I've added AIvAI war maps and lowered settlers cost, but the auto-play simulations are still unsatisfying.
No AI expands decently, they just found 3-4 cities around the area and that's it.

Any ideas to give them a shock?

Cheaper units to some of the civs? (not too cheap, 5-10% cheaper).

Cheap defenders so the AI sends out settlers.

Maybe ask Better AI? They seem to have gone through practically all the AI logic.

Have you edited the tech tree at all? Maybe they're waiting for more technology that makes growth work better. Possibly also lack of money.

I'm afraid I think the problem is the map, too big.
The distance from Macedonian capital to Babylon (not even Persepolis...) equals the distance from England to Persia in RFC, or from Mongolia to Israel, to get the idea.
Even raising most units movement to 2, I think we're just asking too much to the AI, no matter the improvements in AI logic.

I wonder how RFC Europe works out, with such a bug map. Does Germany ever conquer Spain?

I'm afraid I think the problem is the map, too big.
The distance from Macedonian capital to Babylon (not even Persepolis...) equals the distance from England to Persia in RFC, or from Mongolia to Israel, to get the idea.
Even raising most units movement to 2, I think we're just asking too much to the AI, no matter the improvements in AI logic.

I wonder how RFC Europe works out, with such a bug map. Does Germany ever conquer Spain?

This has been a problem for us. It's not really possible to get the "proper" fast expansion that would be historical for many of the civs. Independent cities help (pre-placed barb cities are even better), and we've added some pre-built roads which help as well. If you do tweak the AI logic in an effective way I hope you'll let us know.

Have you tried to require cities to have at least 2 tiles between them?

It probably really is the map. One simple way to reduce the map size without removing tiles would be to turn it 45 degrees. That way you lose the ice on the edge. Of course that means redoing the whole map. Not sure that be an optimal solution (or one which I would like to do)

Have you tried to require cities to have at least 2 tiles between them?


That would hamper Greek civs.


It probably really is the map. One simple way to reduce the map size without removing tiles would be to turn it 45 degrees. That way you lose the ice on the edge. Of course that means redoing the whole map. Not sure that be an optimal solution (or one which I would like to do)

If you rotate a circle, you will always get edges at the corners.

I've resized the map files to 63x60: the detail il still quite high, but there's much less empty space between starting locations.
I'll do a few tests before redoing resources, rivers and city/settler/war maps

Have you tried to require cities to have at least 2 tiles between them?

I'm always a fan of this, though in certain areas like Greece, Italy, and the Holy Land it may not be wise. Someone said that the code was editable in different parts of the map.

If you rotate a circle, you will always get edges at the corners.

I've resized the map files to 63x60: the detail il still quite high, but there's much less empty space between starting locations.
I'll do a few tests before redoing resources, rivers and city/settler/war maps

Right, what was I thinking? On second thought, because the tiles are arranged in a chess-pattern, there would be less tiles? Difficult to say.


What if you discard much of the landscape in Europe and the Sahara? This doesn't really have to be big. Especially France and Spain only has to have room for one civ (and romans and carthaginians), so you could make them extremely distorted, the Whole "Eastern Europe" as well.

In Hecataea, which had no RFC expansion mechanics, the AIs did expand enough. Often they did not expand in the direction I wanted, but I did influence that a little with Worldbuilder attitude modifiers. I also used the Aggressive AI setting, but I think you must be using it too.

I'm afraid I think the problem is the map, too big... I think we're just asking too much to the AI, no matter the improvements in AI logic.

I can appreciate the challenge of reprogramming the naval AI to get civs to fight across the Mediterranean. But for land connections, I don't understand your concern. Why do you think the map size is a problem for the AI, such that it would be too hard to reprogram?

Is the AI capable to expand from Mongolia to Egypt in RFC?
Probably, if it had 1200 turns. Maybe. Under certain lucky conditions.
Not to mention the fact that Huns and Germanic tribes were getting lost in the forests rather than attacking civs.

I agree that it takes a lot of turns to expand across the map. For the original size,
I think it should be a goal to allow [a Domination-style victory]. Assuming this mod does not radically redesign the Civ 4 combat and movement mechanics, I think that requires at least 500 turns.
More like 750+, given that civs start out without a lot of existing cities.

But the problem you stated in the first post is that the AIs are stopping at 3-4 cities. That's a different problem. Map size doesn't explain it, does it?

my guess would be lack of units...

do you still have mercenaries for RFCGW?

I would suggest either a greater range of units that can be built, or lower the cost of the unit builds (and the maintenance costs for units as well,) at least 15%

I think I will try testing some games where AI's and humans can actually control HUGE armies and see how conquest goes...

I meant that they don't collide - they may found more cities if they have richer lands, but that's it.

Now, with a smaller map and a longer timeline, keeping the increased unit movement, I can see better results. Such as Macedonian Anatolia, Greek Egypt, Germanic Gaul.
Furthermore, I've forbidden stopping a war in the first 10 turns of play. That will help Rome.

Do you need city names again with the smaller map?

Yeah, as soon as I finish editing map detail. No sense in making a city map while I'm still changing the territories

it takes a lot of time to expand because of the limits imposed by the game. You may soften them with:
- basic buildings for growth/stability pre-built in new cities, and more than 1 starting citizen (2 or 3). Initially, at the dawn of civilization, forming a big city was a slower process than later on, when men already had the know-how on how to speed up the process. This is not represented well in Civ, or not at all.
- lower maintenance cost for cities and civics.
- more ?

- lower maintenance cost for cities and civics.


Yes I am interested to see what will be done here; will there be buildings to replace to modern stability-reducing buildings in classic RFC like security bureaus, jails and what not? The classical era civs will be quite large if its going to be accurate especially since we're dealing with a much smaller portion of the world than classic RFC.

Maybe make certain religious buildings or national wonders help with stability with the argument that it both unifies the people and increases the sense of pride or national/cultural identity to their country?

We could leave jails in.
Or replace with something slavery-related?

Courthouses have been moved up the tech tree right now, but we could try to figure out another late-game stability-adding building.

I suggest making Rome, Macedonia, Persia and possibly some other civilizations less likely to accept other AIs as their vassals. After running an autoplay simulation, I noticed Rome had Celtia, Sparta and Macedonia (and Carthage at one point too I think) all as vassals which really limited Rome's expansion capabilities.

I think vassal mechanics should change in this mod. Rome had several vassals, and annexed most of them later on. I think it would be a nice feature to have the option to annex a vassal under certain conditions, also for the general game.

That's even better, although the problem is deciding on these conditions so it's not too easy/not too hard to annex your vassal.

We could follow a EU3-like set of rules which is:
1)Vassal must like you enough
2)Vassal must have been your vassal long enough
3)You didn't annex other vassals for some time
4)You are stable enough
5)Must be of same religion

I agree with all those except 1). It is much easier for a human player to get a vassal state to good relations than an AI. This means these rules would benefit humans too much imo.

Well first lets see this feature implemented and then balance it out :).