AI Aggression levels (please discuss here).

[Uncle_Joe]

I dont know, I fail to see what is being a 'wuss' about wanting to play the game as it was apparently INTENDED to be played rather than going balls out to conquer the AIs every game.

If thats fun for some folks, there is the 'always war' option too. I want a challenging but varied AI. I believe the stock AI provides good variety, but its execution of its strategies are somewhat weak.

Warmongering is fun at times, but I can do that in plenty of games. Civ4 is fun because I DONT have to conquer the world to win or fight numerous wars, even have the most powerful military in the world (as long as you play your diplomatic cards right). I can try those things if I choose, but the military is NOT supposed to be the focus of Civ4

By your own admission, you say that launching the pre-emptive strikes and 'rushing' the AI is a good strategy because the AIs have a hard time against that. Isnt that wussing out instead? Isnt that taking the easy way out?

 

[Roland Johansen]

Come on, the 'wuss' thing was clearly a joke.

 

[Elandal]

I've mostly quit games midway lately when it's been clear that I need a stack of at least 40 units (10-15 of them cats) to get a single city. Yes, I could've built a supersod and gone on rampage. But stacks of that size make the UI crawl (takes a few seconds to simply select one unit out of the stack), so I'd wear myself down faster than the WW hits my civ..

Anyway, I'm not playing multiplayer games. I don't really want to play against opponents that exist to make my life miserable (as in, other humans or an AI using the same strats as a human would with the same efficiency). So I definitelly will be using the wuss-settings. Going down difficulty levels doesn't really work, as I already did that - I think the AI production and research bonuses on Prince are a bit too low for me already. For now, it's rather about needing umpteen units to go to any war at all.

For now, I'll stay and wait for the unit spam reduction build. Maybe if the AI can get around with fewer units, maybe use them better - maybe if I can go to war with a stack of a dozen units again like it used to be, only now needing to use my stack well and have it backed up by active defense ministacks..

 

[Uncle_Joe]

Come on, the 'wuss' thing was clearly a joke.

Yeah, I know, but its just indicative of the mentality that if you dont want to be playing with hordes of units against hordes of units that you are a wuss. At least in my case, I just find it tedious, not intimidating or 'hard'.

Elandal sums it up perfectly in his post:

For now, it's rather about needing umpteen units to go to any war at all.

 

[jkp1187]

I can't really see a problem with making Aggressive AI the setting for "I'm brutal warmonger and these AI's put up precious little resistance" setting... a major complaint about Aggressive AI is that it makes a "AI vs Human" mentality - well that's on the way out - so if you're a thug and want to play with other thugs, turn on Agg AI. IF you want to play wolf amongst the lambs, well play on Normal... you'll have the choice.

This seems like a fair way of splitting the baby.

 

[Blake]

I dont know, I fail to see what is being a 'wuss' about wanting to play the game as it was apparently INTENDED to be played rather than going balls out to conquer the AIs every game.

For now, I'll stay and wait for the unit spam reduction build. Maybe if the AI can get around with fewer units, maybe use them better - maybe if I can go to war with a stack of a dozen units again like it used to be, only now needing to use my stack well and have it backed up by active defense ministacks..

While I am making the reduced spam I'm sure you recognize that an AI with 5 axemen, however well it uses them, cannot beat - or even resist against - a human who brings 15 axemen to the party and I've yet to really see an AI "out mass" a good human, once a human gets their economy rolling. When it comes to the massing game, "the humans started it".

Regardless of how the game was intended to be played - the fact is that massing a military and conquering the AI's is an EXTREMELY straightforward route to victory

The Wuss jab is justified because it IS forcing the AI to build an inadequate military as an imposed variant.

 

[glokkonn]

The Wuss jab is not justified at all.
There is no need to hurl insults at your customer base.
Not a good business strategy.
Not even in "jest".

You seem like you have a plan of action already set in play.
Make it happen.

Make the AIs effectively use the units they make.
The agressive AIs will attack effectively (and bribe for help if they are having trouble).
The builder AIs will make a decent amount of units and defend agressively (and bribe for help when attacked).

With luck this increase in AI playing strength will knock everyone down 1 or even 2 difficulty levels (hard to mount an attack on neighbor 2 when neighbor 1 declares war and sends in a powerful SOD and actually uses it well).

This will partially solve the massive number of units problem because 2 levels down the AIs have lower production bonuses.

I recommend only making one version of the game.
Just the standard version.
Once you get this going and in a state that you like then you can add in a more macho game. With your programming expertise this would likely only be beatable on Warlord difficulty and maybe not even on THAT level.

You have the ability to make aggressive AIs attack well, and builder AIs defend well. Do that first. Tweaking can come afterwards.

And never ever insult your customers.
Not even in jest.

 

[Elandal]

I don't see any problem with the "wuss" jab. If someone feels insulted, their problem.

In any case, the unit massing ends up with more unit massing that forces more unit massing ad infinitum. I don't find that fun in any way or shape - it's just more units. Going to war with 40 units instead of 20 makes the game a lot slower without adding any new tactical aspect to it. And if the human can mass 40 units, then certainly the AI should mass even more - double the defenses? Forcing the human to again double his attack stack.. No, there's nothing in that route, nothing but more units.

OK, there's one change in game the massing causes: at some point, it's impossible to attack due to economic effect of too many units. At which point there should be no more war (it's cold war, everyone has too good defenses for anyone to attack) and thus the one with most land (and best economy) wins when everyone else gets cobbled by the expenses.

So, the humans started massing units first. That's because the humans were the first to go to aggressive war, and the only way to do it is to have attack stack stronger than expected defense stack. But if the response is to mass stronger defense stack, then the result is what we have: more units on all sides. If taken to the extreme, the only possible wars are dogpile wars - no single civ can afford an offense alone anymore.

What this does lead to is that I think the AI should indeed stop massing at some point. Yes, the human can continue where the AI is simply forced to stop. And thus the human will win. Win what? A war. Not the game. If the stack needed is huge, then the economic disadvantage to the human player is huge as well, and the whole balance comes down to whether that short term economic disadvantage can still gain enough land to be turned into later economic advantage big enough to still win.

As I believe that in SP games the AI is an obstacle, not an opponent, I feel that is acceptable. The AI will lose the war because it's not meant to win. If balanced properly, the human player can then use the land gained (which must be a lot) to make a comeback and still race to the win. If that is not possible, then the game has failed.

 

[Elhoim]

And never ever insult your customers.

He isn´t selling us anything, so we are not his customers.

Besides, I prefer to be a wuss than an insecure boy that needs to build a BIG stack to compensate the lack of other atributtes. j/k

Anyway, I´m sure that Blake will find a way to make the AI use the units in an intelligent way (like a human do*), and also build an army balanced to its needs.

* "The victorious general first seeks to win and then go to war, while the defeated general first go to war and then seeks to win". I only attack when the odds favor me, and force the enemy to attack me in a place where I have the advantage. And I´m sure that Blake follows this quote when designing the AI.

 

[Elandal]

If we assume that the defender always has an advantage (should have - defender has roads to move units around, won't get WW from fighting in own cultural area, has city defenses..) then the attacker always needs to have more units than the defender - barring a big technological advantage. We can say that the attacker needs twice the units of the defender in a city capture attack. Sometimes the figure is smaller, sometimes larger (it's been said that to take a fort you need ten times the numbers - this would be insane in civ), but 2x isn't usually that far off the mark.

So, attacker is always at an economic disadvantage: more units = more expenses, and attacker is paying supply in addition to maintenance, where defender is only paying maintenance. Of course the attacker can pillage to gain some gold, and when playing with small numbers of units this may well have been enough - pillaging a farm doesn't pay for a 20 unit stack though.

As I noted in my post above, the short term economic disadvantage grows by the number of units needed. This depends on the overall economies, with later eras having higher commerce per population point. A unit costs exactly the same in maintenance and supply in both early and late eras, so it's expected that in later eras the number of units around is larger. Paying 10g supply for an attack stack is crippling in ancient era, but pocket change in modern era.

Assuming the attack is simple march to city, bombard, suicide sieges, send in the CR units, send in the mopup units, stay for a few turns to heal and wait for the reinforcements (replacements for lost units as well as garrison for the acquired city), then the only thing affecting the outcome is the number of units. And raising the number of units leads to, at some point, a short term economic disadvantage too large to allow for offensive of this type anymore.

A stalemate of this kind means that the starting position is everything. If you can rex and secure backlands for later settling, you're in for the win. If you can't, then you lose. If there's a runaway somewhere, you lose. If you fail in one war, you lose. If you're smaller than your neighbours, you lose.


So what alternatives are there really?

There's of course diplomacy to make sure all wars are dogpiles. A single civ defending against multiple civs should have no chance here - each of the attackers goes for one city, takes the short term hit, but the defender gets broken. The one who loses the war no longer has economy to keep it in the game - it's going to fall behind in tech and will then be suspectible to being attacked with superior units.

Then there's the concept of smarter war. But I don't know how that can be done. CIV isn't a wargame, so it has very limited and simple warring system. Mixed stacks go only so far in smarts - the basic warring is still the same, only now requiring more units.. But if there's a solution, it must be somewhere in the "smarter warring" area.

 

[Arlborn]

What this does lead to is that I think the AI should indeed stop massing at some point. Yes, the human can continue where the AI is simply forced to stop. And thus the human will win. Win what? A war. Not the game. If the stack needed is huge, then the economic disadvantage to the human player is huge as well, and the whole balance comes down to whether that short term economic disadvantage can still gain enough land to be turned into later economic advantage big enough to still win.

As I believe that in SP games the AI is an obstacle, not an opponent, I feel that is acceptable. The AI will lose the war because it's not meant to win. If balanced properly, the human player can then use the land gained (which must be a lot) to make a comeback and still race to the win. If that is not possible, then the game has failed.

I just cant believe on that, sorry. Did you read the thread in this sub-forum called "What is the Better AI'' ?


But I liked your last reply to this thread and somehow I agree.

 

[Elandal]

Yes, I've read that Arlborn. I don't see it reading as "the goal of the project is to make SP into MP with AI opponents that equal human opponents". I do understand if some people want that, and if it's possible to separate these two aspects with the Aggressive AI switch, all the better. The goal is a holy grail of computer strategy games, definitelly worthy reaching for. But it's not a real single player game goal, it's an MP goal.

I'm just afraid of that the project ends up making the AI not smart enough to play against humans but rather massing so much that the game loses focus completely. Of course it can be argued that more units is a smart move, but I've already outlined how I see that road going and thus can't see that as enhancing gameplay.


Note that I haven't ever played civ multiplayer. I've played very few games in multiplayer environment at all - a couple of lan parties a decade ago, a bit of fps'ing at the office after closing the door, some mmorpging.. Most of my multiplayer strategy game experience comes from tabletop, not computer games, and even there it's limited to a maybe 10 years or so and nothing at all recently.
So I do indeed focus on the single player experience knowing that if I want multiplayer I want human opponents and I would play with my friends, not random ****s in the 'net. And from my past experience playing MP strategy games with my friends I know very well that the wars would not be decided by units (even if those are needed too) but rather by diplomacy.

 

[Quagga]

I think the difference between SP and MP is that in MP all of the players know it is a game and behave accordingly. (Well, all the humans, anyway.)

In SP, the AIs should, and generally do, behave like civilizations. Only when they won't trade you Rocketry do they go out of character ("We want to win," or whatever the exact message is. I see that as an error in judgement on the part of the developers.)

Given that, although the AIs should take actions that lead to a win, they must be guided -- perhaps hampered -- by their natures. If that approach is not taken, the risk becomes devolving all AIs into a win-at-any-cost machines, just like when humans play MP. I think most SP players won't like that at all.

 

[Roland Johansen]

We are customers? I didn't know that Blake and Iustus were getting money for this. Man, this sounds like a good deal for them. I should also go into the modmaking business if it pays so well...






(the above is a joke, not to be taken seriously.)

Besides, I prefer to be a wuss than an insecure boy that needs to build a BIG stack to compensate the lack of other atributtes. j/k

Also funny.

 

[Uncle_Joe]

OK, there's one change in game the massing causes: at some point, it's impossible to attack due to economic effect of too many units. At which point there should be no more war (it's cold war, everyone has too good defenses for anyone to attack) and thus the one with most land (and best economy) wins when everyone else gets cobbled by the expenses.

Elandal hit the nail right on the head and my 3 recent games all back that up. In my latest, I share a continent with Toku (playing as Ottomans). We have a Mexican standoff for while and we split the continent about 50/50. In the late Classical, he DoW's me. We fight a relatively inconclusive war (although 50+ units are killed on both sides). We have a period of piece and buildup and then he attacks again in late Medieval (although I know he'll probably do it so I have built up too). This time, I manage to take one of his cities by the end of the mess and again, we both lose copious amounts of units.

Somewhere in the middle of our war, contact is made with the other continent. One of the AIs (Hatty) is a vassal to the Vikings. Another AI is alone on their own continent. And the last AI is (Vicky) is best buddies with Ragnar due to religion. The result? All of the other AIs (except Hatty) have nearly double the points of Toku and I and are miles ahead on the tech tree. The reason: they didnt go into the ridiculously expensive 'death spiral' of going to extended warfare.

Elandal is indeed correct and it was something I had sensed, but never really put the finger on. When the numbers of units are as big as they are, the cost for going to war is IMMENSE. Even a 'successful' war will often leave you behind others who are at peace (but still building up their massed defenses). When stacks of 20+ are clashing, you could be lossing 1000s of hammers of production in a single turn!! And you MUST replace those units or else be weak and eventually overrun.

It goes back to what I first stated...the massed militaries cost so much that they mute the other effects in the game. And in the early game, you might gain some success relative to the cost of those troops by attacking a neighbor. But from Medieval on, the cost of going to war is just crippling. And thats why the more warlike AIs always tend to peter out and fizzle...they come out strong and may take out a rival or so, but then they hit that wall of expense vs payoff with the huge militaries and they crash.

If you happen to be one of those involved in the early war with a warlike AI, you'll struggle to remain alive...there is no doubt that the AIs can prosecute wars much better now, but the cost is astronomical. But its likely you'll never recover from the early wars (assuming you win). And on the flip side, if you manage to stay at peace (through diplomacy or simply building a bigger stick), then you'll likely cruise to victory because the AIs will either destroy themselves and then fizzle or else poorly balance econ vs military and also fall behind.

I also think that Elandal is again correct on the consequences of having the AI not build as many units. He is correct that yep, they may lose a war to an aggressive human player. But thats fine. As long as the cost of prosecuting that war is balanced to the gains, there is NO PROBLEM. I agree that in 2.08, the cost/benefit ratio of attacking many of the AIs is very lopsided in favor of attacking (Blake's 'pinata' effect). But 1/25 is the polar opposite. It doesnt need to be anywhere NEAR that high. As long as the AI makes a 'credible' defense, its fine. 5-6 units per city will indeed lose to human stacks of 15. But that doesnt mean that the AI should be building 15-20 to resist. That just encourages the human to build 20-25 and the cycle begins again, only with the cost of warfare overshadowing all else.

At some point, it just has to be accepted that the 'better' idea is to simply make the cost of an attack the deterrent. And I believe that 5-6 units plus some reserves would do that. It would no longer be the cakewalk that it can be in 2.08, but a real cost will still be assessed (15-20 units still isnt cheap when the rest of the world is teching/building infrastructure).

So, FWIW, my opinion is that the AIs should build perhaps 2x-2.5x their 2.08 levels. If at war or actively building to attack, that should increase obviously, but for 'peacetime', the AI should cap their unit production and continue to develop. If one or more is killed off, so be it...the rest will still be presenting a challenge by continuing to progress.

 

[Quagga]

Somewhere in the middle of our war, contact is made with the other continent. One of the AIs (Hatty) is a vassal to the Vikings. Another AI is alone on their own continent. And the last AI is (Vicky) is best buddies with Ragnar due to religion. The result? All of the other AIs (except Hatty) have nearly double the points of Toku and I and are miles ahead on the tech tree. The reason: they didnt go into the ridiculously expensive 'death spiral' of going to extended warfare.

This happens without Better AI on the Continents map. Yes, maybe it's worse with more unit build up, but if there is a peaceful continent and a warring continent, the research will be imbalanced.

 

[Uncle_Joe]

This happens without Better AI on the Continents map. Yes, maybe it's worse with more unit build up, but if there is a peaceful continent and a warring continent, the research will be imbalanced.

No doubt it happens even in 'stock'. But the point is that the disparity was HUGE (much larger than what I routinely experienced in 1.0 or 2.08). Which led me to exactly where Elandal already was...the insane cost of war when units are dispatched 20 at a time while the costs of 'peaceful' development remain unchanged.

Thats what I was trying to say before in the old thread. The balance in the game between 'guns or butter' has been completely thrown out of whack.

 

[Arlborn]

This happens without Better AI on the Continents map. Yes, maybe it's worse with more unit build up, but if there is a peaceful continent and a warring continent, the research will be imbalanced.

Yep. Now I dont see the reason why the building of units made you have war with Toku. He would declare on you anyway, just perhaps now its a little harder to take him off of you?

I'm really curious to keep playing that game I started with the last version where 2 AIs didnt give me any resistence, and I had only 5 melee units(4 axeman and one spearman) and 2 horse units(the UU of Mongol). And I only needed the 2 horse units againt Cathy..And Ramses I pretty much would only need them also..
Now I gotta wait untill I make contact with other continnets(6 AIs left), but untill now, no mass spam of unit by the AI, I checked with chiplote and with only 3 defensive units and like 10 offensive units(axeman-swordman-horse archer era still) I had twice more power than any other(that includes Monty)...

 

[Uncle_Joe]

Yep. Now I dont see the reason why the building of units made you have war with Toku. He would declare on you anyway, just perhaps now its a little harder to take him off of you?

I think you missed the point. The building of units didnt CAUSE the war with Toku. But the cost of fighting those wars was astronomical compared to 2.08. Instead of 10-15+ unit dying (to the tune of maybe 400-600 hammer lost), 40-50 units were dying (about 1600-1800 hammers worth). Of course you have to rebuild the losses (or else your military rating sucks and you get attacked again). So, if you are spending 1600-1800 on simple replacements for war that netted you nothing you can see how that hurts compared to those without that expense.

The cost of development and techs stay static, but the cost of military sky-rockets. Imagine that instead of the AIs building more units and the players following suit every game that military units simply cost 4-5x what they do now. So, the AI would still be defending with 2-3 units per city and attacking stacks might be 4-5, maybe 6 unit. Thats basically what this emphasis on unit production is doing...its simply multiplying the cost of military units by a large margin (400% or so would be my guess).

I just dont see how anyone can look at the military expenses going up while the rest of the costs stay static and NOT think that something will be knocked out of whack in the game model.

 

[Blake]

Something else I want to note is that my design goal is something like a "friendly" Multiplayer FFA game - assuming all players are of equal skill, only one of them is going to win (barring moral victories) - if you are one of the players in that FFA game, you should only have a 1/N chance of winning (where N is the number of players). Much the same should be true in a game against the AI at a suitably hard level.

I have no intention of making an AI you can win against every single time, I want an AI that will beat you and that will try to win itself (I'm basically committed to removing all anti-human bias - so all AI's should be subject to the 1/N rule, without some of them deliberately choosing victory-denial strategies).

Where does the "Better" come into it? Well, I also want balanced gameplay - a multiplayer game is balanced superbly despite the "1/N" thing, but an Immortal 2.08 game is not balanced the same way - what's the difference? Well when the AI uses unbalanced strategies, the humans need to use the opposite strategies - if the AI's always neglect defenses and tech mega fast, the human needs to conquer them, every time. Even if the victory odds come out to 1/N, it'll still only be possible to win if you pursue the "One Right Strategy".

What I want, is 1/N odds where you only get those 1/N odds if you correctly adapt to the situation - where it's not about pursuing the "One Right Strategy" but making the right moves and hoping for some lucky breaks. So at the ideally hard difficulty, you should sometimes be able to win by turtling and teching - the AI should not be able to tech so fast that it can always win such a tech race.

You can obviously get better than 1/N odds by playing a lower difficulty level.

Most players seem to want approximately (N-1)/N odds - that something bizarre or extremely unlucky must happen for them to actually lose. I could quote someone saying that Deity should be winnable every single time. I disagree, those who disagree with me, can deal with it. The goal of this project is not and has never been making an AI you can beat every time.

To put it another way, you'll get shafted sometimes.

And I know that in the original game, the AI provides "Whack a mole" gameplay - quite literally in their war strategies. The AI provides some resistance to your victory and you need to whack them if they get frisky, but they aren't really playing the game. I wouldn't say that was the design intent so much as limited time to program the AI, plus limited understanding of strategies at the time the AI was programmed.