Optimum AI Flavour settings

Hey guys! Been a lurker for a while, never posted a thread. I was wondering what you think the perfect AI flavour settings would be for a challenging civ (if you want to go in depth you could go for each victory type)?

I was thinking about it because in the hands of a real player, Hiawatha is just... average. His AI takes advantage of the AI's bonuses perfectly, and I was wondering if you could design a civ that would take control of the AI bonuses better? Could be interesting for Deity challenges to have a game with 4 civs, each one perfect for science, cultural, diplomatic or domination victory.

For AI seeking mostly peaceful victory: Korean AI flavor, especially in both Vanilla and G&K.

For AI seeking science victory with some successful war mongling: Persian AI flavor.

I find the AI in general does best with a rather wider strategies. It needs to be expansionist without the pitfalls of ICS, taking advantage of its bonuses to growth, production, and Happiness to get bigger.

It also needs some science flavor to maintain or gain a lead.

It needs enough military flavor to be willing to exploit a weak neighbor, but not so much that it over-builds military or makes itself an unpopular warmonger. It almost needs to be opportunistic and deceptive enough for a decisive sneak attack.

It doesn't need a ton of focus on Culture as it seems the AI is horrible at winning Culturally. It cab do okay with a Religion but shouldn't be over-aggressive with it as it'll anger all the neighborhood.

Hiawatha's AI comes close.

One of the biggest things an AI can't get away with long term is overt aggression or a tendency to be bribable. Especially the tendency to be bribable. Bribable civs become backstabbers if the player is doing anything right.

I find the AI in general does best with a rather wider strategies.

Click to expand...

On wide AIs; I find most of the expansionist AIs actually shoot themselves in the foot via not building enough science buildings; resulting in falling way behind.

Catherine also reaches the hall of shame by going into (and staying in) unhappiness even with the AI happiness handicap.

Zulu, Iroqois and Korea all seem like AIs that tend to do well with their settings in different areas. Zulu = best wide warmonger, Iroquois = best wide builder, Korea = best tall science. Persia also *can* be a real beast, normally wide with a mix of science and conquerer. I think the AI generally does fail at going tall, only reason why Korea performs so good is because of his ridiculously overpowered UA, so if there's a way to teach the AI to favor science more, that would probably help it when attempting to go tall.

joncnunn said:

On wide AIs; I find most of the expansionist AIs actually shoot themselves in the foot via not building enough science buildings; resulting in falling way behind.

Catherine also reaches the hall of shame by going into (and staying in) unhappiness even with the AI happiness handicap.

Click to expand...

True, but this is more that the AI cannot develop cities to save its own life, combined with relentless expansionism. Hiawatha and Cathy expand too much, and put too much effort into building even more settlers and workers in lieu of increasing population and infrastructure.

The empires that stay tall tend to fail worse in my experience. The AI cannot handle Venice in the slightest, China often falls flat on its face in my games (founding 3 cities then staying static), Morocco never accomplishes anything, etc.

GoStu said:

Morocco never accomplishes anything, etc.

Click to expand...

In my most recent completed game, Morocco was the last civ to beat in my cultural victory. They were also the most advanced AI in science.