I also tried much higher values, and here's my experience so far:

If you make materials anymore than 20, the AI won't build pioneers at all, this is apparently due to how the AI decides what units to build, it is based on material and metal cost, and the most "expensive" units will only cost 20 material/metal. This is actually an okay setup if you like to play on tiny maps, as it will end up being a battle of capital cities most of the time.

At 20 materials, some AIs who start near materials has a big advantage. They will still expand like mad while others will be stuck with 1 city. This is also true to some extent with 10 materials, although it is admittedly less of a problem.

If you add metal to the cost, some of the AI will be screwed (that doesn't have an iron mine at the start) and end up not building pioneers at all. They would need to get lucky by picking up metal in their exploration, or start with resourcefulness.

Adding higher pop values generally make factions with royal sovereign too strong, since other factions will gimp themselves by making pioneers.

Gold and training time seems to be the only "fair" values thus far. I've tried 200 gold with +10 turns, and 300 gold with +15 turns, both were pretty successful at stopping massive city spam. However, the problem remains in how the AI selects where to build their cities. That is to say, instead of reaching out further to grab more land, they will still build their 2nd city near the closest unclaimed resource despite the increased difficulty of getting a pioneer.

what happen to AI when we change the value of Pioneer? Will it okay?

@ Brainsucker: the AI adapts fairly well, they still play okay with modded values, but there are various problems depending on how you mod the values like I mentioned in my earlier post. Most of the issues arises from bad starting location rather than anything else (IE: if they aren't near a tech library and can't research the goldmine techs). But in most cases, those bad start location AI are pretty much doomed anyways. One thing of note is that if you make the pioneer take a long time to train, the AI seems to be less hostile because the pioneer building eats up a big part of their military unit production time.

@ Zard0z: The AI unit decisions that I was talking about is in CoreAIDesf.xml nearing the middle of the document where it talks about their unit design. Cheap Units = 10 mats, no iron; Basic Units = 20 mats, no iron, Decent Units = 20 mats, 5 iron; etc... Basically, currently the best unit the AI designs will use no more than 20 mats, 20 iron, and 5 crystal (doesn't seem to be a gold or training time limit). At the time I was experimenting with high mat cost pioneers, I thought this was just the value for unit design and it would be okay since pioneers are default design, however, the AI in that run didn't build any pioneers, which leads me to believe that this also affects what they build as well. However, I only did 1 test run on those values (and it seemed pretty broken), so it could be an aberration (it could have been that all the AI in that run had mats problem, but rather unlikely since I always test with 10 AI).

   Being restricted on your ability to expand should be effected by one's environment. I like this potential. I have never been a fan of a "fair & balanced" start. Gives me an excuse for why  lost!

materials are what comes from lumber.

I also tried this and fiddled with the settings in elementaldefs.xml also, for unit wage scaling.

in a current game I have  a faction carrying a balance of -6000+ gildar, though it seems to slow city spam a bit.

The problem with adding food cost to city hubs is that your first city will kill you. In other words, you'll starve as soon as you found your capital. Even if you were to fix this somehow with a capital bonus, you'll only create a situation where it's tough at the start. After the first few cities, you'll still have the city spam situation due to trade caravan producing food. If you make a city, then make a caravan that gives a 30% food bonus to your capital (that might have 12+ base food production), then you essentially gain food. Enough to get that new city to level 2 or 3 even, not just a little lv1 outposts.

I've been experimenting with giving higher level city hubs Gildar base production (think of it as tax) and more Prestige to make it more alluring and easier to get to, thus making food scarcer overall (since people grow fast with higher prestige, you'll be forced to keep building houses to keep up). I also added a 6th level to the cities for good measure. Unfortunately most of the AI in my recent test are failing miserably to keep up with me, heh. Aside that, what I found was that to get enough food to supply my major cities (which ate a lot), I had to spam little cities where ever I can to make caravans. Thus, as long as the caravan gives a bonus to food, I don't think you can use food as a limiting factor for city spamming.

Maybe we can balance / reduce the city spam just by add the pop need to 11. By then, only a city with a house will able to create a pioneer. Just thinking, will it hurt the AI Economy?

As mentioned earlier, increasing pop requirement on pioneers hurts the AI pretty badly. Their settlement won't be able to grow fast enough to match, as a result, certain factions (kingdom with royalty sovereign) will just steam roll the others.

Edit: As for your suggestion PiersAS, on a typical kingdom capital, if you research the first 2 adventuring tech immediately before expanding, you'll have 12 food. Unless you plan to make caravans completely useless or just removing them entirely, it's not going to stop expansion.

Why do we need to TRAIN pioneers?

Pioneering is one of the greatest challenge to the AI. And in this game the AI has to choose between a lot of oportunity which leads to very different effects. So free the AI from pioneering.

(My last game i gave myself a great handicap. For 60 turns i didn't do anything. I built my first outpost and my sovereign slept there for 60 turns. No building, no adventuring. At the 60th turn every nation hated me because of my week military. But nobody attacked me. They had - on ridicoluos level - 2, 1, 5, 3, 4, 2, 1 cities after 60 turn. At the 100th turn i was equel. I had four cities (if i want i have more) and only one nation had more powerfull army. To tell the truth this made the game more challenging. So recently without war i haven't got any place to expand, but this can't be a solution ( or can, try it out )

In history pioneers wasn't trained as armies. They made their own choice. They were promoted, supported (don't know the proper world), but not trained.

So what about a simple counter.

For example: when you establish your first outpost the counter begins (or when your outpost has 50 popu). If you have your second city with 50 popu the counter get a bit bonus. And every city give it a bit bonus if they are large enough. There could be some tech either which accelerate the counter. The refugee camp could accelerate either. And when the counter reach a number in your largest city a pioneer emerges.

With a similar method the AI could keep up with the player or could be better if he develop the proper tech first. (as regard the number of cities).

And to prevent the redundant late game pioneers. You can have maximum 8-12 pioneers with this method. But if your entire population is above 1000 you can train them on the recent way.

What do you think?

Brainsucker: That would work, it would certainly make fertile land even MORE valuable, which isn't too bad of a thing since you get one to start with. However, there is still the "food caravan " problem, and there is also one more. What if your pioneer is killed enroute to building a city? You lose that 1 food "forever"? That's sort of silly since the food spoils.

I think the alternative way to look at it is it COSTS 1 food to establish a new city with a Pioneer (but not your Sov). Basically, level 1 cities cost one food. It would still encourage city-spams (especially to secure as many fertile lands to spam more cities). But it does force you to evaluate whether to build hovels in your capital or not. Build a Hut, but don't find a new close-by fertile land? You might be in trouble.

So how about changing it so that we need to be size 2 before builidng caravans, (as well as money chanres, archivists and lore shops). And restricing the second caravan to size 3 cities.

(I think I can see how to mod the improvments but not a caravan city size change.)

(Again not trying to stop city spamming in the end game, just at the start to change the pace of the game)

Ah crap this was a great idea but now the AI (stupid, stupid AI) has problems developing it's settlements. That damn AI just sends them unguarded into the woods to die, in a way it's like the way they spammed them was to compensate for how stupid they were about escorting them.

Why not create a new resource and a new resource building to go with it? You can make the building build-able once per side then you can control exactly how fast you can build new pioneers as you control the rate at which the new resource comes in at, as well as the cost per unit. Assuming that all sides build the building at the same time then everyone will always be able to build the same number of pioneers, as even if you start building the pioneers later you will have accrued the resource so can start the next one sooner etc.

You could alternatively add the new resource to the city hub, thought his would create an exponential increase in the speed at which you can pump out pioneers. Limiting it to just the capital in this case (if possible) would be a bad idea as if you lost your capitol then you can no longer build pioneers (unless you have a resource stockpile). Although it would add more importance to your capital I doubt the AI would take it into account.