Lots of Features AND Smart AI

[nato]

Short Version:

Just because a feature is in the game, doesn't mean the AI has to use it, or even should use it. Let it ignore the fancy stuff completely and excel at the basics ... but still provide the fancy features for the human player to use. For expansion packs, add new interesting features for the player to use, but don't bother the AI with it.

Long version:

Why not design the game so that there are many features, but the AI simply ignores the complicated ones that it could not make intelligent use of?

Artillery is what made me think of this. The AI was actually weaker because of its poor use of the concept, than if it had ignored it altogether. But artillery is a great game feature, and not including it would have lessened the game.

So keep interesting features, but let the AI ignore it!

The obvious complaint will be that this gives the human player options and an advantage the AIs won't have. This is very true ... however ...

Human and AI players are already inherently different! They already "think" completely different, "see" the map completely different, and make decisions completely different. Giving the player some extra capabilities is really a minor detail, as long as you are careful that they are not too powerful compared to not having them, like SMAC's vaunted "chop and drop" tactic. Having a feature the AI lacks doesn't mean it has to be overwhelming.

Again consider artillery. Even when AI didn't have it, it could still compete, just by being smart and efficient with the basic tools it does have. Further, it is sure to have advantages the human doesn't.

Imagine if you didn't have artillery, or some other non-basic feature. You could still put up a good fight. And if you got cheaper units than the artillery guy, its not really clear who has the advantage. It changes AI cheats into AI advantages compensating for your extra options.

Yes, it is very good to make human and AI players as similar as possible. I fully realize and want that too. However there are two problems:
1. It can never really happen anyway because the AI is a computer and you are a human
2. It is not worth trying to achieve this if the cost is fun and interesting game features.

I would much rather win because I made smart decisions using somewhat complicated features to overcome an AI who was a master of the basic features and had advantages with them. This would be much more satisfying than overcoming an AI who has the exact same tools, but can't possibly be as smart as a human like me.

Therefore I say add some complicated features in expansion packs! Give the game the complexity that people miss from SMAC and Civ2! Just don't bother the AI about it ... let him remain excellent at the basics.

This will add features, make good blurbs on boxes, quell complaints, and add a lot for multiplayer, where its all humans anyway. And it should make programming new features less difficult, because the AI doesn't have to worry about it at all.

This way you could get the best of both worlds.

Thanks for reading.

[Ijuin]

Regarding the AI not using features such as bombardment properly, I would suggest that the AI be changed to use artillery and air units in a simpler, more defensive manner. Instead of asking the AI to use such units to atack the player's cities, the AI should simply build artillery and fighter units and leave them garrisoned in its own cities as defense against the player's attacks.

In Civ III, any time an enemy unit attacks a unit that it stacked with an artillery unit, the artillery gets a free turn of bombardment to counterattack, which softens up the attacking unit (I forget whether the artillery gets to counterattack once only, or once for each defending artillery unit). Also, a defending fighter aircraft assigned to Air Superiority has chance of intercepting, and thus preventing, a bombardment by enemy aircraft within its intercept range.

If the AI were to merely create artillery and fighter units and leave them stationed in cities as defense, it would make things much simpler on the AI than forcing it to figure out how to use them offensively. I, for one, would enjoy seeing my first attacking unit forced to advance through enemy fire when I attack cities.

[nato]

Well the artillery was just an example, not my real point. But what you say is kind of what I mean ... the AI is stronger by not trying to do too much.

About the defensive fire, the AI does use artillery for that to my experience. It just builds way too much for that one purpose, and it is a waste of shields and all gets captured (this may have changed with 1.21). Like I said though, the artillery was just an example.

Welcome to Apolyton, btw.

[Skanky Burns]

That will create more exploits.

For example, the AI never uses artillery. The human does. So for every battle the human initiates, they could potentially be attacking 1hp enemies, whereas the AI would always be attacking 3, 4 or 5 hp units.

If you give the player more advanced options, but tell the AI to ignore them, then the AI will always be at a disadvantage.

[Tuberski]

Originally posted by Skanky Burns
That will create more exploits.

For example, the AI never uses artillery. The human does. So for every battle the human initiates, they could potentially be attacking 1hp enemies, whereas the AI would always be attacking 3, 4 or 5 hp units.

If you give the player more advanced options, but tell the AI to ignore them, then the AI will always be at a disadvantage.

The AI never uses artillery now. If you remove that option from the AI, it will concentrate on other things. Instead of the peicemeal artillery you see now. I have yet to be hit by artillery when the AI attacks.

[nato]

Exactly Tuberski! The AI will give a better game by focusing on the things an AI can actually do.

And like I said in my original post:

Giving the player some extra capabilities is really a minor detail, as long as you are careful that they are not too powerful compared to not having them, like SMAC's vaunted "chop and drop" tactic. Having a feature the AI lacks doesn't mean it has to be overwhelming.

[Tuberski]

Sure nato,

We already have several advantages over the AI that they COULD use but don't. So get rid of the things the AI doesn't use, it will use what it has more effectively.

[Evil Robot]

Agreed.

[Capt Dizle]

Who cares about the AI?

Walks away shaking head.....

[Cyclotron]

Originally posted by jimmytrick
Who cares about the AI?

Well, someone must, or else this thread wouldn't be here. Most people like playing single player...

[nato]

Who cares about the AI?

Walks away shaking head.....

If you mean you only care about MP, then that was a really dumb post.

A major thing holding back features for MP humans to use is that the AI must be able to use it. Letting the AI ignore complicated features would allow those features to be put in, and thus available for MP. This would be one of the best advantages because it would improve MP a lot.

If you didn't mean you only care about MP, then it was an even dumber post because it makes no sense at all.

Guess I'll shake my head too.

[Capt Dizle]

Soren and folks dumbed the game down to enhance our single player experience. In other words, to help the AI.

I actually agree with Nato about the need to add back features and complexity to the game. But Nato says:

"A major thing holding back features for MP humans to use is that the AI must be able to use it."

This is untrue. Soren, probably working on orders from Sid to make the game as lite as possible, is the one that insists on the AI being able to use things. In other words, dumb it down.

The reason you won't see added complexity is that Firaxis is not the company it once was. Shallow games with little replayability is the goal, so that you will buy another game soon.

[Grumbold]

Peeling away unnecessarily complex features of a game so that the AI does not have to deal with them I can understand. In some ways it has made Civ III more challenging to play than previous Civ games. Unfortunately something as fundamental as artillery seems to be going too far. If the AI can't use it even halfway sensibly then I believe they should keep working on it until it can. That aside, allowing the AI to build artillery and encouraging it to do so when it has no idea how to use them is certainly worse than telling it not to build them at all.

[Deornwulf]

Perhaps the answer would be to give the AI units and basic abilities unavailable to the human player. Not necessarily as cheats but as advantages unique to the AI to make single play as exciting as possible.

As nato stated before, the AI will not think as a human so quit trying to force it to do so. Along the same train of thought, quit "adjusting" the game to force the human players to play the game exactly like the computer. Every uniquely human action is not an exploit.

[Ogie Oglethorpe]

I wish it would be considered, but in light of numerous comments now from Sid himself, he considered SMAC a failure because of its complexity. From a market acceptance (read huge sales) POV, he feels that SMAC turned people off from TBS games due to its

a) SCIFI genre game

and (driving a dagger thru my heart)

b) it was too complex thus affecting gameplay

In hearing the second comment it makes one wonder whether he means Firaxis couldn't code a credible AI to handle the complexity thus game challenge is a problem, or did he mean the unwashed masses never took to the game because it was too complex.

Looking at it this way one starts to understand how CIV3 came into being. It succeeded addressing both points but left the advocates of SMAC and to a lesser extent CIV2 feeling a bit left out/deserted.

Just my two milliwatts

Og


PS. Since CIV3 seems to be a commercial success (at least short term) Sid prolly thinks his read on the market was the right one. (i.e K.I.S.S.)

[Captain]

nato, about your original post, totally in agreement!

Originally posted by Tuberski
The AI never uses artillery now. If you remove that option from the AI, it will concentrate on other things. Instead of the peicemeal artillery you see now. I have yet to be hit by artillery when the AI attacks.

tuberski, what do you mean remove that option from the AI?
i assume you mean you've edited the game, but how?
did you make artillery unavailable to all AI civs? or did you set the AI to never produce? or did you deselect the bombard/attack unit AI roles setting ?

just trying to get a sense of whether the AI has been prevented from ever using it, or just 'discouraged' from using it (and smart enough not to).

[SirSebastian]

Actually, in 1.21f the AI does use artillery somewhat offensively. =P Of course, that was just used as an example. And, does that jimmy guy have to post everywhere that MP is all that counts?

SMAC was a good game, but Civ III is better. I certainly can understand O.O. being sad about it, but at least he understands what Civ III is. It's "streamlined." Personally, I like the different direction Sid went with it. However, he should be cautioned against dumbing down the game... but I don't think he did this (how could you ever say he did with resources + culture?). It is unfortunate if he truly thinks that SMAC was a failure, it wasn't. It was just different. There is room for both the streamlined and the complex, and SMAC did some things right that no other game has done, IMO. Civ III, however, is just beautiful.

PS - I do a killer Nwabudike Morgen (sp... haven't played SMAC in ages...), "Human behavior, is economic behavior"

[Deornwulf]

What OO brings up is very sad. Perhaps CivIII marks a new era of throwaway TBS games sold only to entertain the masses for a few moments until the next distraction is released. If this is true, then replay has no value to Firaxis or Infogrames. Someone tell me this isn't true .

[nato]

Well no one but Firaxis really knows why design decisions were made the way they were. The game could be simple because:
1. By intentional design
2. For the AI's conveniance
3. Ran out of time

Probably some combo of all three. I guess I tend to think it is more 2 and 3, and hence my post. If it was 1, maybe the less than perfect reception of the game will make them want to change course. In any event, expansion packs offer a real opportunity to change their decision and add features, if they wish.

Endless moaning about Firaxis-is-evil is useless. It just turns all threads into the same discussion. We've heard it all before. I say this even though I lean to the whiners side slightly! In any event, since my idea is arguing for change, it seems counter productive for whiners to turn it into the same old argument.

Anyway, if you really believe there is no hope ... WHY ARE YOU HERE? GET A LIFE!

Artillery is just an example! Replace it with terraforming, social engineering, espionage, specialized units, specialized city improvements ... any concept beyond the basics, that might be too complicated for an AI to use intelligently, but adds fun for the human.

I know, you might not like one of those examples either, and be just dying to say so, but in general more options = more strategy decisions = more fun for the human players.

Peeling away unnecessarily complex features of a game so that the AI does not have to deal with them I can understand. In some ways it has made Civ III more challenging to play than previous Civ games

You have a good point ... I just don't agree because it basically forces humans to operate on an AI's level. If the AI can't handle it, poof! its gone. It seems backwards to me to design the game around the AI instead of the human player. This leads to jokes like "fun was decided to be an unfair advantage since only the human player could enjoy it, so it was programmed out for the AI."

Being streamlined does have advantages, I admit. I guess I am just casting my vote for features over simplicity.

Perhaps the answer would be to give the AI units and basic abilities unavailable to the human player. Not necessarily as cheats but as advantages unique to the AI to make single play as exciting as possible

This would work for me. Cynical people describe the game as AI expoits versus AI cheats. I would rather it be human strategy & options versus AI advantages at basics.

I look at it this way... imagine a war between two forces. One side is small but very elite and clever. It uses tactics, weapons, and capabilities the other side doesn't even know about. The other side is not stupid, but fairly simple. It uses a few simple but effective tactics and basic weapons ... and it uses lots of them because it has more troops than the other side.

So it is a battle of brains versus brawn. I find that compelling. I would love to be the first force, using my clever strategy and special weapons to outfight superior but simple numbers.

Since humans will always be smarter than AIs, arming both sides with the same options will never work. It will have one of two results:
1. The game is on the humans level, and the AI is doomed
2. The game is on the AI's level, and the human is bored

Arming both sides with different options, to take advantage of the human's and AI's unique strengths, can give an overall even match, and an interesting and tough challenge.

Thanks a lot for reading and replying, sorry this one was so long.

[Cyclotron]

Deornwulf, looking at your sig I'm surprised you agreed with Ogie's statement...

and (driving a dagger thru my heart)

b) it was too complex thus affecting gameplay

In hearing the second comment it makes one wonder whether he means Firaxis couldn't code a credible AI to handle the complexity thus game challenge is a problem, or did he mean the unwashed masses never took to the game because it was too complex.

[Mark_Everson]

Great idea nato!

I think the core AI could be clearly enhanced if Firaxis could have been able to focus on doing a few things better. For my money, AI has always been the weak point in the Civ series. I'd probably still be playing actively if the AI were more challenging.

However I differ with one of your previous comments, that Firaxis had spent too much time working on making the features limited by what the AI could handle. I think they would be better going Further in considering what they can provide a good AI for in terms of rules than they do now. YMMV

One other issue is that if the AI lives in a much-simplified world, it could give problems in gameplay. FE if when taking over an AI city all of a sudden it doesn't Work. As in its population crashes, or other jarring things happen. I'm not saying your suggestions for simplified AI would cause that necessarily, but its something that might limit your approach a bit. Lets hope someone in a position to do something listens

[nato]

Wow thanks!

Those are good points, especially the second one. There are always unintended 2nd and 3rd order side effects...

I was under the impression a lot was removed and changed for the sake of the AI. I guess I got this impression reading posts here.

It might be tricky to make the human options and the AI options mesh ... but I think it could be overcome. Especially if, in general, it is just the AI not having options the player does. For capturing cities, for instance, the human player would get a city that didn't have any of his special human-only buildings, but did have the basic ones. But thats just an isolated example like artillery, so the concern still exists.

You are very right, there could be a lot of information or details missing. You'd have to make sure the low detail AI stuff could give all the info the high detail human stuff needed, or that this info could be generated on the spot.

Still I think it could be done ... it would probably be a lot easier than trying to get the AI to understand new features, at least.

[exeter0]

I am sorry but I must disagree.

I do not think that the AI should be disciplined not to use bombard (artillery & bombers etc) tactically.

To be competitive (from the time the cannon is available) the AI must understand the notion/value of combined operations.



I also think that the AI is not too far off... it CAN use artillery and cannon.. it just needs to be disciplined in coordinating attack/defense manouvres. It needs to develop more concentrated cominations of units. I do not thaink that this will be a major

If you suggest to Firaxis and convince them that we do not need the AI to develop combined offensive/defensive tactical manouvres then I think the game will become very stale very quickly.

The AI doesn;t need to be hard coded with instructions for each unit.. just each type of unit.

I sort of envision that the combat model could evolve with an evere growing library of 'set tactical manouvres'.

If I am over simplifyng the task then maybe Firaxis should get a proven wargame AI architect guru.

Look at Close Combat... a brilliant game with a vastly more complex model to achieve because the game operates at the combat unit level: each individual soldier is controlled with different weapons (rifles, sub-machine guns, grenades, flamethrowers, panzerfausts, mortars, heavy machine guns etc.) and will tactically manouvre hedgegrows, buildings, ruins, walls etc for cover.

Please do not lower your expectations of CIV in this regard. I think Firaxis are on the right track and are committed to maturing the AI combat model... look at the features in the XP that have just been announced!

[Tuberski]

Originally posted by Captain
nato, about your original post, totally in agreement!



tuberski, what do you mean remove that option from the AI?
i assume you mean you've edited the game, but how?
did you make artillery unavailable to all AI civs? or did you set the AI to never produce? or did you deselect the bombard/attack unit AI roles setting ?

just trying to get a sense of whether the AI has been prevented from ever using it, or just 'discouraged' from using it (and smart enough not to).

No, I haven't edited my game, it's just that the only AI artillery I see, i capture inside of cities. I never see it in the field.

[Grrr]

I too have yet to see Artillery of the AI's on the field.

[I-iZ-1337]

I've seenz the AIz with catz tho, but they never uzed them d00dz.

[Panag]

Originally posted by Grrr
I too have yet to see Artillery of the AI's on the field.

hi ,

seen it , when you play "deity" , ...

have a nice day

[Ogie Oglethorpe]

Originally posted by nato
Well no one but Firaxis really knows why design decisions were made the way they were. The game could be simple because:
1. By intentional design
2. For the AI's conveniance
3. Ran out of time

Probably some combo of all three. I guess I tend to think it is more 2 and 3, and hence my post. If it was 1, maybe the less than perfect reception of the game will make them want to change course. In any event, expansion packs offer a real opportunity to change their decision and add features, if they wish.

Endless moaning about Firaxis-is-evil is useless. It just turns all threads into the same discussion. We've heard it all before. I say this even though I lean to the whiners side slightly! In any event, since my idea is arguing for change, it seems counter productive for whiners to turn it into the same old argument.

Anyway, if you really believe there is no hope ... WHY ARE YOU HERE? GET A LIFE!

Nato et. al.,

Firstly apologies if I am considered one of the whiners in the throngs of the disconented. It has not been my intent in postings to do another Firaxis is evils Firaxis S&cks thread.

I've made my peace (at least to myself) with the Firaxian way of doing things. So my response comes from a cynical yet resigned realism. It is my belief that the TBS genre is at a crossroads.

I believe that Sid/Firaxis's vision of TBS is one that targets the new consumer, not the remnant consumer group of die hards. It remains to be seen if a simplified game will hold the long term appeal that its predecessors did. (and yes I believe it is simplified despite adds of culture and strategic resources. Even these two adds have had their sets of issues resulting in preferentially razing vs capturing and re: strat resources often times causing games to be more a matter of luck when scarce resources are involved. But I digress....)

As to your 3 points, Re: point 1 My point was this. As far as Firaxis is concerned, sales have done fairly well. Critical responses only appear to be coming from boards like this while it has gotten rave reviews from the gaming magazines. Truth of the matter is from a commercial POV the game was a success. (sure it was rushed even by their own admission and buggy as a consequence but the inherent design of keeping it simple is prolly justified in the Firaxian culture/mind set as the appropriate design decision.) A couple of adds for folks already familiar with the game and now wanting more to be thrown in for an expansion pack makes for a decent marketing strategy.

I agree with your point though that all three points probably played a role in the shaping and subsequent release of CIV3. I simply think though that from Firaxis's POV they prolly think they have a success on their hands and more to the point think they were absolutely correct in abandoning the complexity/feature laden approach they were following subsequent to CIV3.

Finally in response to your question:
"Anyway, if you really believe there is no hope ... WHY ARE YOU HERE? GET A LIFE!"

I wouldn't say I have no hope. Matter of fact I do. I just think that perhaps our hope lies with Brian at BHG as he really was the mastermind of CIV2 and SMAC. So again I appeal to Brian to remember his roots and after he completes RON, return to the TBS game approach.

Og

P.S. In any event I still find myself from time to time firing up CIV3 as a kind of diversion. Not nearly as immersible as the others mentioned but a kind of CIVlite alternative instead of playing solitaire. All in all an average to good game but not something I'm gonna have an addiction over as was the case with CIV2 and SMAC. I do keep watching though!

P.P.S. I applaud your attempts to get it through to Firaxis that features and AI can be an attainable goal. My apologies if my cynical realism (i.e. my belief that Firaxis thinks CIV3 is a success as is) throws the proverbial bucket of cold water on your optimism.

[Raion]

If Civ III was Civ II or SMAC or SMACX, then what would some of you do?

Smile!
Or
Not Smile!

So, basically the flavor of CIV III is different, and it is that difference that some contend that a game 'catches on'.

I do not know, but I had to 'catch on' to Civ in the Past, and SMAC.
Especially SMAC with globs of yuck on Planet!
I still am not use to that - as 'its yucky'.

I like the terrain in Civ III better, as not having the 'yuck' factor of SMAC.

[Ogie Oglethorpe]

Originally posted by Raion
If Civ III was Civ II or SMAC or SMACX, then what would some of you do?

Smile!
Or
Not Smile!

So, basically the flavor of CIV III is different, and it is that difference that some contend that a game 'catches on'.

If it was made differently, some would love some would hate it. Some love CIV3 as it now stands and hate SMAC for exactly those differences. (hey thats individual opinion everyone got one right)

My posts were speaking to the larger issue as to what is best for the TBS games future. (I don't know the answer to this)

Is it best to recruit a new generation of fans via the current implementation of CIV3 (IMHO simplified CIV style games) vs. providing more fully featured versions?

On The One Hand you've potentially shortened the learning curve thus allowing better immediate acceptance.

OTOH you run the risk of completely pissing off your core hard core consumer group. I submit much of the reason why there aren't as many complaints vs. earlier days is that some have already started walking away from the game. Notable exceptions being the trolling JT does every so often .

Only time shall tell what the future of the CIV genre will be.