@Spatzimaus: The full idea of what I'm trying to do is make marathon-length games a bit more interesting. I like the idea of marathon-length games, but I feel like the implementation just causes you to sit around with little to do, pressing the next turn button a lot.
The problem is that the method you're looking at just won't do this very well. Making a flat reduction to the production cost of each unit just means that the AI will make far more of them in relation to building infrastructure, because of how the Flavor system works. But since you're not also changing the unit maintenance equation, the AIs will quickly go bankrupt from all of the units they're trying to support, and even altering how warlike the AIs are really won't fix that.
There are a lot of things like this; the game's balance is very interwoven. You can't just drastically change one part and expect the rest to still balance out. If you drastically increase the number of units on the map, it'll affect things like:
> Barbarians. You're making it far easier for each empire to flood its land with units, so what hope do the Barbarians have of being a threat? And with units covering the map, there'll be nowhere left for Barbarian camps to spawn.
> Strategic resources. If you make units much cheaper, then the units that require resources will now be only a tiny part of your army; the advantage will really go to those civs whose UUs require no resources.
> Skirmish warfare. Units that specialize in hit-and-run tactics, like the Horseman and its descendants, get a lot less useful once your opponent just has a wall of foot units walking towards you, especially if those units are resourceless, like... spearmen and pikemen?
> Counter units. If every late-game unit is within spitting distance of an AA gun or SAM, then it becomes pretty much impossible to do anything with air units. The same goes for most counter combos like spears-vs-horses; with larger militaries covering the same amount of land, it's considerably more likely that any horseman left near the frontline will get killed by a spearman on the opponent's turn. The worst part of this is that the AI doesn't think this way, so it's a significant advantage to the player.
> Ranged attacks. Before I might have had 1-2 archers that could reach your nearby unit, but now it might be 4 or 5. Given that every fight deals a minimum of 1 damage to defender, it becomes very easy to set up a kill zone that just obliterates every good unit the AI sends at you. It's bad enough in the vanilla game where a smart human can put 7 or 8 bombers in a city and pound down nearby enemies one at a time; if that applies now to range-2 units, it's hopeless for the AI.
This doesn't even apply only to ranged units; the human's considerably better at making multiple melee attacks against a single foe to kill it, instead of spreading the damage across multiple foes. You'd just make that worse.
> Policies. What's the point of picking the Honor tree if you're just cranking out tons of untrained units to get the same effect (and with the previously mentioned Barbarian issues)? And why bother with a Freedom policy that gives you a measly 8 free units?
> Pillaging. Normally if you want to pillage an opponent's improvements, it costs you in that you're leaving a unit inside enemy territory, where your opponent has significant advantages, to deal damage instead of wounding units or fortifying. If you can send in a human wave, then it becomes trivial to burn everything.
> City capture. Cities only have 20 hit points and deal a set amount of damage based on their strength, and if you can throw three times as many units at them then they'll be captured much more quickly (again, minimum 1 damage per fight); attrition becomes a non-issue, since you can simply move wounded units out of the way to make room for fresh troops. In fact, if the opponent finishes off a wounded unit, it makes your job easier as you now have an extra hex to move into!
The list goes on; I really COULD do this all night. The balance in Civ5 is very, very fragile. A large, unilateral change to one aspect of the game will just destroy what semblance of balance remains.
Also, if you're going to do an across-the-board change, like reducing costs by a fixed fraction, it's FAR better to use SQL. Think about it; you might have changed the Warrior and Jaguar to cost 14 instead of 40, but what happens if there's a DLC adding a civ with its own UU alternative to the Warrior? It'll stay at 40. So either do the reduction by some more generic variable (unit class, tech prerequisite, era) or use SQL.
Just so you know, I've never bothered starting at any tech period other than ancient, so I'm not super-concerned about the effect on starting in other eras.
If you're ever considering publishing this mod so that someone other than yourself can use it, then you do need to be concerned about this. If it's purely for your own use, then go right ahead.
eek2