from the years 1500-1800.

A turn is one year,

turn-based and singleplayer,

The units represent large bodies of men

you have no unit stacking?

Yes I do

How big is a hex?

what's the stacking limit?

how big is a unit?

obviously small enough to consist of all cannon or horses?

you're early napoleonic era tactics, pretty much.

infantry in divisions, but cav and arty only in regiment sizes as attached units.

armies at that time (circa 1650) will also be more ad-hoc - not likely to find an org chart for an english inf division from 1650 if you know what i mean. <g>.

by going with a traditional wargame type system, you get atk, def , rng, and move.

rng for all units will be 1 hex or country most likely, given the scale you describe. IE adjacent.

atk and def can be the same. so your cavalry engagement in the field is a roll of the die.

move is just movement rate. obviously horses are fast and cannons are heavy.

then you get terrain bonuses for atk / def (mostly def). bonus for defend in town, bonus for defend in fort. bonus for defend in woods, etc.

cav is just weak inf that moves fast. its used to mop up once the inf breaks and runs. kill them the first time, you don;t have to fight them a second time.

the other use is to atk inf in the flank or rear of inf already engaged by other inf, etc.

or to atk undefended arty units.

arty units are big atk, but even weaker than cav atk on defense.

by choosing appropriate values for atk, def, move, and terrain bonuses, you'll have a simple yet quite effective system.

then you simply resolve combat like in a wargame. add up atk of all attacking units, add up defense of all defending units, calc odds, roll die, consult table, get results. you can injure or destroy individual units, or disburse them (ineffective for a while, very vulnerable if attacked again) or you can force them to retreat, destroy the whole stack, etc. given the time scale of 1 turn = 1 year, damage, destroy, or retreat might be the only options that make sense.

there's no room to correctly model real armies with real composition of different troop types

not necessarily - the stack is the army! the units are just the companies and regiments that make up the army. So, my "army" has 4 inf, 1 arty, and one cav unit.

artillery units cost wood, metal AND horses

Arty requiring horses is fine. Inf might need pack horses as well. Then all 3 units would require horses as a resource. Maybe do that, maybe not.

You might want to convert the 'native' type into something more like

mob/rabble -

and light infantry (troops trained for bad terrain)

and 'irregulars' - often fairly well equipt but with poorer 'morale' who fall apart faster when faced with casualties (and other control problems)

'militia' varies too much to have as a separate classification

'natives' can give a pretty good accounting when in the right terrain and situation

If you want sufficient complication you might also want engineers/sappers

One of the first things you'll have to do is decide where you want the game to fall on the realism scale. If you want high realism, you'll be limited to the types, numbers, and organizations of troops based on historical records.

If you don't care about realism, you can make up anything that "sounds good".

And then there are points in between those ends of the scale.

As you move from historical realism to fictional made up stuff, you lose the inherent believe-abilty of historical, but gain in the form of fewer design constraints, and perhaps an easier time balancing things, as you may not have to realistically justify any required balance adjustments.

One of the very first decisions I make in any new game I contemplate is where it would fall on the realism scale.

Once you decide on a level of realism, it makes it easier to figure out what you'll need to model in the game, what you can make up and what you can't, and what you can and can't do due to realism or lack thereof.

Colonial militia - 1 strength, +200% vs natives (the reason is you want to use them vs natives but not vs other colonies)

Infantry - 3 strength

Rangers - 3 strength, treat all terrain like plains for movement purposes (more expensive infantry that can move on rough terrain fast)

Cavalry - 2 strength, +1 mobility, +1 sight range, chance to retreat to a neutral grid after defeat instead of being destroyed

Artillery - 1 strength, +500% when attacking, ignore fortification bonus, can not enter mountain terrain

General - 1 strength, +1 mobility, adds +25% strength bonus to all units within 5 grids radius, bonuses from multiple generals do not stack

Supply wagon - 1 strength, can not attack, provide supply like city

Rule: all units, except natives, further than 5 grids from the nearest city, fort or supply wagon get -50% to strength

Rule: all units, except natives, further than 5 grids from the nearest city, fort or supply wagon get -50% to strength

supply is usually calculated as having a clear path of hexes to a supply hex of some sort (city, port, depot, wagon train, tanker unit, tender, etc). its called "line of supply", not "zone of supply" for a reason.

i suppose on a large enough scale, zone of supply could work. IE supply trains can only go so far in reality. the classic problem of out-running your supply. that's why in modern armies 3/4 of the troops are support, and only 1/4 are combat. and if you include the home front and production, its more like 10 people in support roles for every grunt on the ground.

if you use areas instead of line of sight, they should be large enough to reflect the maximum line of supply capabilities of a supply hex. but this could result in large areas. an army could attack washington while being supplied by new york. and troops in an area but surrounded would still be in supply , which is totally unrealistic. that's probably why areas are not used and line of supply is. its all about being cut off from your supplies. without that, there's no need to outflank, surround, or cutoff. and what about sieges? is a fort near a city always in supply forever, even when under siege? these are the ramifications of that design choice. you always have to carefully think through the impact of any game rules you consider, and how they interact with everything else in the game. hopefully before you adopt a new rule, or implement it, or release it, or realize post-release there is a flaw.

you also might want to read up a bit on military history and try some wargames to become more familiar with the subjects.