AI getting stronger I think

Cities contract for defenders, rather than always building them for themselves (cities without high priority builds can also respond to contracts by building units, and in that case the contract broker picks the best placed city, based on the unit it can make, the promotions it would provide, and the distance it is from where it is needed)

Is the AI able to bring defenders in their SoDs for defending new conquered cities?
 
Excuse me?:mischief:

Well I am the one who is uploading it, I should have at least some say in the matter. :mischief:

Also, I'm only basing suggestions on what I'm seeing and feeling, ie that the cycle is pretty much over in terms of major additions and it's starting to slow down a bit on the forums. That combined with the time since V28 makes me think that starting the freeze next week would be a good idea.
 
The thing that makes S&D hard is also why some of the other things you mention would be hard, so it's the common underpinning that needs tackling sometime anyway. Specifically, the underlying issue is that the AI has absolutely no concept of central control, or really of cooperation. In the original BTS AI, every stack makes totally independent decisions about what to do, so essentially it's locally greedy but has no concept of a bigger picture.

<snip>

Sorry for the doublepost, but I wanted to touch on the specific issue of central control. How hard would it be to create some kind of mechanism on CvPlayerAI which can override the default decisions of the stack (ie, change it's stack UNITAI, change it's destination, change what it plans on doing, etc) and can split and merge stacks as well?

The reason I suggest this (and I know very little about this compared to you) is that it would allow all of the essential mechanisms of CvUnitAI to remain untouched, which will probably be important for the new Combat Mod tags, while giving the AI a much needed boost. The lack of a central intelligence on the part of the AI has been in my empirical observation the number 1 reason it fails against a good and dedicated human attacker in wars.
 
Sorry for the doublepost, but I wanted to touch on the specific issue of central control. How hard would it be to create some kind of mechanism on CvPlayerAI which can override the default decisions of the stack (ie, change it's stack UNITAI, change it's destination, change what it plans on doing, etc) and can split and merge stacks as well?

The reason I suggest this (and I know very little about this compared to you) is that it would allow all of the essential mechanisms of CvUnitAI to remain untouched, which will probably be important for the new Combat Mod tags, while giving the AI a much needed boost. The lack of a central intelligence on the part of the AI has been in my empirical observation the number 1 reason it fails against a good and dedicated human attacker in wars.

I'm planning to allow an override a bit like that, but not to cvPlayer, but to a pseudo unit AI entity which uses a different (new) unit ai type and has it's own (persistent) state. That way ,ultple local commands can exist, and in principal they can have different types (since it would just be a unit a type). Since the local command AIs will themselves be pseudo units, there isn't any fundamental difference between them and the underlying units in erms of structure, which mean they can also delegate upwards hierarchically, ultimately to an overall strategy level (eventually)
 
Is the AI able to bring defenders in their SoDs for defending new conquered cities?

It doesn't explicitly bring defenders (though you're right, maybe it would be a good idea if it did). What happens when it takes a city is that city will decide it needs a certain level of defense, and the SOD will shed units (good defensive ones) to form a city garrison, and xontneu with what's left (or pause and build p more if need be). This is not always the best strategy, but it's a reasonable starting position. Ideally it would take defenses with it, and maybe decide to raze rather than hold if holding involves stalling it's attack on the st of the enemy civ too much. I could fairly easily cause it to take defenders, but the downside is that the ai would have to build up that much more before it eve launched an attack if it waited for enough defenders in addition to it's attack needs, so it's a tough balance. It really needs a deeper strategic evaluation of whether it rally qanat to hold certain cities, or just inflict their loss onits enemy.
 
Huh? What does a field commander getting XP have to do with barbarian generals? If you get a GG (whether it is from fighting barbs with the option on, or from fighting other civs) and make it a field commander you want as much XP for it as it is practical to get since a field commander without promotions is completely useless. Whether or not barb generals is on has nothing to do with it. The point is not to get more GGs, although if the option is on that can be a nice additional benefit, it is to get promotions for the field commander to make it useful.

If you are actively pursuing a war with another civ, it is probably better to send it to the front since it will get XP there too and the instant it gets the first promotion it can start contributing. But if you are not, then taking it hunting is the way to go.
 
Really? I never noticed that. I must not have ever taken one hunting with barb generals turned off (although I usually play with it on anyway), probably since with that off they come much later and when you get one it is because you are at war with someone and probably doing a lot of combat with them (or you wouldn't have gotten the general).
 
Really? I never noticed that. I must not have ever taken one hunting with barb generals turned off (although I usually play with it on anyway), probably since with that off they come much later and when you get one it is because you are at war with someone and probably doing a lot of combat with them (or you wouldn't have gotten the general).

Yeh, I hadn't realized either since I normally play with the option on.
 
It'll also enhance the value of S&D strategies as well!

Thanks for the nudge though H! :D

Note; I will be implementing the SAD tags in the Combat Mod in core units, those are one of the few things in it I'm excited about. I'll assume that Koshling knows about that, because I'm waiting on that until the SAD AI gets in.
 
Note; I will be implementing the SAD tags in the Combat Mod in core units, those are one of the few things in it I'm excited about. I'll assume that Koshling knows about that, because I'm waiting on that until the SAD AI gets in.

I am aware that the combat mod has S&D affecting tags in it, but not familiar with the details. I suspect the implication of any such tags will be second order relative to basic S&D AI anyway though.
 
The point is that with barbarian generals enabled, the field commander will gain xp for barbs/animals being killed. When that option is off, he'll only gain xp for killing units from other civs.

Not so. All units get some exp from killing animals or barbarians. The Barbarian Generals option adds points to the GG generation (and GGs for barbarians) from barb and animal combat. Without it on you don't get GG points from barbs or animals.
 
Not so. All units get some exp from killing animals or barbarians. The Barbarian Generals option adds points to the GG generation (and GGs for barbarians) from barb and animal combat. Without it on you don't get GG points from barbs or animals.

But you still get xp then? In that case it's fine as it is.
 
I am aware that the combat mod has S&D affecting tags in it, but not familiar with the details. I suspect the implication of any such tags will be second order relative to basic S&D AI anyway though.

The main thing I think would need to be considered is that units can very well be specialized for surround and destroy roles, both being the supportive surrounders and the ones to take additional advantage of the support bonus. So if the AI is given to consider particular units as having particular roles in setting up and utilizing surround and support strategies, then those units would be further compelled to promote to extend their role, as well as be chosen to play out those roles.

For more expressed info, refer to the 4th Combat Mod Intro thread. Could be helpful to have these unit abilities in mind before working the S&D AI. I've been a little reluctant to push them into use until some AI was developed.
 
Back
Top Bottom