Good game, but....the AI is very bad. [Archive]

proteus

May 8th, 2004, 07:44 PM

I like this game. I usually play single, thus I have some major problems with the AI.

1. The AI is massing weak troops. Why? I see hella lot of light infantry and cheap troops. Pointless, they are toast.
2. The AI is making weird strategic decisions. It is moving around with strong commanders without a point, and without a decent army.
3. The AI is not equipping its heroes properly, sometimes not at all.
4. The AI is not protecting its heroes/commanders.
5. The AI isnt using the mid/late game spells very well.
6. The AI wont summon too many "heavyweight monsters". -> I wonder what the heck the AI is doing with its magic income.
7. The AI's battlefield tactics is kinda laughable as well.
8. The AI isnt building forts, just very rarely. Maybe this is causing our next problem, that the AI is making LI mostly?

I patched the game to 2.11 of course, before you ask.
I dont wanna be rude, but all of these should be fixed with the next patch.

P.

[ May 09, 2004, 12:48: Message edited by: proteus ]

Leif_-

May 8th, 2004, 08:01 PM

Originally posted by proteus:
I dont wanna be rude, but all of these should be fixed with the next patch.You probably shouldn't hold your breath. AI is a notoriously difficult area, and the nature of Dominions makes it a very difficult game to write an AI for (basically, it's got a very large amount of different inter-releated factors which make for a much larger search-space than most comparable games.)

You might see incremental improvements to the AI, but you're not going to see the AI "fixed" in a patch. The issue is simply so complicated that to do make any huge improvements would likely take far too much work to be worthwhile.

I suggest you try a few multiplayer games instead. The human mind might not be the best computer that can be made, but it's the only one we can build with unskilled labour.

proteus

May 8th, 2004, 08:21 PM

Yes I do play multi, but you know, it takes too much time to play multi games in these type of games.
[Especially in this, in mid or late game]

I think that it is not impossible to script a good AI, in fact I am sure that all of these mentioned weaknesses can be fixed...and hopefully they will be.

P.

Leif_-

May 8th, 2004, 08:25 PM

Originally posted by proteus:
Yes I do play multi, but you know, it takes too much time to play multi games in these type of games.
[Especially in this, in mid or late game]

I think that it is not impossible to script a good AI, in fact I am sure that all of these mentioned weaknesses can be fixed...and hopefully they will be.

P. Of course they can be fixed, it's just going cost more than it's worth. Basically, what you'd end up with is a huge expert system, probably near the size of the game itself.

proteus

May 8th, 2004, 08:29 PM

Originally posted by Leif_-:
Of course they can be fixed, it's just going cost more than it's worth. I dont agree. Huge amount of players are only playing singleplayer games. They should get the best what is possible for their money.
This game costs a lot of money, like the top seller high-end games. This is why the devs should focus on this, since this is the weak part of the game right now.

P.

Leif_-

May 8th, 2004, 08:37 PM

Originally posted by proteus:
I dont agree. Huge amount of players are only playing singleplayer games. They should get the best what is possible for their money.
This game costs a lot of money, like the top seller high-end games. This is why the devs should focus on this, since this is the weak part of the game right now.

P. I think you severely underestimate the effort and work that goes into making a good AI - particularly for something as complex as Dominions. Besides, the AI in Dominion isn't really any worse than the average for strategic games (in my opinion it's surprisingly good, considering Illwinter's position). Let's face it - even for games that are created by the big hitters, the AIs are generally weak.

proteus

May 8th, 2004, 08:48 PM

Yes you are right in some points, there are a few games with a decent AI, most of them are weak.
However I dont think that it is wise to bring up things like that. Just because the majority of the strategy games have a mediocre/weak AI, why should Doms II. have a weak AI as well?
The game costs 50, which is hella lot in the gaming market, so the devs must work for their money. Lot of months have passed now since the game release, and the AI is still bad.
We had patches they were good and nice, but this problem is not solved still. I almost see no changes in the AI. Not good.....not good. http://forum.shrapnelgames.com/images/icons/blush.gif

I guess you all understand my point.
As I said I play multi as well, but I like the singleplayer too, however the quality of this AI rendering the SP part...er..not so fun.
I really hope that the devs will do something about this, because time is passing, and nothing happened yet.

[ May 08, 2004, 19:52: Message edited by: proteus ]

MStavros

May 8th, 2004, 08:56 PM

Well..wow.wow.wow...
This was being discussed months ago, it was my topic!!

Look at here! ->
http://www.shrapnelgames.com/cgi-bin/ultimatebb.cgi?ubb=get_topic;f=74;t=000191;p=20

Anyways I totally agree with proteus.

Leif_-

May 8th, 2004, 09:01 PM

Originally posted by proteus:
Yes you are right in some points, there are a few games with a decent AI, most of them are weak.
However I dont think that it is wise to bring up things like that. Just because the majority of the strategy games have a mediocre/weak AI, why should Doms II. have a weak AI as well?Because good AIs are very, very difficult to make. As an educated guess, I'd say that to make a good AI for Dominions 2, would require at least 80% of the effort that's already gone into making the whole game.

The game costs 50, which is hella lot in the gaming market, so the devs must work for their money.The game is also a niche product from a tiny, half hobby-driven development team. There are limits to how much you can expect from them.

Lot of months have passed now since the game release, and the AI still bad.
We had patches they were good and nice, but this problem is not solved still. I almost see no changes in the AI. Not good.....not good. http://forum.shrapnelgames.com/images/icons/blush.gif As I said earlier, it's difficult to make a good AI, so you can't expect to see any huge changes to the AI done in patches. We're talking about several man-months worth of work here (at the very least), and remember that Illwinter is a part-time buisness.

[ May 08, 2004, 20:03: Message edited by: Leif_- ]

Gandalf Parker

May 8th, 2004, 09:08 PM

Originally posted by proteus:
<blockquote>quote:<hr />Originally posted by Leif_-:
Of course they can be fixed, it's just going cost more than it's worth. I dont agree. Huge amount of players are only playing singleplayer games. They should get the best what is possible for their money.
This game costs a lot of money, like the top seller high-end games. This is why the devs should focus on this, since this is the weak part of the game right now.<hr /></blockquote>Im real sorry if anyone feels they arent getting their money worth. I used to hear the same thing about Master of Magic because it wasnt multiplayer. Or GalCiv (the #1 AI game by the way) because it wasnt multi-platform. Now we have Illwinter's PBEM multiplayer game because the AI isnt good enough for decent solo play. Maybe the game description needs touched up.

Kristoffer O

May 8th, 2004, 09:41 PM

Hello MStavros. http://forum.shrapnelgames.com/images/icons/icon10.gif

I find it rather amusing that you try to influence us with split personalities http://forum.shrapnelgames.com/images/icons/icon12.gif

Hopefully there will be improvements in the AI, but as before I can't give you any guarantees.

proteus

May 8th, 2004, 09:55 PM

Originally posted by Gandalf Parker:
Now we have Illwinter's PBEM multiplayer game because the AI isnt good enough for decent solo play. Maybe the game description needs touched up. Correct me if I am wrong, but isnt this supposed to be a singleplayer - multiplayer game, just like many others?
If your argument would be valid, why is there singleplayer at all? I mean if the developers dont care about the singleplayer part, they should totally drop it for dominions 3, OR try to fix the singleplayer bugs/weaknesses.

Than they wont see Posts like this..yes the
game wont sell that good, but at least I wont spend my money than. What do you think?

Yes I feel cheated for my money sorta. The game is great, but its like a half game for me.
Good mp, crap sp.
For 25$ I wouldnt say a word, but for 50, I think the devs should try to do their best!

[ May 08, 2004, 21:00: Message edited by: proteus ]

proteus

May 8th, 2004, 09:59 PM

Originally posted by Kristoffer O:
Hopefully there will be improvements in the AI, but as before I can't give you any guarantees. Yes we do hope as well. I guess you dont really care about it as I see. Sad.. http://forum.shrapnelgames.com/images/icons/shock.gif

proteus

May 8th, 2004, 10:09 PM

From the Dominions II. features page:

"Epic turn-based fantasy gameplay. Up to seventeen people can play at once thanks to the simultaneous turn structure. For those who would rather play solo the computer opponents put up fierce competition for those 3 AM gaming sessions"

So Mr. Parker?

I just checked the downloads section as well, and all of the fixes/improvements of the patches.

I havent found a singe improvement in the AI, maybe the spellcasting AI, but thats nothing.

Norfleet

May 8th, 2004, 10:12 PM

There was a mention of the spellcasting AI being improved, but by and large, I think that Illwinter considers the SP strategic AI a secondary concern at best, since even the primary game blurb tends to indicate the game is mainly multiplayer, with single player as a sort of afterthought/training mode.

Writing AI is hard! Computers are inherently stupid, and excel only at doing dumb things quickly. It's hard to make a computer do smart things, except by having it do many dumb things and pick the Last dumb of the options.

proteus

May 8th, 2004, 10:17 PM

Sorry, there was no information that the SP part is totally useless, and that it is for train and practice only, in fact read the features page. I wouldnt buy the game than.

Dont get me wrong, I understand that its hard to write a good AI, but hell, a half year passed since the release of Doms 2..or even more?!!!
Still nothing, and as I understand Kristoffer O. they havent even touched the AI!!!

I cant believe this. http://forum.shrapnelgames.com/images/icons/blush.gif http://forum.shrapnelgames.com/images/icons/shock.gif

PDF

May 8th, 2004, 10:21 PM

While I agree that fixing the "strategic" AI, or even the combat AI, would be a big endeavour, I'm also pretty confident that the "AI builds crap armies" thing *could* be fixed quite easily, with only changes into the army build routines. The AI should weight its decisions mostly towards "heavy" units and discard light ones after a certain (low) number.
This alone will make the solo games *much* more challenging, at least we'll have to fight "real" armies, rather than an half dozen of worthwhile units lost in an ocean of militias ... http://forum.shrapnelgames.com/images/smilies/rolleyes.gif

Graeme Dice

May 8th, 2004, 10:29 PM

Originally posted by proteus:
Dont get me wrong, I understand that its hard to write a good AI, but hell, a half year passed since the release of Doms 2..or even more?!!!It took most of the Last century to develop an AI that could play chess competently. I highly doubt that even five or six years of work could do the same for a game as complex as Dominions. You are asking the devs to do the impossible, so it's not particularly likely that it will happen.

You might also want to get a new argument MStavros, since changing your alias and using the same text word for word doesn't change your argument.

[ May 08, 2004, 21:33: Message edited by: Graeme Dice ]

IKerensky

May 8th, 2004, 11:05 PM

Hum I think that proteus have several valid points.

And I DO think that there is easy way to improve AI behaviour is THOSES peculiar points.

First thing will be to have AI template built for each nation giving it a "model" ( better several for less randomness ) army it will try to emulate, thoses modesl including battlefield scripting.

Spells have to be classified by types and ranked in value thus the Ai can script the correct chaining.

For the more strategic point... then we get the weakest link of computerised AI and I guess we will have to use a sorta frontline AI, one wich will abstract the battlefield into a series of nodes controled by entities with the AI aiming at expanding his net of nodes favoring direction by calculation ( indies first, 1 foe, Last province lost... ). All of this is pretty complicated.

Pretender/commander equipment should also be scripted for the AI with a selection by level of CONS and kind of commander and/or current foe. Difficult but possible for a full time payed developper to come with.

To sum it all, I guess the shortest way to improve the DOm2 AI is to give it some recipes on how to play and bind it to that. It will be pretty predictable by nation ( but that's not surprising ) but make a far better job at playing one logically.

st.patrik

May 8th, 2004, 11:09 PM

Originally posted by Kristoffer O:
Hello MStavros. http://forum.shrapnelgames.com/images/icons/icon10.gif

I find it rather amusing that you try to influence us with split personalities http://forum.shrapnelgames.com/images/icons/icon12.gif

Hopefully there will be improvements in the AI, but as before I can't give you any guarantees. ah! I wondered from the tone of the Messages if it might be MStavros again.

MStavros/Proteus: wasn't once enough? do we really have to go through all the tired old rants again?

Kuroth1

May 8th, 2004, 11:25 PM

Originally posted by Gandalf Parker:
<blockquote>quote:<hr />Originally posted by proteus:
<blockquote>quote:<hr />Originally posted by Leif_-:
Of course they can be fixed, it's just going cost more than it's worth. I dont agree. Huge amount of players are only playing singleplayer games. They should get the best what is possible for their money.
This game costs a lot of money, like the top seller high-end games. This is why the devs should focus on this, since this is the weak part of the game right now.<hr /></blockquote>Im real sorry if anyone feels they arent getting their money worth. I used to hear the same thing about Master of Magic because it wasnt multiplayer. Or GalCiv (the #1 AI game by the way) because it wasnt multi-platform. Now we have Illwinter's PBEM multiplayer game because the AI isnt good enough for decent solo play. Maybe the game description needs touched up. <hr /></blockquote>Gandalf I am confused... Are you saying that the SP game of DomII IS BAD because of Bad AI and they should change the Description to read PBEM only???

Gandalf I had been meaning to ask you a question and now I guess is a good time... I think I read one time that you play SP 95% of the time.... My questions is what do you think of the AI??? Is the game fun and good as a SP game????

I am just asking because I always respect your opinion...

Thanks!

Gandalf Parker

May 8th, 2004, 11:29 PM

Originally posted by proteus:
Sorry, there was no information that the SP part is totally useless, and that it is for train and practice only, in fact read the features page. I wouldnt buy the game than.

Dont get me wrong, I understand that its hard to write a good AI, but hell, a half year passed since the release of Doms 2..or even more?!!!
Still nothing, and as I understand Kristoffer O. they havent even touched the AI!!!

I cant believe this. http://forum.shrapnelgames.com/images/icons/blush.gif http://forum.shrapnelgames.com/images/icons/shock.gif What cant you believe about it? you are talking about what? 1 patch? 2 patches?

Im mostly a solo player but I dont expect them to spend alot of time upgrading a part of the game that isnt mainline to what they created.

The fact that some improvements are coming MIGHT be due to the fact that some players finally started doing test runs.

JaydedOne

May 8th, 2004, 11:49 PM

Originally posted by proteus:
Yes we do hope as well. I guess you dont really care about it as I see. Sad.. http://forum.shrapnelgames.com/images/icons/shock.gif Constructive criticism citing that your topmost interest as a player for future patches is an AI improvement (and offering a few suggestions on how to go about it) is one thing. Flat-out Knocking a developer who has shown more support for his community than all but a handful of his better-paid, full-time colleagues is another.

Guess which one is more likely to garner respect and a proactive response from both devs and fellow community members.

[ May 08, 2004, 22:50: Message edited by: JaydedOne ]

Gandalf Parker

May 8th, 2004, 11:50 PM

Originally posted by Kuroth1:
Gandalf I am confused... Are you saying that the SP game of DomII IS BAD because of Bad AI and they should change the Description to read PBEM only???Well Illwinter has never made any great claims about the AI. What was quoted was from the Shrapnel site. (they are the game production company. they are supposed to market it)

And yes, I think the description would be easier to fix than the AI for now. Im sure its on their list but I dont know what would have been left out of the Last couple of patches to give it to us. The speed control in battles? (I can hear he reaction to THAT)

Gandalf I had been meaning to ask you a question and now I guess is a good time... I think I read one time that you play SP 95% of the time.... My questions is what do you think of the AI??? Is the game fun and good as a SP game????I would love to see it improved. I have no problem with asking. Im thrilled when I see Posts that actually try to pin down and test some suggested improvements. Its worked for me and Ive been making suggestions since the first days of Dom1. In my own small way Ive also tried to improve it thru maps. (www.dom2minions.com)

What I dont like is to see people saying its broke, making it sound like some falsified rip-off, demanding work be done on it, and generally turning it into something that the devs dont want to touch (well one dev programmer actually). Its not the subject I disagree with, its the tactics. Not only do I think they wont work well, I think they go in the opposite direction.

[ May 08, 2004, 22:54: Message edited by: Gandalf Parker ]

guybrush threepwood

May 8th, 2004, 11:55 PM

Hi,

First, I certainly think that every krone spent on this captivating game has been spent well!

And I certainly understand that the AI can be an extremely difficult thing to come up with good algorithms for.

But I just wanted to point out that improvements to the AI would certainly be highly appreciated by many people. If it is possible to find the time to do it of course. http://forum.shrapnelgames.com/images/icons/icon7.gif

Anyway, I find comments about "not getting what you paid for" hard to comprehend.

Cheers,
Thomas

proteus

May 8th, 2004, 11:56 PM

1. I am NOT MStavros he is my older brother, and we are using one computer.
He introduced Dominions II. to me, and you know what?? We bought 2 copies not just one !!!!. I am playing multi games with him usually.
I didnt knew about his old topic, so Ive started a new, after we had a discussion about the AI at home.

I am a kinda new D2 player, I have like 2 months of experience, and I love the complexity of it.

I think that this is positive critism, isnt it??
Scripting must be hard I am totally sure, but this is nonsense to me still. Half year and no AI improvements....definitely not a good way to raise the 'fame' of IW.

Mr Parker, if you are a solo player, I simply cant understand your points at all. Even I can beat the crap out of this AI easily, and I am a newbie compared to you.
So what is the fun for you in the SP than?

IMHO the devs should make a statement, that they wont raise the quality of the AI, and people like me, who would like to enjoy the SP part as well will shut up once and for all, and play lot less with Doms II. [Only MP if we will have time for it.]
However I really hope that the devs wont forget about the lot of SP fans...eh well for 6 months they did actually.

You know I am reading that old topic, what my brother mentioned and all of those hella old problems with the AI are still not solved.
Perhaps IW should hire a new coder, if the current coder doesnt have the time to upgrade the AI...because it seems so.. http://forum.shrapnelgames.com/images/icons/confused.gif

Norfleet

May 9th, 2004, 12:03 AM

You seem to believe that AI coding is easy, and that you can just "hire another coder".

I'm sorry to inform you that this is not the case. AI coding is damn hard, and even if you hire some overpaid code monkey, there's no guarantee he won't produce something which sucks even more. Name ONE game that has truly had good AI. Then tell me what its production budget was.

Do you think Illwinter can afford that?

[ May 08, 2004, 23:03: Message edited by: Norfleet ]

proteus

May 9th, 2004, 12:05 AM

Originally posted by Gandalf Parker:

What I dont like is to see people saying its broke, making it sound like some falsified rip-off, demanding work be done on it, and generally turning it into something that the devs dont want to touch (well one dev programmer actually). Its not the subject I disagree with, its the tactics. Not only do I think they wont work well, I think they go in the opposite direction. I apologize I never wanted to be 'demanding'. Simply this is my opinion.
I respect IW and their work, the game is really good. The only thing what I dont like, that the Doms II. AI improvement idea is totally dropped as I see.

A dev comment that it might be fixed...however they cannot guarantee it after lot of months..made me pissed.
So sorry for the tone of my Posts, but that made me very angry.

proteus

May 9th, 2004, 12:12 AM

double post sorry. http://forum.shrapnelgames.com/images/icons/icon7.gif

[ May 08, 2004, 23:12: Message edited by: proteus ]

proteus

May 9th, 2004, 12:12 AM

Indeed the game is complex -> hard to make a good AI.
there are lot better AIs for various games, take a look at HoMM3 for example.
It is not as complex that is true, but there the comp is making huge and strong armies, its really decent. Also it is using them very well.

Just go and play with it on harder settings you will have hard days, I guarantee it. http://forum.shrapnelgames.com/images/icons/icon7.gif

The Doms2 AI is hella weak compared to the homm3 AI for example. These little light infantry armies and the various other AI problems wont offer a little challenge at all.
I guess -as someone said- lot of people would be thankful for AI improvements, lot of people bought the game to play and enjoy the SP part as well, the game was advertised as a sp-mp game.
The core game had a bad AI, but why do we have a bad AI after a half year still? THIS is the real question here.

[ May 08, 2004, 23:13: Message edited by: proteus ]

Graeme Dice

May 9th, 2004, 12:14 AM

Originally posted by Norfleet:
Name ONE game that has truly had good AI.Other than games where a decision tree can be mapped out for all possible combinations of moves like Chess, I don't think that such a thing exists anywhere.

Gandalf Parker

May 9th, 2004, 12:18 AM

There have been plenty of people here talking about the AI. Some of them program at various levels and have said that it would not be easy, simple, or quick. A single-threaded AI is going to be a hard project to patch. And other methods would be a major rewrite (scripted AIs are a great thing but everything in the game has to have a script-handle written into the code for it to work)

Thre are some threads that are helping the subject along. Do a search for
SUGGESTION: AI castle building algorithm
SUGGESTION: AI troop building algorithm

both of which need people who are wiling to test their ideas and post results.

MStavros

May 9th, 2004, 12:24 AM

Well I am here now. http://forum.shrapnelgames.com/images/icons/icon6.gif
I am mainly playing multip. games now, unlike my bro, but this is just because the AI is so crap still.
After that huge discussion in december (20 pages long thread about the AI), even I am annoyed a bit, that after all these patches the AI is the same. This means one thing. The developers can't be bothered about the AI. You don't like these opinions? YOu have the right, but don't forget it seems so that they can't be bothered about it. Is it good? Not at all!
Personally this is my favourite game, but Illwinter is not my favourite gaming company. http://forum.shrapnelgames.com/images/icons/icon12.gif

Gandalf Parker

May 9th, 2004, 12:25 AM

Originally posted by Graeme Dice:
<blockquote>quote:<hr />Originally posted by Norfleet:
Name ONE game that has truly had good AI.Other than games where a decision tree can be mapped out for all possible combinations of moves like Chess, I don't think that such a thing exists anywhere. <hr /></blockquote>Actually Galactic Civilizations would be fairly ultimate in mentioning AI. In conversations in the real Artificial Intelligence sites it gets mentioned. Its still fairly new to the Windows envirnment but still keeps alot of its abilities.

But there is a big difference between building an AI and adding a game to it, versus building a game then adding an AI for testing and soloplayers.

[ May 08, 2004, 23:26: Message edited by: Gandalf Parker ]

proteus

May 9th, 2004, 12:27 AM

Mr. Parker, the galciv AI is really good, I must agree, but the homm3 AI is even better.

The comp rocking there. Sometimes it is making so huge and powerful armies that I need all my skill to beat it somehow with nasty tricks!
Sometimes I even loose against the AI.
Hardcore strategy fans should give it a try, its a huge challange. http://forum.shrapnelgames.com/images/icons/icon7.gif

In dominions 2. I win all games against 1-8 comps on all settings.
This tells about everything.

[ May 08, 2004, 23:29: Message edited by: proteus ]

Molog

May 9th, 2004, 12:31 AM

One easy "improvement" would be to give the AI cheating powers.
Add a new level super-impossible, where the AI pays only 25% upkeep, has a 5% chance to have a castle build in a random province and gets a 50% research bonus.

This would satisfy the people who want a challenge and is relatively simple to do.

MStavros

May 9th, 2004, 12:33 AM

mmmmmm, I had an idea in the past, regarding these LI AI armies.
Maybe a mod what removes all LI from the game should help in the 'weak LI AI armies' problem?
I wonder what would the AI do without LI...maybe it would make kickass armies. http://forum.shrapnelgames.com/images/icons/icon7.gif

This wouldn't solve the other problems, but maybe one! [weak LI AI armies.]
Any opinions about this?

MStavros

May 9th, 2004, 12:34 AM

I hate cheating AIs. http://forum.shrapnelgames.com/images/icons/icon7.gif .. thats the Last thing what the devs should do if you ask me. http://forum.shrapnelgames.com/images/icons/blush.gif

Molog

May 9th, 2004, 12:36 AM

I know AOW, MOM have horrible cheating AI's. Most players don't seem to mind though.

Vynd

May 9th, 2004, 12:46 AM

I like Dom II a lot, and I think that SP is fun. But difficult? No. Would I like to see the SP AI improved? Yes! Do I think its going to happen? No. But do proteus and his brother and whoever else have the right to complain about it? Heck yeah!

proteus

May 9th, 2004, 12:47 AM

The HoMM3 AI is not cheating for example, and I cant tell you with words how good it is. http://forum.shrapnelgames.com/images/icons/icon7.gif
Try it and you will see.

Kristoffer or any other devs...can you tell us the content of the next patch?
What is what you are going to add/fix?
If I am correct we still wont have AI improvements, that wasnt hard to figure out, eh. http://forum.shrapnelgames.com/images/icons/tongue.gif
If not now, with the second patch from now, please try to improve the AI, and solve the problems step by step. http://forum.shrapnelgames.com/images/icons/icon7.gif

Kristoffer O

May 9th, 2004, 12:56 AM

Originally posted by proteus:
A dev comment that it might be fixed...however they cannot guarantee it after lot of months..made me pissed.
So sorry for the tone of my Posts, but that made me very angry. Perhaps all the pointless complaining made us pissed.

There was some constructive AI discussions recently that made me interested, but I'm quickly loosing interest. Right now I'm more interested in doing something I am feeling positive about like improving Pangaean themes or finishing the Shepherds of Creation scenario, or for that matter work at school.

The reason I do not say that there will not be an AI improvement is because we also feel that the AI should be improved.

Graeme Dice

May 9th, 2004, 12:59 AM

Originally posted by proteus:
In dominions 2. I win all games against 1-8 comps on all settings.
This tells about everything. Actually, that tells us nothing, since you aren't discussing the nation you are playing, any of the game settings, which map you're using, and which opponents you've selected.

Kristoffer O

May 9th, 2004, 12:59 AM

Originally posted by proteus:
The HoMM3 AI is not cheating for example, and I cant tell you with words how good it is. http://forum.shrapnelgames.com/images/icons/icon7.gif
Try it and you will see.

Kristoffer or any other devs...can you tell us the content of the next patch?
What is what you are going to add/fix?
If I am correct we still wont have AI improvements, that wasnt hard to figure out, eh. http://forum.shrapnelgames.com/images/icons/tongue.gif
If not now, with the second patch from now, please try to improve the AI, and solve the problems step by step. http://forum.shrapnelgames.com/images/icons/icon7.gif No AI improvenents as of now, but a couple of balance changes that has been topics lately. Can't remember exactly.

PrinzMegaherz

May 9th, 2004, 02:03 AM

This discussion is pointless.

The AI is as good as can be expected from a 2 man team. And I surely would not say that it is that bad. It is definitly a match for new players, and if I want a very challenging game then I play Mictlan (the nation I suck most with) against difficult Ermor.

The main problem with doing the AI is the following: You have to work out solutions for every possible situation, even those you have never encountered yourself, or situations where you actually don't know what to do. This might be fairly easy by such a simple game as heroes of might and magic, where you can calculate your decission down to a few parameters... however dominions 2 is so much more complicated with so many aspects to consider that the devs would surely need another ten years or so to code every smart solution.

What Illwinter actually did, very smart in my oppinion, is to take advantage of the computers strengths. Even today people discuss about the perfect army, the perfect strategy, and there is so much knowledge around here the AI could never match. Instead the AI uses some options that nearly every human would be too lazy to employ: It does the maximal recruitment possible out of every province and assemble huge armies.
Indeed, in most of my SP games and MP games involving AIs, they will leave everyone far behind in sheer numbers. Even tough their armies might not be the best designed, they are still a threat nontheless (in my current MP game, one of the more peacefull but most powerfull players dropped from the game. When the AI took over, its onslaught nearly brought 3 of his neighboring empires to its knees.

Imho the dominions 2 AI is comparable to that of Warcraft 3. Demanding at first, its not a match for an professional player.

Gandalf Parker

May 9th, 2004, 02:20 AM

Originally posted by proteus:
The HoMM3 AI is not cheating for example, and I cant tell you with words how good it is. http://forum.shrapnelgames.com/images/icons/icon7.gif
Try it and you will see.Ive played HoMM3 and you are correct. But its on a far less level of problem. Less units, less equip, less unit commands. But thats OK I can see the point.

The AI for Dom2 can be improved, has been improved, but will never be improved enough to not have these kinds of threads. If we get it to build LI at a level which works better for some nations then it will cause problems for others. If we get it to stop casting certain spells, or cast some more, or build sensible equip, or... well you see where its headed. There will always be more AI problems in a game like this one.

Kristoffer or any other devs...can you tell us the content of the next patch?
What is what you are going to add/fix?
If I am correct we still wont have AI improvements, that wasnt hard to figure out, eh. http://forum.shrapnelgames.com/images/icons/tongue.gif Against my better judgement (as if that mattered) it appears that some responses to other loud threads here are getting addressed. But it does make sense that its multiplayer things that are loudly being proclaimed "broken".

And also a request of mine looks like its getting added (dont get mad Ive been requesting longer than you have)

If not now, with the second patch from now, please try to improve the AI, and solve the problems step by step. http://forum.shrapnelgames.com/images/icons/icon7.gif He mentioned that a couple of threads had generated interest. Threads where Users discussed fixes and seemed willing to test them.

Gandalf Parker

May 9th, 2004, 02:29 AM

Originally posted by Vynd:
I like Dom II a lot, and I think that SP is fun. But difficult? No. Would I like to see the SP AI improved? Yes! Do I think its going to happen? No. But do proteus and his brother and whoever else have the right to complain about it? Heck yeah! Well I was with you right up to the end. As far as their right to complain, ok. Doesnt that mean others have a right to say "shut up" if its hurting our chances of getting the AI fixed?

And actually I wouldnt use "right to complain" too often on a board that is owned by someone and provided for a specific reason. Doing damage to the reason can wash those "rights" away fast. http://forum.shrapnelgames.com/images/icons/icon7.gif This isnt usenet newsGroups.

[ May 09, 2004, 01:41: Message edited by: Gandalf Parker ]

delacroix

May 9th, 2004, 03:29 AM

Originally posted by Molog:
I know AOW, MOM have horrible cheating AI's. Most players don't seem to mind though. I certainly don't mind, Master of Magic / Orion were funnest and toughest at the impossible settings even though the AI receieved significant bonuses at those levels http://forum.shrapnelgames.com/images/icons/icon7.gif .

Certainly if a player didn't want the CPU opponents to cheat they could just play at normal difficulty instead.

Leif_-

May 9th, 2004, 03:31 AM

Originally posted by Vynd:
But do proteus and his brother and whoever else have the right to complain about it? Heck yeah! This is going to be a rant, just so that you're warned:

There's complaining, and there's complaining incessantly, ignoring the explanations and arguments against their complaints.

They certainly don't have a right to demand that Illwinter does this or does that, which is what they are doing. Unlike what they seem to think, computer games are bought as-is. Caveat emptor. Any support ontop of what you get in the box is a bonus and a matter of good service, rather than something you're entitled to.

In my opinion Illwinter has done a great job supporting the community around the game - remember that for most other games "patch" usually means "bug fixes" rather than "game improvements" which is what we've gotten used to. That makes it especially grating to see them harp upon Illwinter as if Illwinter hadn't given the community any kind of support or service at all, just because Illwinter hasn't prioritized their unreasonable requests.

So they paid $50 for the game. So what? The Dominions AI is not any weaker than what's common for full-price strategy games, so it's not like they could really expect something better for that price anyway. Sure, there are games out there with better AIs, but then it's not reasonable to compare Dominions 2 only with the best AI games out there.

They are also both still playing Dominions 2 after several months, so they can hardly argue that they didn't get their monies worth or that the AI makes the game unplayable in single player.

So, no. I don't think they've really got a right to complain. They've gotten what they paid for, and more; they have been explained why it's unreasonable to demand heavy improvements in the AI; they've been told that Illwinter doesn't have the resources a major software house has, and they've been told that the Dominion AI isn't really weaker than what's standard for similar games. Yet still they persist in their complaining and, frankly, whining.

mlepinski

May 9th, 2004, 04:04 AM

I'll certainly agree that the dominions 2 AI isn't perfect, but it's not nearly as bad as threads like this would indicate.

In particular, it is definately possible to play a challenging game against the AI, you just need to adjust the map to put the AI in an advantagous position.

I find the most significant AI shortcoming to be the fact that the AI doesn't build forts. (The AI builds much more effective armies when it has multiple forts to recruit troops from). Therefore, I often play games where I start the AI with 2 forts (or even 3). Additionally, I'll stick some extra independent forts on the map for the AI to conquer.

In my opinion, playing a single game against 4 pairs of allied 'Difficult' AIs (8 AIs total) each starting with 2 forts on a map scattered with independent forts is an interesting single-player experience which is much more challenging than a 'standard' single-player game of Dominions.

So, even though the AI isn't perfect, I am satisfied because it's good enough to give me a challenging game under the right circumstances.

Just my 2 cents,
- Matt Lepinski :->

Wikd Thots

May 9th, 2004, 05:06 AM

WIMPS!
You payed for the game. What diff if its one poor programmer? You can demand anything you want. I want to control units in fights like my other games. I want to save games so I can load if things are crap. I want cheat codes. And I want to play with my friends if we buy one disc.
I want all of that and I want it NOW because it should already be in it like the other games that cost this much!

Darryl

May 9th, 2004, 06:15 AM

Ok, I generally lurk around here but I just have to say something here. I find the topic of this thread to be completely false and the way the concerns are given to be completely offensive. Point blank. To say that...

Originally posted by proteus:

A dev comment that it might be fixed...however they cannot guarantee it after lot of months..made me pissed.
So sorry for the tone of my Posts, but that made me very angry. Pisses me off. Yes, I know you apologized for the tone, but it's the concept that you were pissed because something you wanted wasn't improved in the game as if they owe it to you to do that. You paid $50, so they should improve the AI (probably the HARDEST aspect of the programming) and do it NOW. I mean it's been 6 months, right? And you're pissed? They actually show up on the Boards and implement player suggestions that have nothing to do with bug fixes (like modding) and you're pissed off and feel like you've been robbed? You can get these changes without buying the "Dominions II Gold Pack" and you're upset because your changes weren't there? Like Illwinter/Shrapnel have "misled" you somehow? Please!

I think Illwinter is a better name than "two guys with computers and another guy poking his head in every so often" but as I understand it that's what it is. And these 2 guys do what major companies have been unable to AND respond to fans directly and you're angry. Not "wishing it would change", but "angry".

Now the very idea that someone has been gipped by buying this game is ludicrous, and to even hint at a suggestion that they "falsely advertised" the AI is beyond words.

Now I know something about programming (I've worked as a software engineer and, more than that, I did the Online grimorie http://forum.shrapnelgames.com/images/icons/icon12.gif ) and I can say that AI programming is difficult. To insinuate that "well it's easy just weight and have them do such and such" is insulting, as if they're in Tahiti drinking magaritas with their ill-gotten Dominions funds when they should take the 10 minutes to improve the AI. I mean how many games are supposed to provide a significant challenge after playing them for a few months?

Sorry for the rant, but I just think this is by far The Best Game I've Played bar none. I also think the devs are extremely accessible and responsive, and to suggest they "cheated" someone is completely out of control to me.

Slygar

May 9th, 2004, 07:52 AM

Originally posted by delacroix:
<blockquote>quote:<hr />Originally posted by Molog:
I know AOW, MOM have horrible cheating AI's. Most players don't seem to mind though. I certainly don't mind, Master of Magic / Orion were funnest and toughest at the impossible settings even though the AI receieved significant bonuses at those levels http://forum.shrapnelgames.com/images/icons/icon7.gif .

Certainly if a player didn't want the CPU opponents to cheat they could just play at normal difficulty instead. <hr /></blockquote>I agree with this.. let the AI cheat. Gem, income bonuses, randomly appearing forts, permenent +3 luck, +300 design points (for scales), whatever. Only on higher difficulty levels, of course.

Heh, a neat idea might be to have a moving bonus to AI scales, meaning that at normal difficulty all their scales go with the normal AI creation routines. At hard, they get all scales as normal, but with a +1 bonus or something? +3 at Impossible?

[ May 09, 2004, 06:54: Message edited by: Slygar ]

Norfleet

May 9th, 2004, 08:50 AM

Originally posted by Slygar:
I agree with this.. let the AI cheat. Gem, income bonuses, randomly appearing forts, permenent +3 luck, +300 design points (for scales), whatever. Only on higher difficulty levels, of course.The AI already DOES get bonus design points at higher difficulties. There's no income or gem bonus, however, above what it gains from having superior scales due to bonus points and better searching due to more magic from said points.

johan osterman

May 9th, 2004, 09:15 AM

Originally posted by Wikd Thots:
WIMPS!
You payed for the game. What diff if its one poor programmer? You can demand anything you want. I want to control units in fights like my other games. I want to save games so I can load if things are crap. I want cheat codes. And I want to play with my friends if we buy one disc.
I want all of that and I want it NOW because it should already be in it like the other games that cost this much! You got what you paid for. You had every opportunity to check the demo out before you coughed up the money. The things you complain about are design decisions, you might as well demand that your chessmaster game should have a spurts of blood and cute battle sounds, or that you should be able to tactically control a battle that played out whenever two chess pieces enter the same square. There is no law that point by point details what features a game should have. Had dom 2 been marketed as having these things you demand you would have a case, but it doesn't. You can make any demands you like, but unless they are reasonable noone is going to pay attention to them.

Norfleet

May 9th, 2004, 09:31 AM

JO, I hope he was being sarcastic. Because otherwise he needs to be taken out back and given a sound beating.

johan osterman

May 9th, 2004, 09:51 AM

Originally posted by Norfleet:
JO, I hope he was being sarcastic. Because otherwise he needs to be taken out back and given a sound beating. You might be right. If you check his profile he claims one of his interests is trolling, and his occupation is irritant. I guess I swallowed his hook and bait. Don't click on his hompeage whatever you do though.

Norfleet

May 9th, 2004, 09:57 AM

Ah, yes, his homepage. I've heard about that one before. I take it you clicked on it? http://forum.shrapnelgames.com/images/icons/icon7.gif

johan osterman

May 9th, 2004, 10:34 AM

Yup

proteus

May 9th, 2004, 11:55 AM

Yes, yes I understand what you say of course.
I never demanded anything imho, I only posted my opinion.
Now a question for the developers: Dont you think that hella LOT of players would be thankful for the AI improvements?? I am totally sure that hella LOT of players are waiting for AI improvements, they just _dont_say_it_ here, and why? Propably because they know the anser from the devs:

Originally posted by Kristoffer O:
Perhaps all the pointless complaining made us pissed.
Is this really pointless??? I am totally sure that this is far from pointles!!

Originally posted by Kristoffer O:

The reason I do not say that there will not be an AI improvement is because we also feel that the AI should be improved. The devs know that the AI is weak. I am pleased to see that they know this at least.

I guess my first post told you about the AI weaknesses:

1. The AI is massing weak troops. Why? I see hella lot of light infantry and cheap troops. Pointless, they are toast.
2. The AI is making weird strategic decisions. It is moving around with strong commanders without a point, and without a decent army.
3. The AI is not equipping its heroes properly, sometimes not at all.
4. The AI is not protecting its heroes/commanders.
5. The AI isnt using the mid/late game spells very well.
6. The AI wont summon too many "heavyweight monsters". -> I wonder what the heck the AI is doing with its magic income.
7. The AI's battlefield tactics is kinda laughable as well.

I wasnt constructive? I dont think so.

Maybe my tone wasnt right, I am moderating myself now. http://forum.shrapnelgames.com/images/icons/icon7.gif
Still I am totally sure that everyone know that the AI weakness is a big problem, even if they dont say. And yes the Doms 2. AI is lot weaker than most of the 50$ strategy games.
I know that 2 men cannot make everything, but 6 months wasnt enough to improve the AI? After all those huge old threads about the AI??
This is what I cant understand.

I have seen an old poll about the sp-mp community here, and the poll ended up that we have more singleplayers! I think, that spending time to improve the AI is not a waste, in fact it should have priority on the 'to do list.' http://forum.shrapnelgames.com/images/icons/icon7.gif .. especially now, because after 4 patches we still wont have AI improvements. http://forum.shrapnelgames.com/images/icons/icon9.gif

Kris and the devs -> I never wanted to offend you, I wanted to be creative, I posted the bad parts of the AI, what should be improved, so I dont think that this was pointless.

P.

[ May 09, 2004, 11:07: Message edited by: proteus ]

Murph

May 9th, 2004, 02:44 PM

If they took out the light infantry, you'd post the same thing again on a different thread, but without the first problem.

I've been playing computer games for about 15 years, which is pretty much as long as computer games have been around, and I've always been a strategy fan. Dom 2 doesn't have the best AI of all the games I've played. There, I said it, and you said it and he said it and it's been said before.

The fact that they HAVE an AI is pretty impressive. And now you don't want it to cheat? For god's sake, get a grip. Illwinter is possibly the most responsive dev group I've ever seen, and I bet they're not exactly rolling in the dough. I was happy to have my 50 bucks go to some guys who work hard on their game, and not have most of it go to marketing, dynamic lighting effects and impressive packaging.

This game is more bang for the same buck and the money goes to the people who actually did most of the work. For 50 bucks (actually for me it more like 80, because I'm Canadian) I got more than my money's worth, I'm I mostly play SP.

I'm sure they will improve the AI, if they can, and now that people are pitching in to help it should go faster. 6 months is no great shake of time on a game this replayable, and I for one feel that all the effort we're putting into rebutting you would be better spent modding, playing or coding to improve an already great game.

You've got a right to complain, but put your time where your mouth is, and do something to improve it, don't slag others who are.

proteus

May 9th, 2004, 03:17 PM

I cannot do anything to improve the AI, just like you cannot do antyhing, this is the coders job.
I had a talk with my brother, he will make a mod and remove all LI, to see that what will the AI do than.

Maybe it will be lot better for SP games until the AI will be fixed, this is the onlyu thing what we can do. http://forum.shrapnelgames.com/images/icons/icon7.gif

Gandalf Parker

May 9th, 2004, 03:53 PM

Originally posted by proteus:
I guess my first post told you about the AI weaknesses:Im going to answer this one in 2 stages. I will get to your 7 points next.

I wasnt constructive? I dont think so.

No not really. Saying somehing is broke is criticism. Suggesting a way to improve it, in a way that might get the other person to look at it, thats constructive.

Maybe my tone wasnt right, I am moderating myself now.
Thank you

Still I am totally sure that everyone know that the AI weakness is a big problem, even if they dont say.

Everyone knows its weak when pitted against a player who knows the game. Not everyone feels its a problem. Or at least not everyone puts it as high on the list of things to improve.

I know that 2 men cannot make everything, but 6 months wasnt enough to improve the AI? After all those huge old threads about the AI??
This is what I cant understand.

Did you read the long list of things that went into each patch?
http://www.illwinter.com/dom2/history.html
Did you see the long threads about those things? (did you notice the difference in the tone of those threads which did get put into the patches?) The battle replay speed was one which had many favorite Dom1 players refusing to join us. That was a feature which involved both the MP and SP Groups. The answer involved alot of different suggestions on how to "fix" it and the final one was abit of a compromise. Not as fancy as some wanted but it was one that could be fit into a patch and has made ALOT of people very happy.

I have seen an old poll about the sp-mp community here, and the poll ended up that we have more singleplayers!

On a game that was built around "fun multiplayer game" you thought that pole said to "do more soloplayer stuff"? Dont you think it could also have meant "more multiplayer improvements"?

I think, that spending time to improve the AI is not a waste, in fact it should have priority on the 'to do list.' .. especially now, because after 4 patches we still wont have AI improvements.

4 patches? And I really wish you would stop saying that. Just because the AI doesnt kick your tail its not the same as "still no improvements". I see AI improvements listed in both patches, and many things that were improvments to both AI and Players though often more for the AI (such as friendly fire)

Kris and the devs -> I never wanted to offend you, I wanted to be creative, I posted the bad parts of the AI, what should be improved, so I dont think that this was pointless.

Posting your feelings are fine but if someone says that posting the "bad" parts of the AI and what should be improved was pointless, maybe they mean because you posted the same list as before? The only point you seem to have brought up by posting it again was to be mad that you had to post it again. You make it sound like it was ignored. As if the devs werent aware of it. You wanted to know why they hadnt fixed your wishlist yet.

Joining in on the threads which had actually made progress on the subject might have been better.

Gandalf Parker

May 9th, 2004, 03:57 PM

Originally posted by proteus:
I guess my first post told you about the AI weaknesses:This has been done often in other threads, and better than I can do it, but lets go over these again....

1. The AI is massing weak troops. Why? I see hella lot of light infantry and cheap troops. Pointless, they are toast.

There is only 1 AI and it plays the same in small or large games, low indepts or high, low magic or high, and any nation. LI are not pointless. They are only pointless to certain nations in certain games. If you think their purpose is to beat up your hvy troops then play some more.

If you REALLY feel this can be improved then show us a forumula. What type of troop, when, how many, what percentage of the formula? Write us the AIs code so we all can test it to see if it holds up in more different game situations than the one now.

If you check out the threads I mentioned then you would see that there is alot that players can do on this even if they are not coders. Do a search for...
SUGGESTION: AI castle building algorithm
SUGGESTION: AI troop building algorithm

2. The AI is making weird strategic decisions. It is moving around with strong commanders without a point, and without a decent army.

Like much of this, this is an opinion stated as a fact. If you are going to work with quantity over quality then random movement is best. It holds up against human players much better than smart movements.

3. The AI is not equipping its heroes properly, sometimes not at all.

Another area where a small effort on your part to come up with a formula would have kept you from making statements like how easy these would be to fix. Get specific and lets test it.

4. The AI is not protecting its heroes/commanders.

Probably true. I think it does a better job when they are at home, and probably when they are first sent out. But eventually the flak troops die off and you are left with just the big ones. Especially bad if thats the AIs god.

Im not sure how well the path-finding works for something like a "go home" plan. Especially since the AI does not store information well for multi-turn actions so each turn it would have to re-decide that the piece needs to go home.

Maybe a "stop and build army if less than 100" command or something so it can boost up with whatever indepts are at that province but that would end up with "crappy stupid armies" again.

5. The AI isnt using the mid/late game spells very well.
Same problem as equipment. Come up with a formula. One of the reasons that AI's in other games can do better is that the choices are less and more obvious. The AI does cast spells, it does select what it casts, youjust think it could be improved. But would your improvement work for all nations in all games without alot of "if this and if that"? (and how well would you take it when somoene says your way is stupid and broken?)

6. The AI wont summon too many "heavyweight monsters". -> I wonder what the heck the AI is doing with its magic income.

Since its a turn-based PBEM game its hard to write in multiple turn planning. Saving up for things is difficult to write. Thats true of things like castles, temples, major spells, major summonings, major equipment. If an AI has 50 gems and needs to decide what to do with them then it will decide to make 5 castings of a 10 gem summoning instead of save up for a 75 gem summoning. To decide any different would have it saving up gems early in the game when it should have been doing small summonings.

7. The AI's battlefield tactics is kinda laughable as well.
Actually Ive seen alot of improvement in this. It now uses more formations and positioning on the battlefield. Unless I use scouts to watch their battles as they approach me I can have my strategys ruined by the formations they use. Of course such surprises are easy to beat if you know they are coming but thats alittle hard to fix in an AI.

[ May 09, 2004, 14:59: Message edited by: Gandalf Parker ]

proteus

May 9th, 2004, 04:24 PM

Thanks for your detailed reply Mr. Parker.
I think that you are right in many things, we should come up with detailed ideas. Propably the devs know about these AI probelms as well...
Castle & Troop building algorythm is what we propably need first of all.
My idea is, that the devs should code the AI to build LOT Less LI units, concentrate on summons and HI & HC.
Maybe this is a problem because of the few AI fortresses. No fortress -> no advanced units.
IMHO first of all the AI castle building should be fixed. Than it may build HI & HC over LI.
If not, than the devs should implent troop build algorythms.
I will post more ideas soon.

MStavros

May 9th, 2004, 05:48 PM

Yeap, Im gonna make a mod like that. I think it will be interesting to see, that what will happen with the AI, if all LI will be taken out. http://forum.shrapnelgames.com/images/icons/icon7.gif

Oh guys a question! Can we remove/disable spells????

mr.white

May 10th, 2004, 01:08 AM

Personally I think the random fort appearance is a great idea for a higher difficulty level. The biggest problem with the AI is the lack of forts which causes it to use almost nothing but indy troops. Designing a few SC kits for summons like bane lords it could then try to build could also shake things up a bit. It won't take that much to improve it a lot in practice.

By the way, the HOMM3 AI does cheat. It gets bonus resources at the start at higher levels and also a small income in all resources every turn even without mines. Cheating is fine in my opinion. Whatever it takes to make the game interesting.

Kristoffer O

May 10th, 2004, 01:25 AM

Originally posted by proteus:

Kris and the devs -> I never wanted to offend you, I wanted to be creative, I posted the bad parts of the AI, what should be improved, so I dont think that this was pointless.

P. Your first post wasn't pointless, but it has appeared several times before. The arguments that often follow are pointless, both in this thread and others.

proteus

May 10th, 2004, 01:44 AM

Originally posted by Kristoffer O:
<blockquote>quote:<hr />Originally posted by proteus:

Kris and the devs -> I never wanted to offend you, I wanted to be creative, I posted the bad parts of the AI, what should be improved, so I dont think that this was pointless.

P. Your first post wasn't pointless, but it has appeared several times before. The arguments that often follow are pointless, both in this thread and others. <hr /></blockquote>Yes arguments might be 'pointless' in one way, I agree. http://forum.shrapnelgames.com/images/icons/icon7.gif
Ignore the past arguments, and focus on my first post, everything is there, what is wrong with the AI. http://forum.shrapnelgames.com/images/icons/icon7.gif
..btw I apologize again for the arguments, those were posted because I felt that you cannot be bothered about these important things. Maybe I was wrong.

proteus

May 10th, 2004, 01:46 AM

Originally posted by mr.white:
Personally I think the random fort appearance is a great idea for a higher difficulty level. The biggest problem with the AI is the lack of forts which causes it to use almost nothing but indy troops. Designing a few SC kits for summons like bane lords it could then try to build could also shake things up a bit. It won't take that much to improve it a lot in practice.

By the way, the HOMM3 AI does cheat. It gets bonus resources at the start at higher levels and also a small income in all resources every turn even without mines. Cheating is fine in my opinion. Whatever it takes to make the game interesting. Yes, in homm3 the AI getting + resources on high levels, but the core AI is rocking on normal level as well.

Yes I forgot to add that the AI is not building enough forts at all, I will correct my first post.

You know, my brother just mentioned something yesterday, that all LI should be removed from the game, and than the AI would be forced to make strong armies, maybe it would even build forts!!!
We must make a mod like that, I really wonder what would happen!!

[ May 09, 2004, 12:51: Message edited by: proteus ]

PDF

May 10th, 2004, 09:38 AM

Originally posted by proteus:
Thanks for your detailed reply Mr. Parker.
I think that you are right in many things, we should come up with detailed ideas. Propably the devs know about these AI probelms as well...
Castle & Troop building algorythm is what we propably need first of all.
My idea is, that the devs should code the AI to build LOT Less LI units, concentrate on summons and HI & HC.
Maybe this is a problem because of the few AI fortresses. No fortress -> no advanced units.
IMHO first of all the AI castle building should be fixed. Than it may build HI & HC over LI.
If not, than the devs should implent troop build algorythms.
I will post more ideas soon. Yes, I proposed the same kind of mod some time ago, but in a more complicated manner (change cost.res of units to "help" AI to not buid crap). But a straight take-out-the-LI approach is certainly simpler. You will still have AI builds a lot of indy militias http://forum.shrapnelgames.com/images/smilies/rolleyes.gif , but at least national troops should be better... http://forum.shrapnelgames.com/images/icons/tongue.gif

Gandalf,
Your defense of LI is good, but really LI is useless to AI, as it plays it as if it were HI ...
A human player can do *some* things with LI because he'll use it purposefully (patrol, front line fodder, mass enchantments, sneakiness...), that's not the case with the AI, so it should rather not build any http://forum.shrapnelgames.com/images/icons/blush.gif !
(edit) : Why are you asking ppl to give "formulas" for AI ? Users complain about a feature they think "bad", they don't *have to* give a solution...
(Should be your techie-oriented thinking http://forum.shrapnelgames.com/images/icons/icon12.gif )

[ May 10, 2004, 08:45: Message edited by: PDF ]

Gandalf Parker

May 10th, 2004, 02:07 PM

I dont think its the techie in me so much as its the diplomat in me. I could give specific examples if I knew what areas you felt you were expert at.

If someone reads you a paper that they think is good enough to get published but you dont think so, then there are alot of response you can give.

You can say it sucks, and walk off.

You can say it that in your opinion it needs something more, and walk off.

You can say it could use some improvement and at the same time give specifics.

You could give specifics and make some possible suggestions of wording changes.

Of course the easy thing is to say it sucks and walk off. If it doesnt change you can say "I told you" over and over and over. Choosing a different way of approaching it is probably as simple as how good a friend they are and whether or not you really want them to look at what you are saying. This is even more true if what you are pointing at are some lame jokes that the writer doesnt feel is a major point to the paper at all.

I do not consider myself a programmer. But I have worked with programmers and I know that you can often meet them half-way (usually called psuedo-code) then they will see the answer and convert it to real coding. It doesnt take a programmer to talk in language which trys to cover many possibilities using "if this" "if that" "but not this" although having been a parent can help http://forum.shrapnelgames.com/images/icons/icon7.gif

PDF

May 10th, 2004, 03:16 PM

Gandalf,
I agree that complains are more productive if possible solutions are given...
But as long as the complaint itself is argumented ("it sucks" being *not* documented, "AI builds lot of crap LI" is better http://forum.shrapnelgames.com/images/icons/icon12.gif ) I find it mildly offensive to request people to find solutions ("Ah you think AI is crap ! So now tell me how can you do better yourself" http://forum.shrapnelgames.com/images/icons/blush.gif ?)
The problem is not that players should fix the AI, and that's the "build AI" is feeble is no debate... Now it's to IW to state if they will either try to improve it, or else externalize the parameters to allow players to mod it (techiespeech back http://forum.shrapnelgames.com/images/icons/icon12.gif )

Gandalf Parker

May 10th, 2004, 05:15 PM

Originally posted by PDF:
Gandalf,
I agree that complains are more productive if possible solutions are given...
But as long as the complaint itself is argumented ("it sucks" being *not* documented, "AI builds lot of crap LI" is better http://forum.shrapnelgames.com/images/icons/icon12.gif ) I find it mildly offensive to request people to find solutions ("Ah you think AI is crap ! So now tell me how can you do better yourself" http://forum.shrapnelgames.com/images/icons/blush.gif ?)
The problem is not that players should fix the AI, and that's the "build AI" is feeble is no debate... Now it's to IW to state if they will either try to improve it, or else externalize the parameters to allow players to mod it (techiespeech back http://forum.shrapnelgames.com/images/icons/icon12.gif ) PDF I didnt think it would be you that drove me outside of my diplomatic stance. Please dont take this personally, I figured it would be aimed at someone else. Read this quick because I will probably regret posting it as soon as I hit enter.

I havent told anyone to do anything. Ive been suggesting it whenever someone seems to be suggesting that a loud complaint will do the same job.

The devs HAVE answered in every *&^%$# thread on this subject. Do they have to answer EVERY post IN the thread for people to feel they are getting heard? READ THE THREADS and you should end up with the same impression I did.

Let me make some CHOICE pointers here....
Kristoffer in "SUGGESTION: AI troop building algorithm "
Hmm, it's a bit difficult to follow without graphical representation. I should test it when I get some spare time.

A nice answer, an answer of hope, given in the thread I keep trying to point to. The one that gives testable cases and possible solutions.

Later...
Kristoffer in "Good game, but....the AI is very bad"
Hopefully there will be improvements in the AI, but as before I can't give you any guarantees.

Im abit surprised he even stepped into that one considering the title itself says its going to go in the wrong direction. But he did answer and quite early in the thread.

Im sorry I was busy and didnt notice the thread in time to maybe keep it from going to....

Kristoffer later in "Good game, but....the AI is very bad"
Perhaps all the pointless complaining made us pissed.

There was some constructive AI discussions recently that made me interested, but I'm quickly loosing interest. Right now I'm more interested in doing something I am feeling positive about like improving Pangaean themes or finishing the Shepherds of Creation scenario, or for that matter work at school.

The reason I do not say that there will not be an AI improvement is because we also feel that the AI should be improved.

Now, let me ask. Is anyone still confused about the best choice of direction for those who REALLY WOULD LIKE to see the AI get some improvements? Is anyone still confused about the best choice of direction to go if you want to gaurantee that the devs move the AI DOWN on the wishlist? Now Im sure that we will hear some discussion on whether some people feel it SHOULD be that way or not but I cosnider that to be another dead end discussion. Im a solo player interested in a better AI and Ive made my choice.

http://smileys.smileycentral.com/cat/26/26_9_19.gif ('http://www.smileycentral.com/?partner=ZSzeb008')

proteus

May 10th, 2004, 06:16 PM

Gandalf, I am 14 years old, I dont think that I could help in any coding/design tricks.

I always thought that this part is really up to the devs. We -the players- are posting our opinion about various subjects, maybe making suggestions.
I didnt knew about those old topics what we had, so maybe it was really pointless to make this post, since that old thread is from november, and the developers never cared about that thread at all, as I see. It was a huge thread with LOT of good ideas, I just saw it.

Gandalf even you posted a lot in that thread, I just saw it, so you must know, that lot of good ideas can be found there, and still, the 'fix the ai' thing was on the bottom on the 'to do' list, that is for sure.
Hopefully it isnt there still..

I am really trying to be constructive, but this AI coding thing is a bit too complicated to me I think, that is why I only mentioned the LI and castle building problems in that case.

P.

Gandalf Parker

May 10th, 2004, 06:33 PM

Originally posted by proteus:
I always thought that this part is really up to the devs. We -the players- are posting our opinion about various subjects, maybe making suggestions. Think whatever you want. It wont change the way things are. I think people should offer solutions instead of problems but thats a common saying in the workplace.

I didnt knew about those old topics what we had, so maybe it was really pointless to make this post, since that old thread is from november, and the developers never cared about that thread at all, as I see. It was a huge thread with LOT of good ideas, I just saw it.Which thread is this that they "never cared about"? I thought they answered in most of the threads?

Gandalf even you posted a lot in that thread, I just saw it, so you must know, that lot of good ideas can be found there, and still, the 'fix the ai' thing was on the bottom on the 'to do' list, that is for sure.
Hopefully it isnt there still..Of course it is. Well not the bottom but lower than many other things. Why wouldnt it be? No one has come up with anything to change that yet.

I am really trying to be constructive, but this AI coding thing is a bit too complicated to me I think, that is why I only mentioned the LI and castle building problems in that case.Bull. You cant play a game of Dominions pretending to be the AI? Try playing with no LI at all? You can beat the AI so often that this is a problem but you cant play like a dumb AI to test a theory?

Do you have any idea what army the AI should have? How do they get that? Do they build all HI until they cant afford more and then build LI? Or should it be LI, LI, HI, Cavalry, repeat until broke. Now how well does that work if you arent playing Ulm? Does it work for Man? Pangaea? Ctis? Caelum? Atlantis?
Should they build only at home? In every province? How well does that work if you are on a map larger than the small maps? 100 provinces? 200? 300? 500?

[ May 10, 2004, 18:07: Message edited by: Gandalf Parker ]

MStavros

May 10th, 2004, 08:01 PM

I have some ideas.

The AI shouldn't build LI at all. The AI cannot make good moves with the LI. It shouldn't build LC after turn 10. This should be a core rule for all nations.
I posted my idea about the AI castle building, in a previous thread.

As I said before, the castle building problem must be fixed first of all. Maybe with more forts the AI wont build as many LI units. If it will, than code it, to dont build LI at all, because LI controlled by the AI is everything just not a good army.

Well I started to make my mod, so we will see what will happen without LI.
One problem could be here, the provincial militias. Those are weak LI units, and I bet that the AI will build them.
Perhaps I will change all provinical militias as well. They will be strong HI units.

Gandalf Parker

May 10th, 2004, 08:14 PM

Well with your mod you will be able to test that. There is a --superhost command which will let you loop the game on automatic. You can add --scoredump so that each turn will spit out the scoreboard. That way you can keep an eye on things and decide when to kill the loop to jump in and see how things are.

MStavros

May 10th, 2004, 08:20 PM

Ah thanks Gandalf, good infos! It will be a few days until Im ready with the mod. If it will work, I will make it downloadable.

NTJedi

May 10th, 2004, 09:37 PM

Actually the AI is okay considering the size of the gaming company. Developing an intelligent AI is probably one of the most challenging tasks gaming companies have. Most gaming companies develop a single artificial intelligence which follows a series of x,y,z steps. Once gamers identify the set pattern it's easy to conquer the computer opponents.

I believe the best choice would be to begin developing improved AI for Dominions_III since the biggest and best changes most likely cannot be done for Dominions_II.

BEST IMPROVEMENTS FOR THE IMPROVING THE ARTIFICIAL INTELLIGENCE:

1) Randomly selected AI personalities at the start of each game. Thus only one or two personalities out of 10 would use the tactic of massing lots of low level units.

2) With the assistance of scripting gamers are able to create or improve existing AI personalities. This was done long ago with Warcraft_II and Age_of_Empires_II.... and even recently done with Neverwinter Nights. I see no reason why it could not be introduced for Dominions_III.

This would allow for a strong replay value for AI opponents.

PvK

May 10th, 2004, 10:01 PM

I'd say the AI is remarkably good. Especially considering both the complexity of the game, and the level of AI in other strategy games, even ones with much less complexity.

My own win/loss record against the AI is something like 3 complete losses (I was wiped out after making some initial progress), 2 convincing wins (not yet completed), and 5 games with very long stalemates and very near defeats before eventually I'm starting to pull ahead (but also not completed).

I'd say that's a pretty awesome record for the AI, considering I'm a very smart strategy game player with decades of gaming experience, and that the game has insane amounts of complexity.

Of course, I'm not just duelling against a single AI in these games - I tend to play against multiple AI's with some or all of them given bonuses, and I don't go nuts trying to exploit techniques to the utmost, but rather I play to enjoy the game by trying out different strategies and techniques to see how they work, and so on. I don't recall every having a more rewarding experience against an AI player in a strategy game.

The "Light Infantry issue" isn't so much an AI issue as it is a game balance issue. The LI fight like HI and so become rather ineffective in most cases, and therefore they are generally a weak choice. I would say that should be addressed by making LI a more viable choice, by adjusting the tactical AI so LI don't often get caught in melee with superior enemy forces, and/or by reducing their gold cost in most cases.

PvK

MStavros

May 10th, 2004, 10:42 PM

Hum, maybe "lot of" problem will be solved, with the removal of LI. We shall see. http://forum.shrapnelgames.com/images/icons/icon7.gif
Damn I cant wait to see, what will the AI do than! http://forum.shrapnelgames.com/images/icons/tongue.gif

[ May 10, 2004, 23:14: Message edited by: Gandalf Parker ]

proteus

May 11th, 2004, 01:16 AM

Originally posted by PDF:

(edit) : Why are you asking ppl to give "formulas" for AI ? Users complain about a feature they think "bad", they don't *have to* give a solution...
(Should be your techie-oriented thinking http://forum.shrapnelgames.com/images/icons/icon12.gif ) Hehe yup, I agree with this sorta. Anyways Gandalf propably meant that we should help the devs with that. However they know that whats wrong with the AI, I guess they can solve those problems with coding/testing.
I will try to help them, to give them examples if I can, but I dont know that it will help at all or not. http://forum.shrapnelgames.com/images/icons/icon9.gif

proteus

May 11th, 2004, 11:57 AM

Originally posted by PvK:
I'd say the AI is remarkably good. Especially considering both the complexity of the game, and the level of AI in other strategy games, even ones with much less complexity.

My own win/loss record against the AI is something like 3 complete losses (I was wiped out after making some initial progress), 2 convincing wins (not yet completed), and 5 games with very long stalemates and very near defeats before eventually I'm starting to pull ahead (but also not completed).

I'd say that's a pretty awesome record for the AI, considering I'm a very smart strategy game player with decades of gaming experience, and that the game has insane amounts of complexity.

Of course, I'm not just duelling against a single AI in these games - I tend to play against multiple AI's with some or all of them given bonuses, and I don't go nuts trying to exploit techniques to the utmost, but rather I play to enjoy the game by trying out different strategies and techniques to see how they work, and so on. I don't recall every having a more rewarding experience against an AI player in a strategy game.

The "Light Infantry issue" isn't so much an AI issue as it is a game balance issue. The LI fight like HI and so become rather ineffective in most cases, and therefore they are generally a weak choice. I would say that should be addressed by making LI a more viable choice, by adjusting the tactical AI so LI don't often get caught in melee with superior enemy forces, and/or by reducing their gold cost in most cases.

PvK Hummm, are you a veteren player?? I am not, and I can beat the AI all the time easily, and I am not using any SCs etc. at all.

PDF

May 11th, 2004, 12:11 PM

Gandalf,
I'm sorry you took my post as offensive, maybe I didn't word it correctly and went up preaching stupidly (English isn't my native language).

I didn't make any remark concerning the devs response, I know they hear us and try to satisfy their customers/fans.
And I've no grudge against the local mods either, that make a terrific work.

Apologies for this stupid episode http://forum.shrapnelgames.com/images/icons/icon9.gif .

Gandalf Parker

May 11th, 2004, 02:01 PM

Originally posted by proteus:
Hummm, are you a veteren player?? I am not, and I can beat the AI all the time easily, and I am not using any SCs etc. at all. Its possible. Why dont you post the game parameters. What map? what AIs? what settings on resources and indept strength? What nation do you play?

sachmo

May 11th, 2004, 02:23 PM

I hope my wife never reads this board. She might learn a new tactic to get attention.

proteus

May 11th, 2004, 02:25 PM

Gandalf, I am always playing with indep str 5, because I heard that its the best for the AI.

Hmm I am playing against 1-8 AIs, different maps.
All settings on normal, except the AI. Some AIs are on normal, some on the highest diff level.
I do this because that way 1-2 AIs will be very strong in the early game, so more challange!

I think I will play against allied AIs, I never tried that. I think that will be hard enough!

MStavros

May 11th, 2004, 07:12 PM

Hm, this will be a bit harder than I thought. I need some advice. What should I do with the provincial militias? Make them 'unavaible' in a way, example all of them will cost 100 resources, so the AI propably won't recruit them, OR make them all heavy troops?
I like the first one better, that way the AI will only use the national troops, and maybe it will build lot more forts, since the provinical militias will cost more than any other unit.
It sounds weird I know, but maybe the SP AI will be lot more challenging. http://forum.shrapnelgames.com/images/icons/icon7.gif

[ May 11, 2004, 18:26: Message edited by: MStavros ]

Gandalf Parker

May 11th, 2004, 08:01 PM

Originally posted by proteus:
Gandalf, I am always playing with indep str 5, because I heard that its the best for the AI.

Hmm I am playing against 1-8 AIs, different maps.
All settings on normal, except the AI. Some AIs are on normal, some on the highest diff level.
I do this because that way 1-2 AIs will be very strong in the early game, so more challange!

I think I will play against allied AIs, I never tried that. I think that will be hard enough! Other things that boost AI's are:
lower indepts (5 is about human normal)
higher resources
higher magic (if you are not big on spell use)
low random events

If you are going to try allied AI's to make the game harder then take a look at some of the maps Ive done at www.dom2minions.com (http://www.dom2minions.com) such as "Poke in the Eye" "Mazes and Monsters" or "War for Independents". Also, I could probably use some suggestions from you on how to make "WE vs THEM" harder since thats supposed to be a game which FORCES players to ally against the AI (if I can make it that hard)

PvK

May 11th, 2004, 08:11 PM

Originally posted by sachmo:
I hope my wife never reads this board. She might learn a new tactic to get attention. LOL! Yep!

PvK

Gandalf Parker

May 11th, 2004, 08:24 PM

Originally posted by MStavros:
I like the first one better, that way the AI will only use the national troops, and maybe it will build lot more forts, since the provinical militias will cost more than any other unit.
It sounds weird I know, but maybe the SP AI will be lot more challenging. http://forum.shrapnelgames.com/images/icons/icon7.gif To at least test the concept quickly the first one would be easier. Leif or I could also write you a routine which populates a map on purpose with certain pop-types and leaving others out. Maybe making all the provinces have knights and longbows http://forum.shrapnelgames.com/images/icons/icon7.gif

Leif_-

May 11th, 2004, 08:27 PM

Originally posted by Gandalf Parker:
[QUOTE]To at least test the concept quickly the first one would be easier. Leif or I could also write you a routine which populates a map on purpose with certain pop-types and leaving others out. Maybe making all the provinces have knights and longbows http://forum.shrapnelgames.com/images/icons/icon7.gif Actually, look at the DovregubbensHall.ran included with the DomScripts.zip for an example of a randomisation configuration that does just that (although mooseriders are much more fun than longbows).

PvK

May 11th, 2004, 08:34 PM

Originally posted by Gandalf Parker:
<blockquote>quote:<hr />Originally posted by proteus:
Hummm, are you a veteren player?? I am not, and I can beat the AI all the time easily, and I am not using any SCs etc. at all. Its possible. Why dont you post the game parameters. What map? what AIs? what settings on resources and indept strength? What nation do you play? <hr /></blockquote>Well proteus, I've been playing Dominions since several months before Dominions II came out, and I got Dominions II immediately after it was released.

As Gandalf suggests, and as I mentioned, the game setup, and the method of play, make a big difference. I tend to include plenty of AI's, and usually end up fighting one or more AI's within the first 20 turns, because they start out only a few provinces away. Or, in another game, I managed to spread out fairly well when I ended up having to fight three AI's at once (who were also fighting each other) and got down to about two or three provinces left before I made a comeback.

Anyway, difficulty depends a lot on the situation.

P.S. That is, it's essentially impossible to program an AI for a game like this that will be as challenging as a smart and experienced human player. However it is very possible to set up the situation (more AI's, AI difficulty level, do nothing for a few turns at game start, etc.), and/or to limit yourself from taking advantage of the AI by not over-using techniques that the AI doesn't, or not using techniques that exploit known AI limitations, or simply to think and play like a roleplayer rather than like a minimaxing munchkin). I haven't found it necessary to do more than include enough AI's to get some extremely satisfying, interesting, and challenging games, though.

PvK

[ May 11, 2004, 19:42: Message edited by: PvK ]

proteus

May 12th, 2004, 12:33 PM

Thanks Gandalf! I downloaded those + the other new user made maps as well now! http://forum.shrapnelgames.com/images/icons/icon6.gif

Gandalf Parker

May 12th, 2004, 02:18 PM

Originally posted by MStavros:
Hm, this will be a bit harder than I thought. I still havent really gotten into MODs. Im having too much fun with MAPs. But I was thinking about it this morning.

Do you program in anything? It should be real easy to generate a text file which addresses each of the (1100?) units. Then with the help of a list of the units such as Ive seen elsewhere around here, you could just have the mod commands you want to use copied into the clipboard. Scroll thru the list and hit Ctrl-V to paste those at each number you want to mod.

If you cant generate a starter .dm of all the units then I can whip one off for you. Or you could download one of Leifs list such as the randomized one, and strip it.

MStavros

May 12th, 2004, 06:29 PM

Yeah, I really need a 'perfect' unit list, I was wondering already, that where to find one.

Also is there any lists for weapon 'IDs'?

PvK

May 12th, 2004, 07:55 PM

Not sure about a complete unit number list, but there is a weapon/armor list on the stickied mod thread on this forum.

PvK

delacroix

May 12th, 2004, 08:03 PM

Stavros / Proteus, copy/paste the below text into notepad and save it as any filename xxxxxxx.dm Then put it into your dominions\mods directory and test it out if you wish. Don't forget to enable the mod in dominions under "preferences" / "mod settings".

It makes all Militia / Light Infantry cost 999 resources so the AI hasn't built any in my quick 25 turn test game. Met armies of Ulm which recruited pikemen + infantry of ulms + archers + sappers + siege engineers and had 4 Tribal cavalries recruited. Mictlan who had 2 100% national armies of about 30-50 units a piece. And only saw 1 vanheim army consisting of 20 hirdmen and 5 Van's.

Haven't even run into any independent High Infantry unit yet although I doubt the AI has enough resources in the independent provinces to make them at this point. But you can report the results of what the AI builds instead once the LI are no longer feasible. I think archers may end up being more frequent... but thats just a guess.

Complete unit list can be found in this thread :
http://www.shrapnelgames.com/cgi-bin/ultimatebb.cgi?ubb=get_topic;f=74;t=002613

*** start of file, copy/paste everything below, don't include this line and save as filename.dm from notepad

#modname "high cost LI"
#description "makes all Militia & Light Infantry cost 999 resources"
#Version 1.0

#selectmonster 18
#rcost 999
#end

#selectmonster 28
#rcost 999
#end

#selectmonster 29
#rcost 999
#end

#selectmonster 30
#rcost 999
#end

#selectmonster 31
#rcost 999
#end

[ May 13, 2004, 18:57: Message edited by: delacroix ]

MStavros

May 13th, 2004, 03:39 PM

Okay thanks for your help! http://forum.shrapnelgames.com/images/icons/icon7.gif

Kristoffer O

May 13th, 2004, 04:59 PM

You might wanna avoid militia 999 gold cost. Upkeep from random events will be horrible in this case. Also a problem if a nation begins the game with LI. Replace initial armies to avoid the problem.

PvK

May 13th, 2004, 05:58 PM

Perhaps high resource cost would be enough.

Or better: low gold cost, at the correct level to balance them so they are not a bad choice anymore...

PvK

delacroix

May 13th, 2004, 06:33 PM

Yep thanks and noted Kris. The only thing I'm worried with keeping default gold costs and 999 resources cost is that the AI may try to build a militia unit and in the process will block production in its independent province forever http://forum.shrapnelgames.com/images/icons/icon9.gif . Am going to test that out now though.

PS: editing the mod text file will take effect during already started games too. (To anyone who tried it out and wanted to avoid the high upkeep syndrome).

[ May 13, 2004, 18:58: Message edited by: delacroix ]

Gandalf Parker

May 13th, 2004, 06:53 PM

Originally posted by delacroix:
PS: editing the mod text file will take effect during already started games too. (To anyone who tried it out and wanted to avoid the high upkeep syndrome). REALLY? Now as a Random Fanatic that brings some intresting possibilities to mind. Yes, very interesting

delacroix

May 13th, 2004, 08:07 PM

Ok tested militia / LI at 999 resource costs with default gold costs and the AI built crossbowmen / archers in 2 of such provinces over 20 or so turns that had Militia available instead.

So its safe to assume the AI won't block production in all of its independent provinces due to the high resource unit costs. Maybe 1 or 2, maybe none ever, but never all of them http://forum.shrapnelgames.com/images/icons/icon7.gif .

delacroix

May 13th, 2004, 08:16 PM

Originally posted by Gandalf Parker:
REALLY? Now as a Random Fanatic that brings some intresting possibilities to mind. Yes, very interesting Yep really. The only thing is you must quit dominions entirely for changes to take effect. I assume dominions reads/checks the mod files on load-up only. Of course the game has to be created/started with the mod file enabled.

I double-checked it again. Initially had the 999 resource militias. Exited game, edited my LI.dm down to 9 resources, restarted game, built 3 or so militia's. Had them appear next turn. Exited game, edited LI.dm back to 999 resource cost and went back into the game and now can't build them again since they changed back to 999 resources.

Haven't tested it, but I would assume that you can change any unit attributes on the fly without problems. Weapons/armor mods likely too. I'm not sure if scales and/or random event frequency can be changed on a per turn basis too, but again, a quick test can answer that.

[ May 13, 2004, 19:32: Message edited by: delacroix ]

Gandalf Parker

May 13th, 2004, 09:10 PM

Originally posted by delacroix:
<blockquote>quote:<hr />Originally posted by Gandalf Parker:
REALLY? Now as a Random Fanatic that brings some intresting possibilities to mind. Yes, very interesting Yep really. The only thing is you must quit dominions entirely for changes to take effect. I assume dominions reads/checks the mod files on load-up only. Of course the game has to be created/started with the mod file enabled. <hr /></blockquote>Hmmmm I will have to look at MODs again sometime. Most of my regrets are based on MAP commands which I play with more. I regretted not being able to change the game thats running based on triggers in the game. Triggers such as the game hitting turn #10 or turn #100, or having a certain player hit a certain number of provinces, or even randomly deciding its the "Season of Dragons" which will run until it randomly decides its that its not anymore. Or "Season of Vampires" or "Season of Trolls" or whatever.

If a MOD can be changed easily by a program (already verified) and the program can do random decisions (my specialty) or based on triggers (all the above are available) and a changed mod will take effect in an already ongoing game (as you say) then WHOOPEEE someone has turned me loose again! (althought I fully expect Leif to beat me to it)

The mentioned triggers are all now available for outside programs to process but didnt do me much good for maps since only graphic changes affect ongoing games in MAPs.

[ May 13, 2004, 20:13: Message edited by: Gandalf Parker ]

delacroix

May 13th, 2004, 11:10 PM

Sounds like you'll be busy then Gandalf http://forum.shrapnelgames.com/images/icons/icon10.gif . I look forward to any works you may have on the matter.

Existing games don't seem to have any sort of check of whether the mods themselves get modded. I just renamed Unwise's Flag mod to LI.dm in my existing test game and started up the game, and lo and behold all the flags changed and militia no longer cost 999 resources.

Although the "you must quit the game for mod changes to occur" is a little bit of a hinderance. I wonder if there could be a way to have the game refresh mod files during the turn generation process or to have the game auto-quit/restart. But I guess you can't have everything you want lol.

If the mod commands could include some of the map-commands that would be a nice bonus too. Such as the ability to add monsters on a specific province (like the barbarian invasion random event), or change province ownership, or randomly add forts on random provinces to help the AI out, all during the middle of the game could make for some creative scripted scenarios. Although all of the above are 'random events' of the game, so they already happen mid-game and implementation of them into mod commands would not be too far-fetched. I don't know if said features already exist as mod-commands, but they aren't documented if they do.

edit : typo corrections

[ May 13, 2004, 22:17: Message edited by: delacroix ]

Gandalf Parker

May 14th, 2004, 12:10 AM

"And on the 50th turn into the game the sun went dark and a global enchantment gone horribly wrong fell upon the world. No one seemed to claim having tried to cast it."

"We discovered that every province Ruby Amazons had suddenly become Vampires (troops) and Vampire Queens (commanders). Not only did the province I was attacking suddenly change (completely wiping out my army) but the province I already had allowed me to recruit them. Actually this would have been greatly in my favor except that the Onyx Amazons (which my enemy had 2 provinces for) had become various types of Horrors."

"Each turn of the sandclock causes me to fervently hope that this horrible enchantment had run out."

[ May 13, 2004, 23:12: Message edited by: Gandalf Parker ]

MStavros

May 14th, 2004, 04:31 PM

Yup, 999 resource to 'disable' a specific unit for the AI is good enough.
999 gold might cause trouble sometime. http://forum.shrapnelgames.com/images/icons/icon6.gif

NTJedi

May 14th, 2004, 05:37 PM

Well there seems to be a more obvious problem with the basic militia which can be purchased on these independent provinces. Clearly no gamer with even some experience would purchase 2 or more of these units.
This indicates the cost of the Light Infantry units as being too expensive. For human players to ever consider using several of these units the gold cost will have to drop.

[ May 14, 2004, 16:38: Message edited by: NTJedi ]

Leif_-

May 14th, 2004, 05:41 PM

Originally posted by Gandalf Parker:
If a MOD can be changed easily by a program (already verified) and the program can do random decisions (my specialty) or based on triggers (all the above are available) and a changed mod will take effect in an already ongoing game (as you say) then WHOOPEEE someone has turned me loose again! (althought I fully expect Leif to beat me to it)Hmmmm.... *ponders* ... Naaah - too much work. You'd basically have to have a program running in the background, analysing the contents of the score files for a particular game to see if the trigger condition is met; replace the mod file and tell you to restart Dominions (and distribute the new mod file to any other players). In addition, there's limits to what you can change with the mod - in particular you can't make any changes to the map (so no adding new units or magic sites anywhere.)

I suppose you could use this to implement "technological advences" though - at first your infantry only has leather armour and pointy sticks, then after a year they get bronze cuirasses and spears, then two years later mail and swords and five years after that they get full plate and zweihanders.

Gandalf Parker

May 14th, 2004, 05:59 PM

Originally posted by NTJedi:
Well there seems to be a more obvious problem with the basic militia which can be purchased on these independent provinces. Clearly no gamer with even some experience would purchase 2 or more of these units.
This indicates the cost of the Light Infantry units as being too expensive. For human players to ever consider using several of these units the gold cost will have to drop. I must be a rotten gamer. I spend most of my money on LI and lite cavalry, especially independent ones. Worse even than the AI does

edit: Hmmm on checking Id be more truthful saying I spend most on the li bowman in independent provinces. But I never let them build up too far before I switch to li to add to it

[ May 14, 2004, 17:00: Message edited by: Gandalf Parker ]

MStavros

May 14th, 2004, 08:21 PM

Hehe, Im doing well with my mod, imho the AI is lot better even now!!!
I have more ideas anyways, I will post more details soon.

Gandalf Parker

May 14th, 2004, 08:33 PM

Originally posted by MStavros:
Hehe, Im doing well with my mod, imho the AI is lot better even now!!!
I have more ideas anyways, I will post more details soon. Excellent. I look forward to it. Is this an improvment across the board? All nations?
Are you going to make the mod available for others to test with?

MStavros

May 14th, 2004, 08:54 PM

Indeed, all nations & independent units will be edited. The result will be a mod without weak national & independent units.
I don't plan to edit the summonable creatures yet. However I plan to edit the "low level" summons later on & I plan to do more minor tweakings as well.

[ May 14, 2004, 19:54: Message edited by: MStavros ]

Gandalf Parker

May 14th, 2004, 09:49 PM

MStavros

May 14th, 2004, 11:34 PM

Yeah, a nastier AI mod. Well said. http://forum.shrapnelgames.com/images/icons/icon12.gif
IMHO this will be a great help for the SP fans, until we will have some AI coding improvements. http://forum.shrapnelgames.com/images/icons/icon6.gif

NTJedi

May 15th, 2004, 02:12 AM

Originally posted by Gandalf Parker:
I must be a rotten gamer. I spend most of my money on LI and lite cavalry, especially independent ones. Worse even than the AI does

edit: Hmmm on checking Id be more truthful saying I spend most on the li bowman in independent provinces. But I never let them build up too far before I switch to li to add to it Obviouslly I didn't make myself clear enough for some folks...
I was NOT talking about archer types or Lite Cavalry((whole different topic)). I was talking about the Militia and Light Infantry units.
10 Light Infantry means 100 gold which could easily be used more effectively except for a few crucial moments.

[ May 15, 2004, 01:13: Message edited by: NTJedi ]

PDF

May 15th, 2004, 12:43 PM

Originally posted by MStavros:
Indeed, all nations & independent units will be edited. The result will be a mod without weak national & independent units.
I don't plan to edit the summonable creatures yet. However I plan to edit the "low level" summons later on & I plan to do more minor tweakings as well. Very nice, I wanted to do this myself but never found the time/courage http://forum.shrapnelgames.com/images/smilies/rolleyes.gif .
A nice add-on would be 17 "national" mods that put back each national "LI" in-game (at normal price) for use by the player nation (ie player activates the mod for his own nation and so has normal/full recruiting choices http://forum.shrapnelgames.com/images/icons/icon7.gif ).
It'll work as well in multi where there are AI nations BTW http://forum.shrapnelgames.com/images/icons/icon6.gif .
I think you should totally separate the Summons issue (in a different mod).