What is more difficult?

[Ming]

Yeah... I do keep saying the same thing. Because this is a matter of opinion. And frankly, I believe the opinions of the people that designed the game more than people that didn't. The Blake quote DaveV posted was just one of many from early discussions on the aggresive AI option.

It's all about balance... the real argument. In aggressive AI, you have civs that will do the tech route and challange you in that area, while the aggressive civs will go the military route and keep you challenged on that side. With non aggressive AI, it allows you to pretty much ignore a key part of the game instead of having to balance military vs tech. It makes the game harder... as stated by the designers.

I can only think that you continue to defend you opinion because that's the option you use, and you don't want to admit that it makes the game easier

[slnz]

All this hype about the challenge led me to take another, closer look at this setting that's supposedly the Holy Grail... Anything that makes Immortal more of a challenge would be welcome since Deity is so obnoxious (feels like Monarch->Immortal is about equal to the Immortal->Deity jump)

Well so far I'm not impressed. The diplo system is even easier to abuse than without Agg AI because of triggerhappy AIs, and while the AI tech pace is faster than I remembered, it's quite unspectacular. Also snagging key wonders is easier than ever since the AIs don't focus on them much at all.

Currently playing an Immortal/Normal/Standard/Pangaea/AggAI game where I got Cavalry and no one else has even Gunpowder or Printing Press. Friendly with all that matter, and about 33% of the top dog's soldiers without building a noteworthy military and just starting a massive military buildup (Cavalry vs Longbows, what a challenge). There is a bigger danger of a runaway AI to form (this setting really seems to favor high unitprob AIs a LOT) but the tech suckiness is a more significant factor. Also got 'Mids, TGL and UoS without even trying (for Taj I had to actually try a bit). Not sure if I'll feel like playing that one out, though some nasty defensive pacts or wartime peaceful vassalage could bring some surprises and drag this one out.

Now one game doesn't prove a thing, but this just...feels like playing half a level below. Agg AI can definitely bring more variance in the game before you can begin to get things under control, though. And if you don't know your diplo it can be annoying. Either way, definitely not a significant change in difficulty, NOTHING compared to rising up or down a difficulty level. I'll have to play a few more games with it to get a more decent feel for it.

[Ming]

I've played some non aggressive AI games which were far easier than the normal games I play at that level. So your "example" doesn't really change my opinion.

Again... you're "opinion" vs that of the people that made the game. I know what side I would bet on.

[slnz]

Again... you're "opinion" vs that of the people that made the game. I know what side I would bet on.

Didn't see that one coming.

[Ming]

[wodan11]

Since Ming keeps quoting from "the designers" without giving the quotes , here are some quotes from Blake (the AI programmer for BtS):

[q=Blake]Aggressive AI means it - not only will they attack more often, they'll also be far more likely to commit to a massive, full scale assault - throwing everything and the kitchen sink at their victim. This can be a risky move but I've decided (for now) that's what Aggressive AI means - they wont hold back in a war, they wont even really worry about defending themselves until a new threat actually presents itself - I mean ideally they'll throw EVERY unit except 1-2 per city into the war effort - they'll take a devil-may-care attitude to war and will "Cross that bridge when they come to it" when it comes to backstabbers (Ruthless AI's may be a more descriptive term).[/q]

A ramification of what he's saying here is that, under aggressive AI setting, all you have to do is convince (or wait until) an AI attacks another AI, and then you can invade either one of them with little opposition.

That's not really directly in our line of discussion but it points out a vulnerability of the aggressive AI setting that the human can (knowingly or unknowingly) capitalize upon.

[q=Blake]You only need one Gandhi to win a cultural, or one Mansa Musa to win space. I do suspect that when a peaceful AI starts next to an aggressive warmonger, the Warmonger will kill the peacemonger unless they end up religious allies. But it's fairly easy for a warmonger to get deadlocked with another warmonger, while an unmolested peacemonger (due to allies or whatever) can just tech off into space.[/q]
This supports Ming's assertion, but only when Gandhi (etc) is isolated on an island continent or something.

[q=Blake]I have made the AI's typically use the same dagger strategy once they get attacked (the likes of Gandhi may remain pacifist though...) - so they start spitting out units to keep the attacker honest, so the question is whether the warmonger can hit with enough force and have enough momentum to take out most of the peacemonger's empire before they properly mobilize for war. Twice now I've had AI Alexander smash right through a peacemonger and take their capital in a matter of a dozen turns or less. On the other hand playing as Asoka I had Monty positively crush me despite a significant military build up on my part, these aggressive brutes are not easy to handle without being agg/cha/pro yourself, and at some point it becomes a matter of "sh*t happens", if you spend enough on military to actually defeat the attacking army you'll be screwed anyway, it can be a better strategy to just hope or manipulate them into attacking someone else, befriending such a warmonger is quite an effective way to gain security and it's a heck of a lot cheaper than fighting them. Since the dagger strat does contain a component of being less inclined to pursue religions the daggerers will usually end up with someone elses religion meaning that at least one civ will probably be reasonably safe from their aggression. [/q]

The above tends to run counter to Ming's assertion (summarized and badly paraphrased: that non-warmongers will keep up with the human in tech parity, while the warmongers will keep up with the human in military). Since the AI can and does fight amongst itself, some if not all of the non-warmongers will fall prey to the warmongers, leaving the human primarily with warmonger enemies and less of a tech race.

[q=Blake]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.[/q]

Ming, was it you who said your games usually end early (and thus you never get to the renaissance or modern eras)? If so, shouldn't we take your experience with a grain of salt? If not, I apologize.

... Let me pursue this logic for a moment. If someone's play style is early and repeated war, then that human doesn't really experience the issue of whether the AI can keep up technologically on aggressive AI vs a human who does focus upon tech. What if, instead, the player did focused, strong, but limited warmongering? Focused upon defense, gaining key territory, and building an infrastructure and tech powerhouse? Meanwhile the aggressive AI setting has the AIs throwing units left and right but not really focusing upon infrastructure or tech. The natural result is that if the human is able to survive and win that key early war, the human will easily out-tech the AIs. Every single time, probably.

Here's more from Blake:

[q=Blake]I'm 100% responsible for the AI in BTS.

A naturally militaristic AI like Alexander or Monty will still make a respectable military effort under normal settings, in AI vs AI wars, it's all relative anyway. Some AI's have to be bad at defense, so others can invade them. In BTS the AI are far less "samey" in their strategy - in short they can pursue goals, but obviously in min-maxing their metagame they sometimes make themselves extra vulnerable to be invaded, it's the price of not being samey/predictable.

Aggressive AI no longer causes the AI to have a relationship penalty with humans. Basically it can be said that the AI expects things to get aggressive. The pacifist AI's aren't actually that much more likely to declare war, they just keep larger armies on hand as to not be easy victims. The naturally militaristic AI's go crazy. In any case if you neglect your army, any AI will notice and with it's larger power will be more likely to declare war and come for you. It is more likely you'll get declared upon, especially if you don't change your playstyle...

The default AI is a bit of a sandbox, you can employ the strategy you want and the AI may interfere with your plans... but on Aggressive AI, the AI can DICTATE your strategy! If Alexander is going to invade you, then you damn well prepare an army or you're going to taken out of the game! Even with the best prepeardedness if you fail to avoid a dogpile you're probably a goner.

Note that Aggressive AI, due to spending more on units, techs significantly slower than the default AI, if you can somehow stay out of the crosshairs it's actually easier to win peacefully - the default AI can be a speed demon when it comes to research.[/q]

I think that pretty much says it right there.

[Sir Og]

I also think that aggressive AI makes the game easier. Agg AI produces a lot more AI vs AI wars which is one of biggest factors contributing to an "easy" game.

The only way that Agg AI can make the game harder is to make one AI become really powerful but this is quite rare.

I haven't played enough with the tech options to comment on them.

I play mostly without vassals because I hated when AIs that are friendly to me vassalise the AI that I am about to crush and I end up in a war with two civs simultaneously. I guess that would mean that vassals were making the game harder for me
I should give vassals a try again. Maybe after all the patches vassals won't be as frustrating as they were.

[slnz]

Thanks wodan, that was a very interesting read. Also seems to correspond quite well to my (admittedly somewhat limited) experiences about it. I can see it has its place but does add some exploitability via diplo and battle tactics. And definitely useful if one plays SP as a training ground for MP.

[wodan11]

Oh and one other thing Blake said that I discovered as I was sifting through threads (don't know why I didn't cut&paste it) is that aggressive AI hurts AI tech progress more on lower levels. The reason for this is that the AI has a huge priority on building units, but doesn't get the hammer modifiers that it gets on higher levels. So, a proportionally larger amount of the AI's resources are spent on making units.

[rah]

I did some testing. (which doesn't prove anything but)
I can beat emp with regular ai with some regularity.
With agressive ai, I rarely beat emp. (and I mean rarely)

But this is also with no tech trading and no vassals so take it for what it is.
I'm sure different combos could change the results.

[wodan11]

Without knowing your play style rah it's hard to draw any conclusions from that.

[Ming]

Without knowing your play style rah it's hard to draw any conclusions from that.

Gee... the same could be said about other people posting their results

[Ming]

Well wodan... I still disagree with your opinins/conslussions

[q=Blake]Aggressive AI means it - not only will they attack more often, they'll also be far more likely to commit to a massive, full scale assault - throwing everything and the kitchen sink at their victim. This can be a risky move but I've decided (for now) that's what Aggressive AI means - they wont hold back in a war, they wont even really worry about defending themselves until a new threat actually presents itself - I mean ideally they'll throw EVERY unit except 1-2 per city into the war effort - they'll take a devil-may-care attitude to war and will "Cross that bridge when they come to it" when it comes to backstabbers (Ruthless AI's may be a more descriptive term).[/q]

RESPONSE
A ramification of what he's saying here is that, under aggressive AI setting, all you have to do is convince (or wait until) an AI attacks another AI, and then you can invade either one of them with little opposition.

That's not really directly in our line of discussion but it points out a vulnerability of the aggressive AI setting that the human can (knowingly or unknowingly) capitalize upon.

But what you are missing here is that the AI can be throwing that balls out attack at YOU. You need to be balanced and actually have a real military or you will be toast. Having two or three units in every city and a small mobile defensive force like you can have in a non agressive game just isn't going to cut it.
You need to actually waste some resources to defend yourself. It's much tougher to be balanced than to focus on simply out teching the AI.

And as far as AI vulnerablity... The same strategy works even with a non agressive settings. It's always good to attack somebody when their attention is elsewhere.

[q=Blake]You only need one Gandhi to win a cultural, or one Mansa Musa to win space. I do suspect that when a peaceful AI starts next to an aggressive warmonger, the Warmonger will kill the peacemonger unless they end up religious allies. But it's fairly easy for a warmonger to get deadlocked with another warmonger, while an unmolested peacemonger (due to allies or whatever) can just tech off into space.[/q]

RESPONSE
This supports Ming's assertion, but only when Gandhi (etc) is isolated on an island continent or something.

HAHAHAHAHHAHAHAHAHAAHA... even Blake points out that they don't need to be isolated. I've seen so many aggressive AI games where a peace loving civ is next to an aggressive civ, and they are best buddies. While the aggressive civ is busy attacking everybody else and trying to win that way, the peaceful civ is just out teching everybody else. You are trying to minimize the key reason why aggressive AI is tougher by implying it's a rare event.

[q=Blake]I have made the AI's typically use the same dagger strategy once they get attacked (the likes of Gandhi may remain pacifist though...) - so they start spitting out units to keep the attacker honest, so the question is whether the warmonger can hit with enough force and have enough momentum to take out most of the peacemonger's empire before they properly mobilize for war. Twice now I've had AI Alexander smash right through a peacemonger and take their capital in a matter of a dozen turns or less. On the other hand playing as Asoka I had Monty positively crush me despite a significant military build up on my part, these aggressive brutes are not easy to handle without being agg/cha/pro yourself, and at some point it becomes a matter of "sh*t happens", if you spend enough on military to actually defeat the attacking army you'll be screwed anyway, it can be a better strategy to just hope or manipulate them into attacking someone else, befriending such a warmonger is quite an effective way to gain security and it's a heck of a lot cheaper than fighting them. Since the dagger strat does contain a component of being less inclined to pursue religions the daggerers will usually end up with someone elses religion meaning that at least one civ will probably be reasonably safe from their aggression. [/q]

RESPONSE
The above tends to run counter to Ming's assertion (summarized and badly paraphrased: that non-warmongers will keep up with the human in tech parity, while the warmongers will keep up with the human in military). Since the AI can and does fight amongst itself, some if not all of the non-warmongers will fall prey to the warmongers, leaving the human primarily with warmonger enemies and less of a tech race.

So... Yeah, some peaceful civs will die. That's why they call them peaceful The key points Blake is making here is that nobody is safe from an aggressive AI, not even you. He implies that it's probably a good idea to stay on the good side of the aggressive civs... but guess what, it's easier for the AI's to stay friendly with each other than it is you. Plus, as he points out, some peaceful civs will have good relations and survive to continue their teching. Again, a balance and much tougher game.


[q=Blake]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.[/q]

RESPONSE
Ming, was it you who said your games usually end early (and thus you never get to the renaissance or modern eras)? If so, shouldn't we take your experience with a grain of salt? If not, I apologize.

... Let me pursue this logic for a moment. If someone's play style is early and repeated war, then that human doesn't really experience the issue of whether the AI can keep up technologically on aggressive AI vs a human who does focus upon tech. What if, instead, the player did focused, strong, but limited warmongering? Focused upon defense, gaining key territory, and building an infrastructure and tech powerhouse? Meanwhile the aggressive AI setting has the AIs throwing units left and right but not really focusing upon infrastructure or tech. The natural result is that if the human is able to survive and win that key early war, the human will easily out-tech the AIs. Every single time, probably.

First, I've played many games where I've played the peaceful builder, doing exactly as you say... This is pretty easy to do with a non agressive AI since you will never see a 100 unit SOD
Even at the higher levels, it's pretty easy to out tech the AI with the non aggressive AI. With Aggressive AI, you still have to prepare for the 100 unit SOD. And sure, the aggressive AI's will be behind in tech (unless they have taken over a few civs and have 4 times the number of cities you have), the peaceful civs that have survived will still be teching with you. Again, having to balance the two is what makes a tougher setting.

[q=Blake]I'm 100% responsible for the AI in BTS.
A naturally militaristic AI like Alexander or Monty will still make a respectable military effort under normal settings, in AI vs AI wars, it's all relative anyway. Some AI's have to be bad at defense, so others can invade them. In BTS the AI are far less "samey" in their strategy - in short they can pursue goals, but obviously in min-maxing their metagame they sometimes make themselves extra vulnerable to be invaded, it's the price of not being samey/predictable.

Aggressive AI no longer causes the AI to have a relationship penalty with humans. Basically it can be said that the AI expects things to get aggressive. The pacifist AI's aren't actually that much more likely to declare war, they just keep larger armies on hand as to not be easy victims. The naturally militaristic AI's go crazy. In any case if you neglect your army, any AI will notice and with it's larger power will be more likely to declare war and come for you. It is more likely you'll get declared upon, especially if you don't change your playstyle...

The default AI is a bit of a sandbox, you can employ the strategy you want and the AI may interfere with your plans... but on Aggressive AI, the AI can DICTATE your strategy! If Alexander is going to invade you, then you damn well prepare an army or you're going to taken out of the game! Even with the best prepeardedness if you fail to avoid a dogpile you're probably a goner.

Note that Aggressive AI, due to spending more on units, techs significantly slower than the default AI, if you can somehow stay out of the crosshairs it's actually easier to win peacefully - the default AI can be a speed demon when it comes to research.[/q]

RESPONSE
I think that pretty much says it right there.

Yeah... it does... It totally supports everything I'm saying.
Straight from his lips... "The default AI is a bit of a sandbox, you can employ the strategy you want and the AI may interfere with your plans... but on Aggressive AI, the AI can DICTATE your strategy!"

And while you point to his line... "it's actually easier to win peacefully" the key word here is IF I don't think I've ever argued that the AI techs better in aggressive AI. What I have said is that some AI's, the peaceful ones, will keep teching while the aggressive civs will be keeping you honest with their military. Again, a more balanced and more difficult game.

So thanks for pulling up more of Blake's posts. It just made my job easier

[Blaupanzer]

At Prince and Monarch levels, playing with my style, I have recently discovered (based on this discussion) Agg AI actually hinders the AI to some degree. I was surprised as I too had followed Blake's discussions and suspected he meant the AI's "disadvantage related to techs and wonders" applied to the levels below Prince. Appears "lower levels" is actually Monarch and below.

My play style is early builder (with lots of soldiers) until Construction and diplo victory off. Important to stay ahead of AI in tech and priority Wonder is Oracle. Once I have cats, go for one early conquest with massive buildup all the way through the end so that no other AI vassals the victim without jeapordizing their own fate. From there I prioritize gunpowder, cuiraissairs, Democracy/Steel in either order. Key wonder is Statue of Liberty. With Astronomy also completed in that set, I then launch round 2 war while picking up rifles and Communism and completing my decision on which victory to pursue. I have usually laid the groundwork for cultural, space and military victories.

In agg AI games, IF my schedule did not get diverted by a nasty double SOD jump, victory is mine as quick as I play it out. Only humans in charge of the AIs could save the game for any of them.

With Aggressive AI off, at least one AI is available to compete in each of those categories and somtimes more than one. The cultural and space race competitors are often vassals of the military types who keep up in tech by being fed techs by these vassals. Vassals can win the game. Ghandi is a really nasty cultural competitor when protected by one of the really nasty military types like Alexander, or Hanibal. Augustus, in one game could have gone any of the three directions too as I arrived at that point. With Agg AI off, one AI had kept up with me in all three categories! I don't remember that ever happening in na Agg AI environment. (I helped him choose to go for military by crushing him.) Admittedly I did not see any of the SOD progression of Agg AI (40, 60, 100, +++). So for a conquerer on Prince the game is skewed in challenge to Agg AI, but for a builder the challenge might actually lie with non-aggressive. I can see that Blake's actual words say that, but would not have caught that nuance without this thread. Way to go Ming!

[rah]

I'll go for oracle for COL
I'll go for GL or PYR, depending on land.
If the opportunity presents I'll take out a close neighbor.
I'll wait for someone to come at me and then take them out.
AND that was the big difference with it off. No one every came with any meaningful stack so I was able to tech away until my tech lead was so good that the rest were toast.