AI Aggression levels (please discuss here).

I've noticed something similar with this build...top tercile in power but a lot of ganging up on me. My guess was that the AI has been bribing other AIs to go to war against me, based on the timing of the declarations. Is that possible in your case? I found I couldn't bribe anyone to join in on my side, so I've modified my "avoid adopting a religion" diplomatic strategy (which coupled with a strong military used to allow me to avoid wars altogether) to one where I try to adopt the religion of a few hopefully stalwart allies. It seems to have worked out better, but I don't have a very large sample size so it could be coincedence.

Darrell

Sharing borders is also a pretty prominent factor. If you turn on chipotle, you can see how the AI feels about going to war with each civ, to see if you really are on the top of the list, or if you are just unlucky in the random rolls.

-Iustus

My guess is that it was a combination of factors in my game. Of course the aggressive AI setting caused the AI to declare many wars. I don't know if the betterAI team has already 'fixed' the negative diplomacy modifier of the AI versus the human player at this setting. But if this hasn't changed yet, then that could be part of the cause as well.

The reason that they picked me the most was at first because I was weaker (you start weaker at emperor level than the AI nations who start with some units and some military technologies). Next a religion spread among a number of other civilizations but it hasn't spread to my lands yet. So some of them share a religion and I have no religion. I could pick a different religion, but then they would hate me even more. So they like eachother more. There is one nation with a different religion but most of the AI civilizations don't know him.

Another reason is that I'm positioned at the center of these nations. So I'm bordering all of them. That means that if one of them declares war on me, then I'll be an attractive target for the next.

There's not a lot that I can do about it as long as their religion doesn't spread to my lands, so it feels a bit unfair. But it is actually quite reasonable, except that I've been quite unlucky with the position and the spreading of the religion.

I don't know if I will continue this game now. I've survived their determined attacks and have outexpanded them because the AI expanded too slow in this build. So I'll probably start a new game with the new build. I expect to win the present game because I have more than double the amount of cities of my strongest competitor.
 
I found I couldn't bribe anyone to join in on my side, so I've modified my "avoid adopting a religion" diplomatic strategy (which coupled with a strong military used to allow me to avoid wars altogether) to one where I try to adopt the religion of a few hopefully stalwart allies. It seems to have worked out better, but I don't have a very large sample size so it could be coincedence.

Darrell

Adopting a religion has ALWAYS been a very important part of diplomacy. In fact I tend to say that going No-State-Religion is a handicap because it makes it so darn hard to get allies in wars.
 
Just played a little with the last version (fixed war declaration percentage) and it definitely feels too cuddly and peaceful. In the whole game (terra, standard size, monarch) there were only two AI-wars (both times Julius Caesar declared on Ragnar) and I could cruise straight to infantry and tanks.

Perhaps it would have been more interesting with Alexander and Montezuma on the map. But somewhere inbetween, more war willingness but also prevention of very long and costly wars, should be the golden mean.
 
Just played a little with the last version (fixed war declaration percentage) and it definitely feels too cuddly and peaceful. In the whole game (terra, standard size, monarch) there were only two AI-wars (both times Julius Caesar declared on Ragnar) and I could cruise straight to infantry and tanks.

Perhaps it would have been more interesting with Alexander and Montezuma on the map. But somewhere inbetween, more war willingness but also prevention of very long and costly wars, should be the golden mean.

Maybe it's time to switch on to aggressive AI?
 
I've got Aggressive AI turned on, and have had several wars begun between AIs. None against me, but that's probably lack of relations (until recently only Hatty has been in contact and we're the same religion) as much as anything.

Wodan
 
Continuing a 4 human 4 AI game with the 12/2/07 build.
Each human has 1 AI as a ally, always war.

We all noted that our AI partners did not seem to be taking any aggressive action now. They are doing O.K. at defence though. I saw my ally take out a pillaging stack of 6 units that another human sent against him.

However, I marched a stack of units up to a AI (Korean) city that had 16 Hwacha's, 2 trebs and several other units in it just to see what would happen.
I got attacked by 1 Hwacha.
(my attack was halted by the AI's human ally who had seen me coming and next go arrived with lots of catapults and other units, so I ran away)
 
New version is definitely too "cuddly". Getting a declaration of war from the AI almost takes undefended cities, an overflowing treasury, and a neon flashing "attack me I've got free loot" sign.

The dramatically reduced chance of having war declared is going to make the game far less challenging. At first I was worried about defending against attacks from my neighbors; after I figured out that nobody was declaring war on anybody under any circumstances, I forgot about them and just focused on building my cities.

If the best strategy for human players often involves taking out or reducing a neighbor early in the game, then the AI should seriously consider that, even without the "Aggressive AI" settings. Warmongerers in particular shouldn't be trying to play builder; they're no good at it, they just get out-cultured and out-built. In my current game, even Ragnar hasn't even thought about attacking anyone--there's no way he's going to out-build and out-tech the others without stealing some of their cities first. I think the warmongering settings need some serious adjustment at the moment--not quite as high as when they were bugged, but much, much higher than what they are now.
 
It depends on how the game shakes out. In my first 2 games, there wasnt much war, but the AI were still competitive all the way up. They attacked me a few times and each other once or twice.

In my third game, there were quite a few wars. The religions were more split than the previous two and that lead to a handful of wars. Alexander stomped on and vassaled two opponents and we eventually had a 'world war' between me and Vicky against Alex, his two lap-dogs, and Peter. Only one nation wasnt involved in that war and he was isolated on another continent.

Perhaps the chances could increase a tad IMO, but its not really necessary for the standard setting. The game isnt supposed to be centered on warfare. I know a lot of players are happy if they can go a whole game without being in a war. I'm somewhere in the middle. I prefer for diplomacy to be viable, but I like having a potential threat to deal with. Having to plan for war exclusively game after game wasnt as much of Civilization IMO.

I think the BetterAI standard should not deviate too far from the norm as far as wars are concerned (thats what Aggressive AI is for, particularly now that its not going to include the 'kill the human' penalties).
 
The 2/12 build is more peaceful than the 1/24 build.

In my current game (Pangea, Large, Marathon, 18 civs, Aggressive civs) here are the wars declared by an AI civ:
350 BC Wang Kong vs Alex
80 BC Rangnar vs Wang Kon
40 BC Kublai Khan vs Brennus
01 Isabella vs Brennus
100 AD Montezuma vs Isabella
140 AD Napoleon vs Washington
(which is where I stopped playing)
Oh, and everyone (except Monty and Brennus) are Jewish.

I kind of liked the fireworks of 1/24, but 2/12 is probably more realistic.
But I think the AI civs are putting together weaker stacks and throwing them suicidally against entrenched defenders behind walls when such weak stacks would do better by pillaging.
 
Three more games (2 with 2-3 players each though) and the AIs were all pretty non-warlike.

Many also seem to be a little light on the troops again. Is it possible another bug was introduced somewhere? Most of the time they have adequate garrisons in most cities (about 4+ units) but I'm not seeing any form of a 'field army' or reaction stack anywhere. So when they get attacked they are generally pretty passive (nothing to hit back or force the invaders out).

That is not without exception of course. If the AIs are the aggressor, they do tend to have some nicely mixed large armies.

My gut feeling is that somewhere between the January builds and what we have now is probably a good starting place. The January builds were extremely warlike and felt like Aggressive AI was always checked but the latest build is probably a little bit too peaceful overall. There are still wars, but AIs dont seem like they are competing for space or resources anymore.
 
Current game: Pangea / Prince / Normal / 7

I'm Isabella versus Ragnar / Monty / Alexander / Hannibal / Wang Kon / Gandhi

Starting positions had me more or less in the middle with Hannibal to my right, Monty to my left, Alex above me, Ragnar above Hannibal, Wang above Alex and Gandhi above Monty.

Warfare has been more or less continuous once the land was all settled (big surprise with Ragnar, Monty and Alex). Needless to say, this was not the type of game that suited Gandhi, and he didn't last long (Ragnar, Monty and Alex all carving out their bits).

My strategy was to appease Alex and take out Monty (it didn't help that Monty first declared war on me around 3000BC with a lone warrior as his killer stack). We had different religions so I knew he'd be impossible to reason with; after 3 wars (one bite at a time), I was finally able to vassilize him.

When I was in the midst of this, Ragnar dogpiled me (at this time he had both Hannibal and Wang as vassals, although Hannibal later broke away), and I was in a world of hurt because both Ragnar and Hannibal had a huge tech lead on me and comparable sized armies(eg they had Artillery and I was still at Rifling / Cannons). Fortunately for me, I was able to bribe Alex with Military Tradition to declare war on Ragnar. He very obligingly did, so Ragnar and minions turned their attention to Alex and made peace with me a few turns later. This saved my bacon, but it was an incredibly dumb move on Alex's part (any human player would have realized that the risk was not worth the reward).

It's now 1950 and after a long chase I have finally (more or less) caught up to both Ragnar and Hannibal in tech and am now chewing on Hannibal (who for some reason is still fighting Alex although he's losing the war back home).

Random observations (not just from the current game):
1) the AIs still love SAMs and Artillery relative to Tanks, Infantry and Marines.
2) the AIs seem indifferent to Flight (because it can be ignored for Space Race ???) which means they are slow to get Aircraft and Gunships so there is a period where you can have a tank-heavy, SAM-light force and do serious damage without the AI having an effective response.

My previous game had much less warfare, so I suspect that the amount will vary based on the particular leaders that are in any given game.
 
2) the AIs seem indifferent to Flight (because it can be ignored for Space Race ???) which means they are slow to get Aircraft and Gunships so there is a period where you can have a tank-heavy, SAM-light force and do serious damage without the AI having an effective response.

I noticed this too. It's good for me, because I rarely go for Rocketry, unless I'm trying to win Space Race. I like to get Radio for Eiffel Tower, but now add Flight (used to skip it) because 7-8 Bombers can be very effective against an AI without Flight. I build the Bombers in cities that don't produce extra promotions (i.e., no Military Instructors), since they aren't as useful producing less-than-crack troops.

Also noticed: A lack of robust response to invasion. For example, in the game I'm playing now, I've declared on Monty twice. The first time, I expected a wave of Horse Archers or something in response. Nothing. The second time, I expected nothing and got it. A few Knights showed up around the newly captured cities, but I dispatched them easily with Cannons and Grenadiers. Yes, I had a big tech lead on Monty, but in the old days, I still would have expected a big response from him. Now he just sits there and takes it.
 
Also noticed: A lack of robust response to invasion. For example, in the game I'm playing now, I've declared on Monty twice. The first time, I expected a wave of Horse Archers or something in response. Nothing. The second time, I expected nothing and got it. A few Knights showed up around the newly captured cities, but I dispatched them easily with Cannons and Grenadiers. Yes, I had a big tech lead on Monty, but in the old days, I still would have expected a big response from him. Now he just sits there and takes it.

Yeah, this is what I meant by a lack of a 'field army'. They have fairly decent garrisons in each city for peacetime, but there is no reaction to being attacked. If they arent the aggressor, there is little response at all.
 
Just had a very enjoyable game (Prince / Fractal / Quick / BetterAI handicaps). I chose Louis and went for a cultural victory. I missed Hinduism and Buddhism but got Judaism. Luckily, none of the AI opponents on my continent had any religion, so I spread it everywhere. I lightbulbed Theocracy from a spare Great Prophet and spread that all over too. I was at 100% spending the whole game and never went into deficit. Cruised to a culture victory in 1926, with only one opponent even having started the space race.

There were exactly 2 wars the entire game. I bribed Brennus to attack Hannibal. I was *way* ahead of Brennus in tech, and I just took the opportunity when I noticed he'd go to war for 5 techs. Seemed like a dumb decision for him at the time -- I just wanted to keep the 2 AI players distracted with war for a while -- but he ended up not getting wiped out by Hannibal (as I'd expected) and ended up ahead of me in score for a while. I never went to war myself. The other war was between the bottom 2 scorers HC and Roosevelt, and nothing really came of it.

So I'd agree with the other observations that the 2/12 build seems a bit light on AI war. On the other hand, our continent was all matched on religion. Also later in the game (after I'd gone Free Religion) Brennus went annoyed at me and I got the sense he was gearing up for war. There was a huge stack (10-12 cavalry and 7+ cannons) in a city right on my border. I immediately upped my defense and offered a few freebies to him, and he never declared, which was fine with me.

All in all, a very enjoyable game, and the lack of warfare on my part meant I finished in just over 5 hours.

Thanks once again to the Better AI team for a great experience!
 
Same experience with lack of war.

There where rare wars, usually for obvious reasons. Most of the world's wars where instigated by me.

This makes me want an "aggression" meter I can play with.

As a second note, when fighting wars, the AI wasn't being very aggressive with me. I had a smaller military, and they declared war -- yet they only sent one or two probes into my territory.
 
In my game there is no lack of wars even though I'm on a continent where everyone has the same religion. You'd expect peace, but after the best settlement spots were taken, one AI went to war with another allthough the relations were very good. It was just that one was stronger than the other. And then the strongest AI on my island joined in and the weakest AI is now being taken apart.

But I use the aggressive AI setting and that changes a lot. You shouldn't complain about lack of AI aggressiveness when you don't try the aggressive AI setting. If you want aggressive AI's, then just use the aggressive AI setting. Now if you're using the aggressive AI setting and still don't see any wars, then something is wrong and the aggressiveness of the AI needs to be increased. But in my game, I don't see the lack of AI aggressiveness. It is of course not as aggressive as in the previous build when I got 12 AI war declarations on my own nation alone before the AD years. But for me that was a bit over the top, even for aggressive AI.

I was wondering something related to the aggressiveness issue. Is the chance of war declaration at a certain turn in any way related to the game speed setting? Because, if it isn't then you'll see more wars at the lower game speeds where there are more game turns.
 
I've been playing the 02-12 build at Prince without the Aggressive AI option and it's become somewhat of a snooze for me. I went up to Monarch, but couldn't get my game to click. If I kept up in tech, I fell behind in military and got stomped.

So I went to Prince with the Aggressive AI option. This is more challenging, but I think still winnable. I've had see-saw battles with multiple AIs, though I think I'm prevailing. The dog-pile wars against me are the worst, but accepting that I might lose a couple low-value cities while I wipe out the invaders has helped me stay calm. :)

In short, the Aggressive AI option does make quite a difference.
 
Yes, I've tried a few on Aggressive now and there certainly are a lot more wars. I'm not sure a lot of them made sense though (attacking me when I had a clear power advantage and simply defeated their army and then took/razed a few of their cities).

I've also seen quite a few AI wars that again didnt amount to too terribly much. Both side wrecked each other's econs with pillaging and the occasional city changing hands and were quickly out of contention.

To me, there is certainly area for a middle ground. On one hand, the non-Aggressive AI feels really passive most of the time. There are occasional wars, but by and large it doesnt seem like the AIs are actually making any attempts to better their positions.

On the other hand, the Aggressive AI often feels like its being 'aggressive' just to do. It often seemed like the wars were just to fulfill the 'aggressive tag' rather than as part of an actual attempt to conquer a neighbor or aquire a needed resource or eliminate a rival.

Personally, I thought the level of conflict felt better before the huge aggression kick that started in the early January builds. The arms races weren't preferable, but the number of wars felt correct. The 1/24ish build that introduced the DoW bug went way too far IMO, but the 2/12 build feels too passive. Perhaps in attempting to curb the unit spamming and to mitigate some of the bug-induced banzais the AIs have been throttled back a little too far.
 
I tried the 2/12 version in occ with custom handicaps, I was gandhi and hatesput,Augustus, Tokugawa, Bismarck, Wang kong, Ragnar as oppenents in pangea agressive Ai on. Ragnar collected a huge SOD of axes,spears,swords and some archers and had them sit in his capital for a long time and finally built one other city. He was very low technolgy most propably due to the huge amount of unit upkeep and lack of cities. Augustus on the other hand had a big Sod of preatorians , axes in one of his border cities with me. He was also trailing a bit in technology and however he built some cities which helped him research . I was Hindu and spread it to Hatty, Bismarck and Augustus and Tokugawa. Wang was Jewish and Ragnar buddhist . Inspite of all this there were no war declarations, till I started bribing Ai's into war. However Ai , Ai war was eventful with big Sod's marching around and no peace for a long time.
 
Yes, I get the impression that there are enough wars with 2/12 and aggressive AI. But the reason why many of them were pointless might have been that some civs just didn't build the units to back them up.

I think that when we had the always war bug, the unit spamming had been over-corrected. Yes the 1/25 was over the top. I'm my OCC game using that build, Augustus sent his SOD of 67 units against me. I lost and reloaded which still turned out close. Praetorians are overpowered even against rifleman.

Now that the AI can keep up economically, it should be able to build more units without hurting itself too much. I'm experimenting with using a unit cost limit of 10% of science plus income minus expenses, rather than 20% of expenses.
 
Top Bottom