From: Secret Underground Lair
Here's one that is probably pie in the sky: topography.
Instead of merely two elevations (well maybe 3 if you consider swamps as 'lowlands', or even 'four' if you consider the impassable mountains as the fourth) how about three or four?
In the old Civ War game "Civil War Generals II" hills could be tiered. One big hill might have sprawled out over 20 or 30 hexes! The tippy-top of the hill might have been a string of only 4 or 5 hexes toward the center, with the next string being one level lower, though with some levels perhaps being wider or more narrow.
If you guys are curious what I mean here, have a look back at that old game, which is now freeware and see what I'm referring to. I think this would go a long way to making the detailed battle maps more realistic if the random algorithm could be harnessed to generate those kinda maps like they had for CWGII.
Second one: once a random map gets generated for a province, have it be a possibility for subsequent battles in the same province.
In my game, I must have had 8 battles in Champagne and every single map has been different. Now admittedly, Champagne is a big place and therefore no two battles MIGHT have happened in the same place. But maybe in some provinces they WOULD be more likely to happen in the same place?
If there were any way to have the randomly generated maps become saved as temp files that would then have some semi-random chance to be the map for any subsequent detailed battles in that hex it would be cool. In fact, if there were a way for the player (or the AI) to have the opportunity to favor having a battle in the pre-fought map that might in itself be cool.
Third, this one is not so much a wish for you guys WCS, as I realize the sheer work that would be involved but, maybe for the fan community: Digitize the actual land areas of all provinces in the game (or at least the areas that are within 4 miles of 18th century roads and where battles could/might have happened) into hex maps. Maybe this results in 15 hex maps for Champagne, 10 for Ile de Paris, 5 for Le Havre, etc. Save these maps as files somehow in the game, and assign particularly strategic parameters to each (e.g., chokepoint, or port facility, or major town, or provincial capitol, or crossroads, or high-ground, or whatever) and link particular benefits to winning combats in any given strategic locale to the finite number of strategic locales you determine. Now the detailed battles seem less like random sandbox explorations of hypothetical tactical challenges and more like real, tactical battles fought out in a larger strategic context. If you are the attacker and you have initiative, maybe instead of attacking the port facility you attack the cross-roads. This allows you to block movement into and out of the province or to block supply, or to pillage supply from the enemy, or what have you. Or, maybe you prefer to go ahead and attack the port "map." If you win the battle early enough on Day 1, then maybe you get the chance to torch some ships, with an increasing chance that those ships can get away as Day 1 ebbs into Day 2.
Attack the urban center of a province that has high iron output, and cause its output to go down temporarily.