AI (ver1.0): Hosted by mindlace

[Ralph]

For those of you who think my previous post were to pessimistic: Well, i certainly hope you guys are right.
Nothing would please me more than a CIV-3 game-AI that could give these know-it-all hardcore CIV-addicts out there some “What the HELL happened?” and “Oh NO! Reloading a 10 turns before saved file just isn’t gonna help” - type of surprises. That really would be nice – wouldn’t it?

Anyway, i have some suggestions here about AI golden rules:

A GOOD CIV-AI should always pursue long-term strategies (like building terrain- and city-improvements) because it WANTS to – and short-term strategies (like building military-units) because it HAS to.
Time and time again i see AI-controlled factions in SMAC doing it the other way around, and that just dont work against most advanced players.

A GOOD CIV-AI should ONLY allow wars between AI-controlled Civ´s if the human player is behind (or equal to) it in terms of POWER and SCIENCE.
If the Human Player (= HP) speeds ahead in above areas, the AI should immediately terminate all “internal” quarrels, and instead concentrating on competing with HP. By that i DONT mean that alliances/treaties with HP should be terminated, or that everyone suddenly goes to war with you. Some of the more aggressive civ´s perhaps starts a military build-up of 15-20 units against you, before launching it, while the more peaceful ones concentrates on building terrain- and city-improvements, besides upgrading their garrison-units. The latter more peaceful approach is absolutely vital for any late-game competition that goes beyond minor “controllable” border-quarrels.

A GOOD CIV-AI NEVER let its military units wandering around aimlessly just showing off their presence (VERY common in both CIV-2 and SMAC). Any mid- and late-game AI-units (when most of the map is uncovered) should EITHER be garrisoned OR engaged in worthwhile battles. Important AI golden rule.
By “worthwhile” i mean launched military build-ups of at least (= absolute minimum) 5-10 units. Often more.

Above are just 3 examples. Feel free to contribute with your own “A good Game-AI should...” golden rules.

[Ecce Homo]

A good civ-AI should use all abilities that the human civ can use. In Civ 2, the AI does never transform terrain, use Carriers, have nearly as many Diplomats/Spies as a good player, use rivers for movement or build the Apollo Program.

[Seeker]

In regards to ideas about mini-AI personalities to remove micro-managing, look at the old example of KOEI's Genghis Khan or Romance of the Three Kingdoms game, where you would appoint lords for regions of your empire who all had different attributes like economic skill, war skill, and loyalty. In civ cities or continents could be used in place of provinces. A major part of the game would be selecting good leaders and trying to bribe or assasinate enemy leaders.

[yin26]

-=*MOVING THE THREAD UP*=-

[Cerebuz]

Anybody know anything about how you evaluate the landscape and represent strategy points in code?

And I too think mini personalities (Intelligent Agents) are the best idea yet.

/Cerebuz
strategy_gamer@hotmail.com

[Flavor Dave]

A few things that, to this non-computer guy, seem pretty simple, and that would greatly improve the AIs. Ralph, what do you think?

1. Come up with an optimum food-arrow-shield ratio, and have settlers/engineers try to achieve it. Often I've stumbled across an AI late in the game with size 7 cities, but lots of mined hills, some of which can't be used b/c of lack of food. Or the opposite situation, huge cities with lots of trade, but lacking the shields to build universities, stock exchanges, etc. And do you ever see AIs late in the game with RR'd forests? They must "irrigate" them all into plains or something. Sometimes appropriate, often not.

2. Attack in force. Number of units should be proportional to the size of the target city, or the game era. Also, when launching an assault, the AI should use a random number to divide its force. If the random number is .70, then 70% of its attack points will go in one wing of the assault, 30% in the other. Actually, to really be done right, there should be some way to keep this ratio, generally, 75% or more. Perhaps, use two random numbers, and take the one furthest from .50. Anyway, as it is now, once you see the AI attack coming, you can denude all of your other cities, b/c that's where the attack is coming, period. It would be much tougher fighting the AI if you had to worry about an attack coming somewhere else the next turn.

3. Limit the number of fortresses, and make sure they are manned. Some kind of logic loop like--after a city has an attacker and a defender, build another defender to go into a fortress. Once this defender is started, the settler/engineer assigned to the city builds on the following priority. 1st, must have roads. 2nd, prioritize this way--special resource hill, mountain, hill, river, forest, other special resource. Finally, no more than X fortress per city radius. Perhaps, City Size/5--as the city goes from 5 to 6, build a fortress, as it goes from 10 to 11, same. And, +1 fortress if the city has a WOW.

4. No more destroyer attacks!! It works once (if that), then you build the coastal fortress, and the AI is throwing away 42 or 36 shields (deity).

5. This might be hard, but program the units in fortresses to retreat to the city if they survive, but in yellow or red, and have the city's defender take its place.

[Kris Huysmans]

Why not use a learning AI ?
I think you cannot buil a strong AI with computer instinct.
But when you build an AI based on learning neural networks so that he learns of his bad chooses you will have a strong AI.
When you see that the university of Louvain(Belgium) has succesfull build an neural network to buy and sell stocks. Then I think it will be possible to build a neural network that learns civ so good that it will win the world championschip.
the best way to learn looks me:
Let the computer neural networks plays agains Beta testers to learn.

<font size=1 color=444444>[This message has been edited by Kris Huysmans (edited June 05, 1999).]</font>

[Eggman]

Please teach the computer when it is intelligent to ask for a ceasefire (like when it is over the barrell) and when it is not (when it has a huge force of reserve troops to counteract your attack). In Civ2, using the "capture a city or two, accept the ceasefire which the computer will ALWAYS offer, regroup and attack next turn" strategy was, well, silly. Even worse with the UN wonder.

Also, in all my Civ2 games, it seemed obvious that the computer player would build militaries that were 90% defensive infantry and only a few offensive units. If you crushed his offensive army, the CP would become a non-threat for a long time. If you managed to spook the computer player, they built NOTHING but defensive units, stacking them as high as 12-15 in some cities. Considering how I used to overrun their defensives eventually with artillery and especially howitzers, this is a stupid strategy. Clue the computer in that they need a balanced military. 10% offensive just ain't gonna cut it. While I like winning, being attacked effectively every once in a while would be nice.

Teach the computer to leave troops in reserve for counterattacks. I can't believe how many times I destroyed the computer's defense when I bought out (or nuked, if I was in a bad mood) the nearest border city which happened to have almost all of their offensive units. Of course, it didn't really matter since I have never seen the Civ2 CP counterattack, even in the worse situations. I had a size 25 city defended by a damaged howitzer of all things with no chance of reinforcements, and they begged for a ceasefire...

Of course, the worst I have EVER seen is when the Americans attacked my Souix border town. They destroyed the 2 defending riflemen with artillery over railroads (I was surprised! It was good strategy!) and then NEVER BOTHERED TO OCCUPY THE CITY! When I later crushed their empire, their cities were stuffed to the brim with various infantry (and no offensive units, BTW) with some cities in double digits. But the computer apparently never considered that, well, maybe I should unfortify a unit or two and grab that city from my scientifically advanced neighbor that I just declared war on in a surprise attack. Nah...

[yin26]

-=*MOVING THREAD UP*=-

[Zen0n]

Funny thing
When I'm the only one who knows the nuclear fission or/and I'm the only one who has completed the Manhattan Project, I go to enemy city, plant a nuclear device and the message appears "Mysterious nuclear blast in city of ..." .Nothing happens. It's very nice, but I'm the only one, who could do this atrocity. Please, let the computer know that.

[yin26]

-=*MOVING THREAD UP*=-

[EnochF]

Ja, der entire purpose of buildink a Doomsday Veapon is lost... if you keep it a secret! Vhy didn't you tell the vorld, eh?

That should be a new rule. One-sided nuclear deterrence. Once the human's got the Bomb, the AI should play it safe for a few years while they're building nukes of their own.

[yin26]

-=*MOVING THREAD UP*=-

[Ove]

I have posted this on the OTHER and DIPLOMACY also, but this deals mainly with AI agents.
They don't have to be terribly bright, they just have to evaluate their own priorities and then present and compete for the Ruler's (AI or Player) acceptance.

AI ministers/governors
The ministers in CIV2 were an amusing interruption of the usual chores (micromanaging cities, micromanaging units, etc.). They didn't really give any useful information, they just complained if you were not doing what they thought should be done. Who is the ruler here?

It would be better if they were able to act upon directives from you (or each other, depending on how much power you grant them) as an extra (micromanagement-reducing) layer between you and the city/regional menus. After they have been given tasks or general directives, they present you with their suggested solution(s) which you may accept/modify/decline.

example:
You have given your diplomacy minister a directive to improve relations with your neighbour.
Diplomacy informs you that your neighbours insist on a special trade relation in which they will buy weapons for food.
The ministers of trade, production and military tries to dissuade you from that course of action, because; you will lose money, your production facilities are already engaged with other orders and it is dangerous to arm your neighbour.
You instruct diplomacy and trade to make the deal anyway, but to delay the weapon shipments. Production are ordered to commence the weapon production and turn them over to military who will use them to make an attack army that can crush this insolent neighbour.

Diplomacy's directives could be shaped like this:
MISSION (to Persians):
x Improve Relations o Provoke War
Get
o Territory o Bases o Passage rights o Technology o Money o Goods o Trade agreement o Prohibition against Slavery/Ethnic/Pollution/Drug/Religion/... o Acceptance of Slavery/Ethnic/Pollution/Drug/Religion/...
Give
o Territory o Bases o Passage rights o Technology x Money (200 gold) o Goods x Trade agreement o Prohibition against Slavery/Ethnic/Pollution/Drug/Religion/... o Acceptance of Slavery/Ethnic/Pollution/Drug/Religion/...
Willingness to achieve mission goal(s) (1-10): 7
Willingness to offer gift(s) (1-10): 5
Who can sign agreement? o Agent x Emperor
Duration of mission? o Immediate o Fixed # of turns x Until an agreement is reached

[Urban Ranger]

You guys are all mad.

The suggestion to use heuristic algorithms is a good one.

There are many suggested the use of a script to drive the AI module. The one fatal flaw of this is speed. A worse suggestion is to use JavaScript/Python/etc. They are even slower. Maybe for those of you with boxes that run on 4 Pentium Pros (or at least 2 PIIs) don't care, but there are still those out there with low end (e.g. Pentium 166) machines. Using any sort of script will force the game to a crawl on those computers.

Darkstar suggested using an API approch so he can write his own AI routines. That's a lot of effort that does preciously little. How many of Civ3 customers have such inclinations and how many have the actual skills? Photoshop is a program that allows external plug-ins. Consider how much more it costs.

The client/server model is also silly. If you can only run one game at a time, why bother with the effort? Just slap the whole bloody thing on a Wintel box and be done with it. If the server can run multiple games at a time, it will be a lot more complicated, both in terms of code and pricing struture. Would you have things like a 25-user Civ3 server? Would the servers be sold seperately from the clients? Again, there seems little to be gained.

I think what is needed is a look at the AI
algorithms. The actual implementation doesn't matter all that much.

The AI algorithms should be modelled after top Civ players. Not just Sid and Brian but people who spend hundreds of hours with the games and can kick the computer players blindfolded.

[NotLikeTea]

Keep in mind that the people who can beat the computer blindfolded do so not due to superior planning and tactics, but because they know how to take advantage of the computers faults.

For good tactics to emulate, they should look at the best multiplayer tactics, where AI faults do not come into play.

<font size=1 color=444444>[This message has been edited by NotLikeTea (edited June 22, 1999).]</font>

[Darkstar]

Urban Ranger - "You guys are all insane."

Yep. And so are you.

Without true AI, we have a problem of a non-learning single player opponent engine. Which once you learn the tricks, is easy to defeat. That is the problem with locking it into the code, like it is now in CivII and SMAC. It can't be modified or customized at all after they are done with updates and patches. And only those with the source can do it whatsoever.

That brought up scripts and API and what not. As in other games which are still alive. The *players* of the game are building and updating the A-non-I modules or scripts.

Then we kicked around various ideas how to do the same for CivIII. Both your ideas are the worse possible suggestions... but will probably be the one implemented.

The problem is Civ and SMAC are programmed to react to what the developers found to be most effective. But the few months development time compared to the shelf life is a small amount of time to try and get the "Ultimate" A-non-I built, and the more experienced of us *will* walk all over it as a consequence on the day its released. And they won't release a patch 1 year to 2 years down the road to correct the "bad" or "inefficent" A-non-I. They will be busy developing new games.

I mentioned API as that would be open to at least a small fraction of the players, kept speed for those concerned, be relatively easy to implement from Firaxian side, and would enable people to further extend the system by building script parsers, if that is what the community wanted.

Making the A-non-I open and scriptable from design time would trade speed for maximizing the number of players being able to customize thier play.

And that is what we are after. Customization and updating of the A-non-I.

If you don't want to waste your time with positive suggestions here, then please find another topic where you can contribute positively. We all want to play the best possible Civ3.


------------------
-Darkstar
(Knight Errant Of Spam)

[EnochF]

And I'm still freaked.

[Theben]

1st time here, if someone else mentioned my apologies. Since us human players are so good at tweaking units to our favor (w/ workshop), or building the only descent ones (w/o workshop), why not have the AI copy what we build, as best it can? After it runs in to a couple of our units that give it trouble, it will copy them. Or allow a "AI preference" like to one used for techs in civ2's rules.txt, except for units.

[Flavor Dave]

another threadmaster suggested this idea get posted here. Eggman made the (IMO brilliant) suggestion that the human have the option of selecting normal diplomacy, or realistic diplomacy. In the former, the computer players act like they do now, and once you've become dominant, pitiful civs declare war on the human to prevent the human from winning. In realistic diplomacy, 3rd rate civs ACT like 3rd rate civs.

I pointed out that the AIs have some pretty obvious triggers where they change their attitude toward you--1750, 1850, space techs, launch. So I thought this would be easy to program, just eliminate these triggers if you choose "realistic" diplomacy.

One further thing this will do is add a level of difficulty. Realistic diplomacy will be alot easier than the way it is now. And that means you can make the AI alot tougher, making deity a challenge again (at standard diplomacy). An enhanced AI, with realistic diplomacy, would be as hard to beat as deity is now.

[Gregurabi]

I agree with Darkstar and Rong. The biggest failure of CTP and (to a lesser extent) Civ2 is the lack of a competent AI. Civ1's AI cheated too much, so it's disqualified from the running. We need to be able to extend and improve the game as we learn to beat it so that it will grow with us.

I disagree with Urban Ranger's comments. First of all, it is not yet known whether a scripted AI will be too big of a bottleneck. It hasn't even been implemented yet! If the scripted AI runs while the human player is thinking about what to do and fumbling with the interface (or micromanaging), it should have plenty of time to make its own decisions. You need more CPU time? Just turn off the damned animated whirlygigs!

Also, Urban Ranger, the client/server model is not silly. In fact, it's already been done -- check out Freeciv at http://www.freeciv.org/ . (What do you think all of the Linux users played before CTP? "Just slap the whole bloody thing on a Wintel box and be done with it" is exactly what is wrong with 95% of the game companies today!) And while Freeciv won't currently support 25 users in a single game, there's absolutely nothing preventing you from running 25 simultaneous Freeciv games on one system, if you've got the hardware to handle it.

Finally, I agree wholeheartedly with kmj's comments. If I want a second-rate game, there are any number of places I can get one. From Sid Meier (and Brian Reynolds) I expect more than that -- I expect a game that will captivate me for years. And if the AI can't keep up with me, I won't stay interested. I'm a picky user, and I want QUALITY GAMEPLAY.

(Blatant plug: Freeciv desperately needs a better AI. It already has a client/server architecture and is released under the GNU General Public License, so you can hack up the code as much as you need to. If any of you are frustrated with the AI in your current Civlike games and have the talent and time to do better, Freeciv can use you.)

[Theben]

*BUMP* How about putting those AI arguements here instead of the suggestions page?

[Maniac]

That AI builds all his crappy little cities two squares apart. Is it that hard to tell the AI to build them 4 squares apart?

After 3 civlike games that problem still isn't solved. Strange. Unless the designers intentionally want the AI to do that...

Then I'm now telling them : I don't want that!

[Emperor10]

Computer AI thinks in two demensions...there's simply no way to get it to think like a human player. A will always lead to B. By giving the computer more "choices" you can make it somewhat smarter (A might lead to B or C), but no way can you make it an unpredictable entity with logical moves. The more complicated it gets, the more it seems to screw up on the logic. For example, it might hate you simply because you're stronger, even if you give it tons of gifts. The computer has no emotion, it simply can't get that you're trying to be friends, all it cares about is making all the numbers even to you, because that's what the program tells it to do. Or it might be the opposite, it might like you more everytime you give a gift. It doesn't understand the fact that you have 20 units along the border could mean its a trick as long as you stay on your side(which is how it divides "hostile" from "friendly"). Either way, it is very hard to make the AI do anything other than math. Everything is a knee jerk reaction.

[Mark_Everson]

Actually, M@ni@c, building crappy little cities two-three squares apart is probably IMO The Best strategy at the beginning of the game. What the worthless AI doesn't know (among Many things) is how to transition this strategy into big cities in the mid-game by starving some of the worthless ones out of existence. I and others have had many discussions on this in Civ2 Strategy from time to time. You might be able to find them if you search under "small cities" if you're interested. Or you could search by my handle since I've probably only got ten or so posts there. I'm sure not everyone agrees with my small-city stance, but its backed up by some mathematical analysis of the game which is more than you can say for most. If you like empirical evidence, its also backed up by multiplayer victories against something like 40 "deity level" opponents with no defeats .

Emperor 10:

I think its possible to do a Lot more with AI than you're willing to admit. Check out http://people.mw.mediaone.net/markeverson/clash_ai1.htm to see our ambitious plans for the AI in Clash. Will all our ideas work? Probably not, but I am Sure we can do better than the mindess junk that passes for AI in civ.

------------------
Mark Everson
Project lead for The Clash of Civilizations
(That means I do the things nobody else wants to do ;-) )
This Radically different civ game needs your suggestions and/or criticism of our design.
Check our our Forum right here at Apolyton...

[Maniac]

Perhaps it's the best strategy, but it's cheating. IT'S ICS.
So I hope you're not gonna learn your Clash AI how to cheat better. You should make it the human player in someway inadvisable to use the Civ2 cheats.
<font size=1 face=Arial color=444444>[This message has been edited by M@ni@c (edited August 20, 1999).]</font>

[Mark_Everson]

No, M@ni@c it is Not cheating. It is playing the game as designed. I agree its a Huge design flaw, but there it is. Unless the players agree beforehand that it is not to be used, and if so what the new rules are, it is Completely Legitimate. Besides ICS only gives a moderate advantage at deity level. Its only at the lower levels where it really gets out of hand.

Clash will not use ICS-type cheats because the econ and population growth models have Nothing to do with the bizarre ones in Civ.

[Maniac]

*BUMP* Post here your AI ideas.

[Mo]

In normal diplomacy the ai should only declare war if it is going to mount an attack or if it already has units with which to attack. Too often has a small civ declared war on me and after 10years/turns I still wasn't attacked. Unless it is inresponse for an action on your part or request of an ally the ai should prepare itself before declaring war.

[KDan]

Well, the other version was messy-ing it all up, so here it is in the right thread...

I would be really impressed if, in civ3, I actually saw some real strategy displayed by the AI.
There are a great, great number of tricks which one can play upon the opponent to confuse him and eventually beat him, and right now the AI uses none of them (for Ctp anyway).
Here's a short, inextensive list of things the AI could and should do. For more reading I would recommend "36 Stratagems", a famous chinese strategy book which should be mandatory reading for all the civ3 team...

1) Feints: The AI should be able to stage up feints, as in giving the impression of attacking somewhere, whilst attacking somewhere else. That is really quite a basic trick.
2) Secretly staging a decisive attack, and watching out for such attacks. A tactic which works very often in Ctp is to build up a small strike force of just one or two very strong stacks, and ship them to the ennemy's strongest cities by sea or space. Then suddenly you drop them in the middle of your ennemy's empire, where he is not expecting them, and take a few key cities, and the ennemy is beaten. That would not work very often in a human vs human game, and the AI doing this would be a welcome (if disastrous) surprise. As it is right now, the only kind of attack I've seen from the computer is the random, unregulated flow of units towards my cities as if they could do something by sheer force of numbers - which they can if the other empire is far stronger. But since most of the time it's not, this strategy (or lack thereof) is disastrous for the computer which looses its units one by one as they clash against untakable objectives and ambushed forces.
3) Attacking to defend. When I wander around my ennemy's territory with slaver stacks I see lots, and I do mean lots of units wandering around from city to city, following me around (and they are actually my main target since they provide me with slaves!). Would it be so hard to get the computer, when under heavy attack, to mass those forces, break through the frontline if possible and go attack some lightly defended cities instead of just providing prey for my stacks?
4) Another feint: Give the impression of emptiness (no units) where there are lots, and give them impression of having lots of units where there are only a few. Right now it's just plain easy to guess where the computer's units are: They're right where my units are!
5) Communications. In warfare, an army with communications is an army which has the possibility of retreating on controlled territory to get back home, and which has a supply of ammo and food and everything. An army without communications is isolated and cannot retreat or refuel or etc... you get it. If I send a stack in the middle of a Ctp computer civ, I can roam around, do a lot of damage, and then retreat when my guys are getting weaker. This should never happen. Especially with the Zones of Control feature of civilization, such an army (say of 4 samurai, 4 archers and a slaver) should be surrounded by 3 stacks of about 4 legion-like units, so as to prevent it from moving without taking massive damage. That already would masively limit the damage. Also the entry point of the stack should be blocked, so that if it does attack one of the blocking armies, it will be crushed in a few turns because of all the damage it has taken. It's easy to get rid of parasites, but you have to do it. The Ctp AI doesn't do it.
6) As a sidenote, sea warfare is very, very poor in Ctp. The AI just cannot control the seas. That sucks. Please fix it. Controlling the seas has been and still is one of the major points of any strategy, from the Ancient Greeks and Romans to the Spanish armada to modern battlecruisers. Without command of the sea the English would probably have been beaten by the Germans.

That should do for a start. I'll try to get my hands on this book (36 Stratagems) to give you the reference for it. I only have it in French but I'm sure at least some people at Firaxis/Activision speak French :-)

Daniel

------------------
http://dwdt.xs.mw