AI: Thing’s I’ve got that you don’t

Rallypoints was on my wish list for E:wom.

Wouldn't it make more sense to gather the army in friendly territory and then move it? Seems very dangerous to leave singular units so close to the enemy city. Even if it's just a few more tiles back, it should at least be outside of sight range for the player.

Edit: Edited back, so frogboy's reply makes sense. Lol.

Not a singular unit, that's part of the trick. They gather in small armies and merge later.

Here's a tactical battle debug.  I don't know how much time I'll get to work on tactical battles prior to the public beta but it's an area I'm very interested in since the strategic AI is starting to get to the point of diminishing returns.

http://screencast.com/t/IBPVKwulBYHK

Another:

http://screencast.com/t/sLI3XqUJ2

Oh, you guys don’t get them. There’s no UI for it presently.

Waypoint DCL

edit: That last video was pretty damn epic.  Can't wait for beta.

LOL.

Well, those pink space ponies aren't cheap.

Of course, right now, I'm having trouble with the AI getting wiped out by quests.

I liked it better when it just avoided them.

Edit: nvm, bad idea.

I think I'll need an AI tag in the quest XML that defines how hard the quest is.

I think players will get wiped out by quests too, but we're jerks and just reload.     Difficulty level would be a good thing to give the AI, or some way of predicting what the return on investment could be.  If the quest is likely to give an item they don't need then maybe it gets devalued, but if they can get a hero, maybe it makes the risk worth taking.  

Love the battles!    Looking forward to the beta.

there anyway I could convince yus ta give us a couple more of dem battles? (I love this stuff)

could be I might throw a bit of karma your way, given such a kind gesture & all

or is this not the currency ye be dealin in?

If "some" of the AI fractions build archer focused armies I will be impressed,

and if "Intelligent" AI sovereigns have the wisdom to cast wind shield spells to protect their units from archer attacks I will be more impressed.

I can't handle the waiting. 

Frogboy - How about if we all promise not to make any posts until after the Holiday weekend if you let us have the Beta Wed night? (You'd probably have to block out Lord Xia).

Seriously, this game is looking like you guys might make everyone forget the heartache of E:WOM.

It would be nice to get the beta next Thursday, but not expecting until another three weeks.

An AI quest difficulty tag- yeah, that's probably needed.  Maybe a few tags, like type of quest- gathering quests like the Wild Goose Chase you'd want a fast unit doing, whereas battle quests you'd want a strong stack.

Yup, looking good.

Right now, the tactical battles need a lot of work.  If it were just a strategic game, I'd be pretty happy with the AI.

I don't anticipate the tactical AI being super hard to do as Charles and Co. have written a ton of new APIs for me to use in it that didn't exist in WOM.

Charles and Co.?

Charles and Co.- An indie developer within an indie developer. 

When a human player starts a quest, she can gather a lot of information from the name of the quest, the description etc to assess how difficult the quest is and if the reward would be really good or not. So it's only fair that the AI gets some difficulty/reward information about quests at the same time the human player reads the descriptions.

Perhaps some AI personalities take more risks with quests, some take less.

Also, I don't know how the quest system works but if you don't want the AI to "cheat" in this sense and it is possible, the AI can use a dummy army to check if the next quest step is a difficult battle or not.

I agree with Sir_Linque about AI personalities and quests.

Some AI's might go after only high reward quests. Some might seek out champion quests to recruit or slay champions. Some might go after only easy quests; i.e. kill rats. Others may seek to complete only magic item quests or Alliance Quests (i.e. rescue my son and my fraction will ally with yours).

Allowing players to see how hard a quest is wouldn't be a horrible idea either.

What would it be?  Percentage Difficulty?  Or like Difficulty Levels - eg. Very easy, easy, average, hard, very hard, almost impossible?  (Or something like Tech Levels required?)

I'm not sure the best route yet.  I just know that losing an army led by a champion in a quest is too damaging. I need to have a better idea of whether I should go on the quest or not.

The AI can't save and reload.  That's probably what makes this so difficult, players will take risks the AI can't.

While a smarter AI that knows it's chances of beating a quest is the best answer, you could always cheat. If the AI loses the battle you could cancel it and re-spawn the army like it never happened, or at least have the champion escape the battle. Players will retreat and cheat anyway by reloading.