One thing I'm disappointed in....

[khearn]

Perhaps it should be along the lines of trustworthiness. If Monty is furious with me and I attack him, no one should be too upset with that. But if I'm best buddies with Ghandi and I attack him, then everyone else should take note. If I demand a tech from someone and he gives it to me, but I attack a couple of turns later, others should remember it.

What should the effects be? Well, a diplo penalty is obvious "-1 You can't be trusted", or maybe more specific "-1 You attacked your friend so you can't be trusted". But it should also cause other AIs to take steps like keeping their border with you better defended than they normally would when they have good relations with you. And if you have a reputation for attacking after being paid tribute, they should be less likely to give you tribute, "Why should I give you anything? Ceasar did, and look what you did to him!"

[Hauptman]

Speaking of fortifying units on a border...

Why does the AI not do this? In civ 1 they would, maybe even civ2 as well. But not civ3 or civ4. Nope no border defense at all in this game, not even a single fort.

Never an ai built maginoh(sp?) line. Especially if they have a 1 or 2 tile bottle neck... fortify a few units there and keep me out darn you!!!

[Silver14]

Lately I've been a total ass to the AI. I completed probably my last game on Prince in the process. Played a Marathon game as the Inca, aggressive civs can be very powerful in Marthaon mode. Given the extra turns I explore a bit and try to found my capital close to another Civ. Produce nothing but Quechas focusing on production even if it means no growth and basically pounce on the ai closest to me, take their city start mass producing quechas there and go out and tromp as many AIs as early as I can. I usually raze captured cities from here on out unless they are close to my capital as the maintenance costs can cripple and stagnate my early empire.

A large stack of Quecha with the 10% bonus of agg can easily splat most enemies early on even if they have archers. The AI simply does not produce enough units early on and the 20-40% defense bonus of a city isn't enough either. I had the Persians, Romans, and Mongols destroyed before 2000 ad. The Mali took some more effort as they had many cities and skirmishers by that point. I was using axmen and focusing on taking them out.

That left the Greeks and Arabs who were on another continent. They were the 'best of buddies' at first sharing Buddhism as their religion. Once I found them and was close to Astronomy I started a military build up again. The Greeks were stronger and had two religions founded so I started with them. Once I razed Athens, the Buddhist holy city the Greeks switched to Christianity. It wasn't long after that the Arabs who were still Buddhists declared war on the Greeks on their own accord and started finishing off some of the Greek Cities.

This was a very bad Idea for the Arab AI as we were now the last Civs standing. I started another military build up and spoke with them every few turns to ensure they had not discovered Astronomy yet. I fished the game in the 1600s with Canon, Cavalry, Riflemen, and Grenadiers with a conquest victory.

The AI should really make notice of the number of Civs you have destroyed before making arrogant demands and backstabbing other AI 'allies'.

[civ4lyfe]

I am currently involved in a game that has some bearing.

Inland sea
Prince
Domination/population victory conditions

I am Catherine in the NW
Ghandi is in the NE
Cyrus in the SW
Saladin between Cyrus and Washington
Washington in the SE
Ghengis is dead (he was between myself and Ghandi in the early game)

I wanted to see how far I could take the game, we are all well in the modern ages, all research is complete.

Scores are approximate
Ghandi 5200
Washington 5000
Catherine 4100
Saladin 3500
Cyrus 3000

I went to crush Cyrus to get the foothold for my round the world romp (as Ghandi and I are on the best terms I wanted more production cities before my run at him). Low and behold the very next turn everyone but washington declared war on me, even my best buddy ghandi. I was amazed the ai had formed such a tight bond, and was readily putting a halt on my war machine.

I defended for my ten turns, which was tough with all those guys against me and washington not willing to help. Negotiated (bought) peace with one at a time while defending myself.

Now I am at peace with everyone, and building the most massive army I can. I dropped my research to 0 and my culture to 50 getting about 700 gpt.

This time I am going to try hitting washington to gain those production cities, but he has a ton of artillery, and gunships, I just keep sabotaging his oil so he cant build ma, etc.

In summary I just wanted to say imo in this particular round the ai made exceptional choices in holding me off, and has made me re-think my strategy with regards to roling over the ai.

gg

What is my best counter for a ton of artillery? lol

[Strategist83]

Velociryx: This is my main gripe with the single-player game, too. I wish the AI could be a little more 'human' - in other words, a little more interesting in winning than in 'role-playing'. I posted about this a while back (although it was misunderstood by the majority of the people who responded): http://www.apolyton.net/forums/showt...hreadid=147065

See my third post.

[Strategist83]

Originally posted by civ4lyfe
What is my best counter for a ton of artillery? lol

Marines/Navy SEAL's gain a 50% bonus vs artillery.

[Blake]

The best counter for artillery is bombers (lots of them).

Another good thing to do is send in single strong units, guerillas or woodsman, the AI will then need to suicide multiple units on each one.

On to the topic. I also find the lack of a "You're a raging psychopath who butchered our friend(s)" penalty disturbing. And the fact you can amass a large pile of units right on an AI's border and it's blissfully unaware of them until you declare war.

It's wrong both from a role-playing and competitive perspective. The role-players shouldn't mind you being a violent thug if they're buddies with you, think of it as hanging on the coattails of the great... but the civs who don't like you should get really, really worried. AI's are a bit too inclined to dogpile those in a weak position, and not inclined enough to take on the big dog. They think a bit too much in the short term in this regard.

The dogpile code would have to be revised, because as it is if one "rebel" AI declares war on the #1 civ, that rebel is then perceived as being in a weak position and promptly gets dogpiled.

Although I don't like tall poppy syndrome in principle, I think it should be more about nobility rather than raw strength. The despicable should be readily dogpiled while other civs are willing to come to the aid of the noble.

[Swiss Pauli]

In my very limited experience, my gripe with the AI is that they seem to disregard UUs.

In my current game Spain has DoWd me twice in the early game as revenge for me beating up America.

But I have Praets (as she knew from OB snooping) and she had (horse) archers. Cue slaughter. TWICE.

If she 'studied on killing me' as she claimed, she'd have been better off waiting for Conqs to come on-line.

Related to this, I don't see much evidence of the AI tech beelining for their UU pre-reqs. But maybe I just got lucky so far!?!