walls. I didn’t have any ranged units this time, so that’s definitely the correct choice. I don’t think I play Mycenaeans right, focus way to much on getting enough units to win a war, and not enough on other era stars. Won one war but got greedy and invaded a classical era opponent. Units held their own against sword and Maya javelins but the battles were too even for my taste ;) Interestingly, most of the world was at war at this time, and the nation I beat declared on me when I let my guard down. If this all is at all related to some of the updates, it is looking promising.
 
I downloaded the beta, increased difficulty one level to Nation, selected large map as suggested and wow what a HUGE difference it made indeed. The AI plays well, defends its cities and does build units after all! I am impressed how just one difficulty level changes the game so drastically. The AI must really be programmed to be super passive at default difficulty and below. I’m definitely enjoying it this time. I’m changing my Steam review to recommend. The terrain and cities look beautiful and tactical combat is very clever. I have a feeling I’m going fo sink thousands of hours into this game.
 
The AI must really be programmed to be super passive at default difficulty and below.

Well, they probably had to do something like that in order to try and capture many players of other competing games that got used to very passive, very easy AI.

If they really did that, programming different levels of AI activation, that shows skills, and that would mean that we can expect better and better things from the game in time. As far as I know, only one game has done that type of AI programming before, and that was Galactic Civilizations 2... but Brad Wardell is from another planet.

I have yet to try the beta, but even without it, the AI plays well at Empire level.
 
I have yet to try the beta, but even without it, the AI plays well at Empire level.

On my empire level games my neighboring empires are very bad at attacking, sometimes they send one unit at a time that I just autoresolve to save time.
They can have big fame score though from all the bonuses on that level. They should build a huge army and invade. (not sure is the combat strenght bonus on this level yet or highest)
 
Yeah, this war had I think Maya had 6-8 armies that I saw, they were ask at war with another nation and their vassal which I didn’t see. Maya is programmed to build armies of 2 swords and 2 javelins and it’s very easy to end up fighting 2-3 of those at a time. They also target multiple sites at a time, I lost one city because I diverted my whole army to defending against their main attack and they sent another army to the other edge of our border.

I’ve started playing regular elevation, but few cliffs, and it better tunes the game to the AI strengths.
 
Since obviously Amplitude bases there changes on our conversations here ;) I noticed the AI does itself a disservice triggering pseudo-sorties by engaging an army about to siege their city. These engagements cannot result in taking the city, though they still benefit from the walls, but they miss out on including an army of citizens in the battle that will ultimately decide the siege that comes immediately after.
 
Since obviously Amplitude bases there changes on our conversations here ;) I noticed the AI does itself a disservice triggering pseudo-sorties by engaging an army about to siege their city. These engagements cannot result in taking the city, though they still benefit from the walls, but they miss out on including an army of citizens in the battle that will ultimately decide the siege that comes immediately after.

Well, yesterday the AI kicked me out of what looked like an easy siege using a sortie, supported by some navy and incoming embarked units. Fortunately I had my own fleet of cogs close by and made them pay a price, but lost 65% of the sieging army and had to retreat. It was a siege on a big island city, so I assumed it was an easy capture.

This was with the beta.
 
i'm all in favor of the AI getting better at defending cities. but it should do it tactically, finding weaknesses and taking advantages of positional strengths
the idea of multiple sorties sounds wrong...

a player, human or AI, that losses an engagement should give up the initiative
 
It seems to me that a battle should end when one side only has ships left. When this happened to me I luckily had just upgraded three muskets and so went and shot all the non-artillery shops. But without range, or with a malicious opponent, a single ship could force a sortie (or other battle) to linger full the full 6 rounds.

What the pseudo-sorties do is let the AI leverage their walls to stall out the end of a war. In a recent war a 6 tile city had 6 citizens, and it took 6 combat rounds to clear with a swarm with Aksumite swords, finishing the turn of the forced surrender. Perhaps if any opposing units are standing next to a city when a battle starts, that battle should be converted to sortie rules with militia and the risk of losing the city.
 
i'm not sure what you mean by pseudo sorties. if it says siege at X, defend your flag then it a sortie period. otherwise, you're being attacked by regular units and default battle rules apply, walls and buildings are simply terrain features... or am i missing something?

i'm trying to understand the rationale behind the idea that a city is able to go on multiple succesive sorties after failing to take the player flag

i can see how it could remain in control of the city, but a loss should take the role of attacker away. otherwise it's effectively buuying turns to bleed you war support abusing the "perfect" reaction time
 
Ah, sorry for the confusion, I’m jokingly defining pseudo-sorties as battles the AI triggers by sending a unit out of the city as you are moving in to siege it, and engages you outside the city before you get a chance to engage the city in a siege. They then get most of the benefits of a true sortie (walls minus militia, but if they have an army; the militia is often not that helpful due to limited space within walls), without giving you the chance to capture the city on a victory. I suggest this should be classified as a sortie provided the defender has one army immediately outside the city walls. Often these battles get triggered while moving everyone into place around the city, so this would allow the besieging party to move their most wide open army to the wall first. The the sortie can either be triggered once the siege begins, or by attacking one of armies before they trigger a siege.
 
One thing to trick the AI in battle is to move a ranged unit into a semi-unprotected position, as a bait.

This poor ranged unit will draw AI units like how a Black Hole drawing everything into it.
 
I just had the rare experience of full-blown war in the contemporary era with multiple opponents (they declared it). I didn‘t have any planes or artillery ready, as I thought I would get to the end without any more fighting. And well… trying to capture cities from the AI at that stage is now that the AI stays within the walls quite hard without support. They almost never move and all their (Soviet) units are dug in which makes them take so many hits. I really like that. No more conquering just with infantry and tanks!
 
Oh wow, that's awesome!

I just "finished" a game (entered contemporary only 200 fame behind the leader, with enough vision on the leaders to know they are stalling), and while the game had lots of great conflict, and the AI took plenty of land off each other, a very simple problem occurred which has been arising in a lot of my games.

Each AI empire (I am allied with all but the one I just won a war against) had one army stack of 7. I think Amplitude needs to add some hardcoding/parameter balance to ensure that AI aims to have something closer to 40 units each in contemporary.
 
Oh wow, that's awesome!

I just "finished" a game (entered contemporary only 200 fame behind the leader, with enough vision on the leaders to know they are stalling), and while the game had lots of great conflict, and the AI took plenty of land off each other, a very simple problem occurred which has been arising in a lot of my games.

Each AI empire (I am allied with all but the one I just won a war against) had one army stack of 7. I think Amplitude needs to add some hardcoding/parameter balance to ensure that AI aims to have something closer to 40 units each in contemporary.
I think the AI is afraid of upkeep costs. Agreed, they should have larger armies even in peace times. Every player keeps several stacks around, and so should the AI.
 
Given the maintenance costs and how easy it is to rush-buy, shouldn't everyone (players and AI) just build a massive army when war breaks out, instead of maintaining large stacks all game?
 
Top Bottom