I thought that we gonna have abilities like first strike in the final version. Froggie has mentioned special abilities in one of his posts here in the dev. subforum.

Archers will have secondary weapons. IIRC this has been confirmed. I hope that dual wielding will be implemented as well later on. [We've talked about it a while ago, and it's not in yet because of some animations related stuff.]

I like version C as well. Another solution would be having everything require a certain resource and you stockpile them just like any other. So in the case of the honey pack, the apiary generates X honey a turn used for honey packs, stored just like materials or anything else would be. It might get complicated depending on how many extra resources there are like this but it also opens up more options for trade and gives you time to retake lost resource nodes while still allowing you to make your staple soldiers. In addition to this if your unit design calls for a material you don't have, you can still create it but lacking the special item you can't afford to have on them. In the case of precious metals it would make Iron plate mail and weapons instead of mithril.

I like option C or A.  If and when an upgrade feature for units is incorporated (ala GalCiv) then either of these methods should work fine and enable you to get back on track when the resource is resecured via force or diplomacy. 

With Option A most players will probably have several alternatives already designed and might want to resort to one of those.  Redesigning a new unit wouldnt be that much of a chore, and is kind of fun. I would bet a good % would do this.  I sort of doubt we will have the sheer number of towns to make this that much of a downer. 

Option C sounds as if it might be the nicest method to implement.  An upgrading feature would be awesome.

Option B seem to damage the strategy aspect of resources.  If you are going to do this one then just let everything be enabled from the start and have the building times reduced as techs and resources are accumulated.

Secondary weapons was talked about, but I don't think it was "confirmed" (as much as anything can be confirmed in a process this fluid!). They seemed to like the idea though.

Unit skills as mentioned in other threads I thought was about champions rather then normal units. Hence the question.

I choose A with an option to store resouces for a rainy day. This makes resources more valuable and gives the incentive to capture and keep resources.

Instead of time, why not simply make the unit more expensive at the time of ordering? (simulating that resources need to be imported at greater costs). this btw, is also the system i advocate for galactic civilizations 3 if anyone's taking notes.

Ultimately the answer to this is tied in to exactly how complex the economic system is in the first place, I can't think of addressing one without covering the other, so apologies in advance if I ramble on.

Questions:

1: What does a unit "cost"?

2: What makes designing a unit worthwhile? What about the design process makes the unit different?

Disclaimer: All numbers used are made up, game terms used equally so, just for points of reference.

1: Costs and economy.

For creation, the first method is that the unit needs some notional investment of "production"; presence of the necessary materials; somewhere to build it and presul ymably a city building enabling this production. This is the Civilization method - These only need to be present, gold can be used to "hurry along" progress, and it's all very stuck in game logic as to why a city constructs only one thing at a time but can apparently do so at the speed of light, but runs out at the speed of light too. It's very hands-off and stylised.

The second method is the Warcraft method. You have a building to build with, you send your peons over to gather the resource from everywhere, and you spend those particular amounts on construction. A time-frame determined by the unit involved is your "training" time. This is a naturally more involved method, bossing around the minor day-to-day activities of your peasants. If a resource is cut off, you have reserves stored in some magical ether. It's not inherently very compatible with a turn based strategy without a little work.

To cover this question, an ideal compromise I believe would be thus:

A) Resources are gathered and spent, not binary states.

In reality, large-scale centralised industrial ventures are a rather new thing. Coal mines didn't happen on some single isolated locale, small mines dotted the mountainside.

Each standard resource has a numerical amount, gathered by your cities and gatherers at a steady, but low, rate per turn. Have a city in the mountains? Each mountain tile has small mines maintained by your populace. Grasslands have farmlands, vital for producing the necessary foodstuffs for a large city. The rarer a material, the less each square would yield, and you must have the appropriate technical capability (and potentially building) to harvest each material. Remember MoM's surveyor? I do. A surveyor could say what a city in each area would gather based on its known surrounding tiles.

Standard type resource nodes (adamantium would be standard, magic honey bees not) produce a larger amount of a selected resource - No special gathering item is necessary on the tile itself, it simply needs to be within range of the city.

If you don't want a city in every spot, gathering points (either universally appropriate trading posts, or possibly ones that give bonuses to specific resources, such as mines, sawmills and farms) can be built cheaply and easily but cover a smaller area (if a city covers three squares each direction, then a trading post covers 1 square).

For non-standard resources, a specific tile-based improvement is required to harness it - Honey in a honey tree? Beekeeping structure. Pack of minidragons? Either a hunting lodge (to harvest those skins), or a training lodge (to train up riding dragons). Non-standard resources should be big, flashy, and provide a potentially massive change to a unit equipped with it. Likewise, any choice between two optional uses of a resource should involve a significant trade-off either way.

This trade-off might even be possible for standard resource nodes. Grove of trees - Great source of mystical yew with a sawmill, or create a druid's grove instead, and let the tree huggers generate mana and unusual unit options? Farm for food and cottons, or ranch to supplement your horse cavalry?

This method of cost is a little bit of A, a little bit of B. You might run an entire empire on iron without a single iron node if you have enough mountains, but it's unlikely you'll gather enough Adamantium from that same empire to make up a single unit of it per turn (and it takes five units to make a single adamantium armoured soldier). With unique materials, you will never be able to make a dragonknight without a pack of dragons, or taking the time to tame at least a few before hunting the rest, nor will simply having the resource be a white card for unlimitted production. If you want an army of tanks, you will need to conquer as many oilfields as possible, not just the one.

The maximum storage available for each resource depends on the number of appropriate buildings in each city for storing such materials. Here lies another aspect of choice B: You can probably magic up some materials, buy some off your neighbour, barter from your neighbour, disband a unit to reclaim the equipment, or flat out steal it (or have your intrepid heroes help. Kill an adamantium golem? Free adamantium!).

There should be two types of universal resources - Gold (or equivalent), and Mana. Gold can be used most easily for trade, whilst with the right spell, minerals can be called forth from the earth, converted from type to type, and generally the laws' of physics should be magically shafted up the large hadron collider. Since it's likely that there is private ownership, surplus can be automatically sold "privately". Likewise, a deficit can be made up - if your nation has a resource available at all - by "buying" it from privately owned sources at a premium.

For conquest, conquering trading posts/mines would both damage their ability to gather as well as steal a proportion of resources relative to the trading post's area production. These same areas destroyed by NPC monsters destroys a portion of the resources and loots the rest, converting the post into something else. A farm slaughtered by undead might become a haunted farm, filled with manic attack scarecrows and children of the corn. A trading post raided by bandits might become a bandit encampment, serving to send more bandits into the surrounding areas. A mine invaded by kobolds and vulkans is an abandoned mine... All great places to go for adventures, both to recover the lost wealth and to seek treasure, fame and fortune. Possibilities are myriad.

Caravans can remain in this system, whenever a trading post exceeds a certain number of units of a resource, it creates a self-automated, self-defending caravan (for free, including a free, if small, mundane defence force). These caravans move directly to the closest town. Multiple caravans group together for defence. The minimum number of caravans to make a trip worthwhile can be set either globally or individually - A trading post next to a far-flung resource that would take five turns to reach a town would need a larger - and therefore better defended - group than a caravan that takes a single turn and so goes as soon as resources are available.

Finally, this trickle of resources can be a setting or a feature of a region/tile itself - bountiful terrain with a high level of passive acquisition, a barren land with little to none, or mineral-rich/food-rich... This would offer a highly variable system with comparitively small amounts of work on the individual settings.

B: Materials are useful.

The wide-swept assumption seems to be that materials become obsolete, like in Civilisation, the higher you go.

This seems a little strange, especially given the fantasy setting, and I truly hope that this hasn't gone in that direction. Every resource should be useful, and for more than providing some generic +1 bonus. Here's a few examples.

Iron: A highly common metal, untreated iron is fairly brittle, heavy, and generally outclassed by steel. Why use it? Iron is lethal to fey creatures. It protects against faerie magic, it harms them greatly, and it's better than bronze.
Silver: A softer metal, usually poor when it comes to equipment. Why use it? Silver is highly enchantable and great against lycanthropes. It's also rather valuable to private citizens in the form of silver jewellery, making it a good source of currency when its more esoteric uses aren't required.
Steel: Slightly softer than untreated item, but less brittle, steel, when it has been researched, comes from the same sources as iron. Steel is a jack-of-all-trades metal - good availability, decent raw stats, fairly receptive to magic and not too heavy... It's this quality that makes steel so universally applicable. It's not the best at any particular job, but it's never bad at it, and you could create an army equipped with steel almost as easily as one equipped with brass.
Mithril: Lighter than steel and highly receptive to magic, mithril is a great material, but hindered by its rarity, an army of pure mithril would be an incredible feat, and the advantage for your standard footsoldier of mithril compared to steel would be minimal, making it an expensive way of essentially showing off. Just because your sword is harder than steel doesn't mean the sharp parts are going to make you any more or less dead, and not everyone has the talents that make it worthwhile.
Adamantium: The hardest, rarest metal. Slow to gather, harder than any other metal, and legendarily.... heavy. Adamantium armour would give bonuses to defence as well as making the defence that's there more likely to be applied against the thing hitting it. Things covered in heavy adamantium won't be fast however, but it's still the only metal strong enough to cut other metal like tissue paper.

This is just the metal, and ignores all sorts of more esoteric materials, but shows how each has its place, its value, and hopefully goes hand in hand with the tenets of fantasy - Rare metal is rare, people using it are rare, armies using it exclusively? Extraordinary.

For unique materials, this goes slightly differently - there are potentially two types of non-standard resource, one I've not mentioned yet.

1: Gathered special resources - the unique honey from a tree, hunted ivory, venomous bloom, pseudo-dragon pets. All these are gathered from the environment already special in some way or another.

2: Created special resources - Enchant an item yourself? Or your hero found it falling off the back of a haunted lorry? That's a created special resource, possibly one of a kind. But this isn't the only means of producing a created resource. Consume food to create poisons or drugs (representing the sacrifice of arable land for sleep-flower production), work basic leather into improved leather, have your mages guild create some scrolls of fireball and much more.

How much effort this may be, and how far this is taken (have a smithy automate to create swords, or assume a sword comes from the raw material directly?) for the player is optional; each building in a city might be globally or individually activatable, quietly consuming resources to produce things whilst the main city production works on important projects (since it makes zero sense that a city can only do one thing at a time, military unit or building, ideally this production would include units as well, quietly consuming population and the necessary equipment for it, but I accept this idea may be unwelcome). A single menu giving global control (and a toggle to make individual cities ignore this global control) would make it easy to manipulate even the biggest empire, whilst a city-specific menu would allow fine-tuning and more controlled production for expensive, vulnerable or specialist units.

A magical nation might be able to gather enough mana to allow every unit built to have a magical weapon. A dwarvern kingdom might have squads equipped with dwarf-crafted extra-thick mountain plate armour. An elven nation might have superior composite bows and automatic machine crossbows... The goal with created unique equipment is customisation and versatility without clutter and tedious (and mandatory) micro management.

These then fall into three different potential effects:

- Unique equipment material with its own bonuses - dragonscale or rune-smithed steel as gathered and created equivalents.

- Unique material with it's own permanent advantage - Unlike equipment material, this would revolve around skills and permanent bonii - a poisoned weapon gives a permanant "poison" skill to the unit, magic honey gives 10% maximum HP, a bow (gives a permanent bonus of being able to fire arrows)...

- Unique one-time use abilities requiring resupply - Unlike permanent advantages, these abilities would be usable once (or however many times available based on the item), then greyed out and unusable until the unit enters the area surrounding a city or trading post (outpost is possibly a better term, and any outpost could potentially be upgraded to a village etc) where it will automatically resupply from available resources. Rather than magic +10 Max HP honey, why not magic "restore 50% HP to a target within 1 square" honey that can be used once? One off abilities provide variety and tactical versatility, yet still be handled on the equipment side of the unit design.

C) Training time ignores equipment.

What exactly is the difference between a knight and a footman? Or a knight and a squire?

It's not equipment. A footman with a mythril plate outfit is still a footman, he still fights as well, he still runs in fear from the Scary Ghost monster, he is still a footman.

The best part about Master of Magic was all the various units. The Dark Elven Nightblade is not just a better equipped Dark Elven swordsman. The thing that makes the difference is the abilities, and it should be these abilities that determine how long it takes to train a unit - churn out well equipped peasants if you like, it won't be cheap, but it will be quick. The time involved in creating the materiel is included in how rapidly you acquire the resource in the first place, and how quickly it can be turned into equipment. This is separate from the production of a unit, and is required up-front (you can't train a unit on sticks, then switch them to steel and expect no change). More on this in section 2.

2: Worthwhile unit design.

The quality of a unit is a combination of its training, its race (selected from the available population in your empire no less), it's equipment, and its special equipment. Attack, Defence, Health and Speed are fine, but there's so much more to a unit. I have to disagree with the suggestion that it's "what you equip them with" that's interesting, there's zero point in making individual armour materials changeable, only their appearance, and that for nothing outside of aesthetics.

There isn't the option to equip them with fighter-bays, a point-defence laser array, a few M.I.R.Vs and a subspace forcefield. A sword is a sword, an axe is an axe. Both swing, both are pointy, and the only difference is 1 point in stats that you never think of again after it's done. What's interesting is what the unit can do with their gear, unless that gear gives them a specific strategic impact by itself.

So.... What differentiates a unit?

A: Stats and gear.

Obviously. The stats are the most fundamental part. Inferior weaponry penalises the attack stat, lighter weaponry sacrifices some points of attack for more attacks per round, superior weapons might improve both. I don't believe that just four stats covers it (how well does it defend against magical attacks? Or fire? Ice? There's a fine line between covering everything necessary and not overwhelming the player (and I admit it zigzags a lot), but surely this is too shallow for an engaging design element? Worldmap speed? Combat speed (not necessarily the same)?

Equipment choice is just as important as equipment material, as is whether a unit has a primary and secondary weapon.

Primary weapon: Two handed weapons are good for penetrating armour, but one-handed weapons allow shields. Shields are incredibly good at defending yourself with (far more than is usually accounted for indeed), but a one handed weapon lacks penetration power against armour. Even the difference between a thrusting rapier and a slashing scimitar are important to consider when facing things that have inhuman resistances.

This falls roughly into "bludgeoning", "slicing" and "piercing" damage, and this doesn't cover "cannonball", "flaming", "chilly", "poison" or any of the more unusual units; even "biting" is a viable damage type here... And of course "magical".

And I haven't even mentioned ranged weaponry yet (which introduces the reload time for ranged fire), or armour (movement speed is every bit as fundamentally important to tactical gameplay as any stat mentioned, skirmishing raiders and flanking bonuses make for an interesting battle), mage weapons, or even two-weapon fighting.

Secondary weapons: "Weight" is an important aspect of equipment. Ranged units with just bows will fare poorly in melee combat using just their bows as staves, but spend a few more resources and sacrifice a little combat speed and you can kit them out with shortswords, or even give your knights some crossbows (or a wand of starbolts) giving them a secondary fire option they can use when the situation demands it (chosen automatically on situation). In this way, a squad of knights could take out the faerie archers with their iron swords, then whip out their steel maces to turn a group of skeletons to powder,

Armour: Equally important to weapon choice is armour. Ignoring the varieties for the most part, there's light (padded cloth, leather, salamander-scale, fur....), there's medium (chainmail is cheap, light, and if mithril probably counts as light armour), and there's heavy (platemail with greaves and gauntlets), each with varying degrees of impact on mobility and weight (this becomes important if you include terrain effects, heavy armoured knights fare sometimes lethally badly in the mud).

Finally, unit changers - equipment that vastly changes the unit. This was mentioned up in 1b - a ranged primary weapon makes for a primarily ranged unit. A mount makes for a mounted unit type for drastically improved health and mobility (infantry might go places a standard mount can't, however). A unit with a mage-weapon might default to being a mage-type unit, and can take mage abilities, each type with their own base stats.

Even whether a unit is male or female might have an impact (physical abilities ignored, a female would be the appropriate choice to handle a nymph and her charming smile for example).

B: Traits and abilities.

What makes a knight a knight, and not a footman? Just the name? What's in a name, will a peasant by any other name still whiff as badly, or fail to hit the broad side of a barn so sweet?

Skills and abilities are the lynchpin of variety - the things that make a unit mechanically different. I mentioned above a ranged unit with a bow, or a mage unit getting a wand, but a swordsman is fundamentally identical to any other swordsman regardless of weaponry... but let's give those two identical swordsmen some abilities:

Swordsguy A:

Sturdy (+10% HP and defence).
Counterstrike (Automatically attacks enemy units moving out of adjacent squares).
Indomitable (+5% attack and defence for each adjacent enemy unit after the first).

Swordsguy B:

Grace (+20% combat speed, +5% defence against area attacks that aren't directly targetting it).
Flanker (If this unit attacks a target which is adjacent to another friendly unit, it gets a 25% damage bonus).
Skirmisher (This unit can move and attack in the same round without penalty).

And there we are, two units that work together but feel entirely different. A is a defensive unit, built to keep units next to it and to survive the experience. B is a skirmisher designed to move in, attack, then move away again.

These traits would be the primary difference between different races, fantastic units and unit types.

A few more examples:

Universal:

Stealth (requires light armour): If a unit moves into certain terrain squares (trees, marshland, long grasses), with no enemies nearby, the unit becomes undetectable until it moves (unless moving through compatible terrain), shoots, or a unit moves adjacent to it. Once revealed, a unit cannot rehide until two rounds have passed.
Magical talent: The unit gains a weak magical ranged attack (limitted ammo / encounter), or if it's a ranged unit, improves its missiles to become magical with a bonus to hit and damage.
Fire Elemental (requires magical talent and "level 2"): The unit's weaponry attacks deal an additional 10% fire damage where appropriate.
Burrow (Fantastic monster ability): Unit burrows into the ground and can move at 1/2 speed in total concealment. It must emerge to attack.
Specialist (weapon): +10% ranged damage and accuracy with the appropriate weapon type.
Dwarf (Racial): +10% health, -20% combat movement.
Dark Elf (Racial): All units start out with magical talent.
Gnoll (Racial): +20% attack for melee units, +5% combat movement.

Ranged Primary Weapon: Ranged unit:

Focus-firing: Attacking a unit already attacked by other friendly ranged units this round gets a +10% hit chance.
Sniper: If firing unseen from stealth, the unit deals an additional 30% damage.
Explosive arrow (requires Fire Weapon and "level 3"): This unit's arrows explode on impact, damaging surrounding units.
DeadEye: Accuracy penalties due to range reduced by half.
Quickdraw (Requires bow): Reloading speed reduced, allowing more attacks per round.
Extra ammo: Doubles ammunition/encounter.

Mage weapon (requires unarmoured): Mage-type unit.

Spellpower: Ranged magical damage is increased by 10%.
Overpower: Total magical shots available per encounter is halved, damage is increased by 50%.
Healer: Gains a one use healing spell per encounter.
Armoured mage I: Can be equipped with light armour.
Warrior training (blunt): The unit can use primary blunt melee weapons instead of wands.

This is naturally just a few, default level abilities, but serves to indicate the massive array of things that can be achieved with them, even with a standard unit.

Skills versus Equipment:

You can refit equipment on any unit in a town, but skills and traits are permanent (and accumulate a few more, as the unit approaches their at highest level), defining parts of a unit. Unlike equipment, which has a fixed number of slots, each unit can accumulate, say, up to three traits on creation (and optionally, given an automatic level-plan to follow straight from the gate), and end up with three more skills (and with various prerequisites of other traits to go). A unit created with no traits would train more quickly whilst three abilities would take many turns (dependent on city quality and the particular skills), but the 0 trait unit could only pick three traits (and without prerequisites, they would be limitted to weaker skills in general) meaning there would be a permanent power difference between a highly trained specialist and a cheap auxilliary.

Equipment would stay as Weapon Slot, Armour Slot, Shield Slot, Secondary Weapon, and two or three special items and never increase.

Fantastic units may have many more, but also won't often get more, or have any variety if they do, whilst heroes would have the most potential traits available, determined by their basic properties as heroes.

This trait system and combined material system, or something along similar lines, would provide meaningful design decisions - choices that drastically alter the behaviour and capabilities of a unit rather than bland numerical tweaks.

Well, there we are. Sorry for rambling on, I do love the subject matter. Hope it was at least interesting for everyone who read it.

Thank you for saying what I have been thinking, and too lazy to put into words, meaningful unit design is a must.  Your ideas are great and well explained.

I really liked the idea of skills. I would love to have it tied it with unit level rank.

This how I would like to have it:

1. Rank

As of now units rank will give unit 1 extra hp at cost of one extra turn to create. So they more or less useless. This how I see that ranks should be handled:

a. Militia

 That's your basic volunteer or draftee if you lucky he'll probably would know which side of sword he should hold. They have no skill and penalty to use of weapons and heavy armor.

b. Recruit

This one had a basic training so he most probably will not cut himself. Also after lots of running his stamina has improved. So basicly recruit will have more HP, will have smaller penalty, and have 1 basic skill. The cost is few additional turns in training and little increase in maintenance. Of course unit can gain rank from combat.

c. Regular

Well after lot's of training this one can be called a soldier. He totally got used to his equipment so he will have no penalty. He also will have access in addition to 1 advanced skill. The cost is same as with recruit, larger build time and larger maintenance. 

d. Veteran

After surviving many battles veteran honed his skills so he will receive small bonus to his main weapon and additional advanced skill. Also his HP will increase. Veterans can't be built only created through battles.

e. Elite

The last but certainly not least, Elite is unit that been there, done that, have t-shirt to prove it.  They are masters of art of war they get 1 expert skill as a result. Their HP is nothing but legendary. They receive bonuses to both primary and secondary weapons. The cost is extreamly high maintenance and you can only create them from veterans.

2. Skills

I have divided skills as you can see from above in to 3 categories: basic, advanced, expert. Basic skills are skill that give minor bonus to combat stats. Advanced give nit some minor ability such as for example Fire Arrows or Shield Wall. And expert skills are battle changing highly specialized abilities such as Assassination or Sniper Shot. Skills should also be available based on equipment and could be retrained at towns for a "small" price.

3. Unit size and organization

Currently unit size simply gives you stronger unit for more money, so there is no downside not to use largest possible force instead of many small ones. I believe there should be both positive and negative to having large unit. For one large units should be slower as large army always slower than single man this penalty could be mitigated by "organized" skill of hero attached to it. In return larger units will have access to better skills. For example if unit is contains one man it will have access to "Cover" which will give him small bonus versus ranged attacks while larger unit will have access to "Shield Wall" which will give larger bonus and the largest unit will have access to "Turtle" formation which will give them total immunity against ranged attacks.

Sorry didnt read the entire thread so if this idea has been mentioned please move on...

Option D:  Allow any equipment combo unit that tech allows to be built but give equipment additional abilities or bonuses for having certain resources. Ie the honey pack gives 10% bonus hps normaly and +1% (or whatever) bonus per twilight apiery you control.  This gives having multiple resources greater meaning without making it completely debilitating not having a particular resource available. 

This will lead to much greater unit diversity and varied land management strategies.

I like option A - Can't Build Unit.

Why, because I don't want to end up with ten different versions of the same unit and I want the constant reminder that I need to find or get that resource back!

That's close to what I was thinking, and it makes more sense than mine. I like it. A raw recruit can be quick to build but have no skills, then accumulate further skills as they level up, or build a more advanced unit from the get go.

Only thing I'd think (and really this is just tweaking) would be two basic skills (recruit and regular), and need to fight for their advanced skill. This would let advanced skills be slightly better (since you can't get them for free - or without the highest-end town improvements just to get one), and shape the unit better (and again, defaults would hopefully be settable from the design stage as well as picked manually).
If player traits have anything in common with Master of Magic's, a trait that allows higher max levels (and more advanced, or elite, skills) would be great for someone dedicated to military over magical power.

I'm not opposed to retraining skills, though it makes them a bit more like equipment I can see a general "refit" option for a unit like you'd see in Master of Orion. It does seem that caution may be appropriate though:

For example: You have your Elite "Bronze Spearguy" you've had since the beginning. He's got basic skills, advanced skills and an expert skill - all basic, early game ones since you lacked the magic and tech to open up anything better back then. Mid-game, you retrain him. He's now an Adamantium Swordmage Guy, he has special unlocked Elite "laser death" skill that you just unlocked with research. As long as elite ex-spearguy is alive, you have free access to any expert skill on demand. That, depending on just how useful and powerful an elite skill can be in the right situation (since you can always get an elite skill for the right situation like this), could be more powerful than you expect.

I was wondering about formations in the game, nice to hear they exist at all.

In the same vein of ensuring "meaningful design", there are some things here I'd suggest -

Each unit is independent.

More units gives stronger/healthier units? Sure, but I sincerely hope it's measured in the same way Master of Magic did it - each unit has their own state of being, their own health (though damage could be divided among them), and their own little attack each turn. More people aren't going to be too helpful against units that attack every unit at once. This opens up a lot of these later suggestions, though in itself doesn't do too much.

1: Terrain versus size.

A massive army might be great on an open field, but in a swamp, each member of a unit would have a (low) % chance of taking damage, getting lost, or some other disadvantage (which may be as minor as -1 attack or half-move for a turn). A small ten man squad on terrain with a 1% chance per unit in the terrain of losing a man will lose, on average, 1 person each ten turns or more.

The same terrain with a five hundred man squad will lose 5 people a round. Implement half-move a turn at 1% chance, there's a high probability that 500 man squad will be stuck in that swamp (armies move at the speed of their slowest member, of course), a 10 man team might have a member fall in a boghole and need rescuing (slowed, roughly 10% of the time), but the army chasing them almost definitely will.

This also works with positive advantages. Hiding in a wood? You have a 50% chance of getting into cover, giving you great anti-range defence. Ten guys will be risking five on average. Five hundred guys have hundreds of people standing around being vulnerable to archers.

Other, non-terrain aspects are: Ranged fire - shooting into a crowd of fifty people gives you fifty chances to hit somebody (a sizeable to-hit bonus), and Area attacks - Setting fire to fifty people works in much the same way as setting fire to five.

Finally.... Experience. A 25 man unit gets 10 Exp for hurting a wookie by 10%. This results in getting 0.4 Exp each. A ten man unit gets 1 Exp each. Working alone might be tough, but it's great for advancing quickly if you can do it.

2: Rows.

A natural aspect of formation is that not everybody can get to swing their sword at someone at the same time. This instantly limits the quality of an army of swordguys. Having a hundred men means ten can hit something at once. This means it is ten times stronger than a 1 man unit offensively, but 100 times stronger in terms of health. A unit of 25 guys might be in rows of 25, so the 100 man squad is only twice as strong offensively, four times stronger in terms of numbers. Keeping things simple, 1 and 2 are single row, 4 up to 10 are two row 12 - 20 are four row... all the way up to maximum numbers. A theoretical 900 unit army would be 30 rows of 30. Capping a unit size smaller and simply having more units is more practical however. A maximum of 25, or just 20, would be ideal; you can still get an army of thousands, just have a lot of squads.

Damage would be divided over the front row of units each round, but wouldn't "leak" unless a unit had a special skill that let them reapply leftover damage. This would obviously be very powerful and common to danger monsters like Dragons.

3: Size constraint.

Not all creatures are created equal. People might fit 25 in a squad, but in that same 25 square space, you could only fit four ogres, or ten cavalry. This makes formation size a valid power of a unit, and opens up a "Gemini" ability (1 "unit" is actually comprised of 2 or more creatures, great for a bee swarm - Gemini 10 - or some highly trained halfling ninjas - Gemini 2). Huge squad of killer bees? Or a tiny squad of Minotaur berserkers?

4: Mixed units.

Having kept them simple, time to make them complex again. You mention formation, but this is more than simply where a fella stands.

Why is a squad composed of all one type of unit? Even the Roman legions and Greek Hoplites mixed things up a little. Mixed formations could allow so much more, if the amount of effort put into creating them (game and programming both) is worth it, then Squad design could be every bit as interesting as unit design.

Mixed 20 man unit:

Front row (the front row is the row that is attacked, and attacks, in melee): Shield-guys: Shieldguys. These guys have defensive skills.

Second row (second row units require spears, the right skills, or ranged attacks to do anything during combat): Spearguys. These guys have long spears and second row support abilities that help them cooperate with the front row defenders better.

Third row (third and fourth rows of this mixed unit can't melee without thrown weapons or spears along with a special skill.... something must enable their ): Healers. These guys heal units next to them. Advanced healers might even have raise dead 1/encounter to bring back 1 dead unit each, or ranged healing to let them heal entire units.

Back row: Skirmishers. These guys have hand crossbow secondary weapons and shortswords, along with the special unit ability "separate". When the ability is used, the skirmisher row moves out of the unit and becomes independent. Their presence is two-fold: firstly, they can separate and use superior speed and ranged attacks to go after vulnerable or exposed units; secondly, their presence in the back row serves to protect the squishy healers from flanking (see section 4b). Skirmisher units with the right skills can leave and rejoin their parent unit as they desired, much like in an ancient Greek Phalanx.

For another advanced formation, this could have Shieldguys in the front row, axe guys with "cooperation" skill in the second row, letting them attack from behind the protection of the shieldguys, and finally a third row of spearguys with the same skill (for the historically minded, this way the formation used by ancient Pict warriors in battle, and was supposedly highly effective). This unit has incredible melee potential, but the units behind the shield guys are vulnerable to flanking, or when the shield guy generally falls down.

A pure unit has no versatility, but has no variability, at 25 guys, it hits as hard as it does with 5 guys. No matter where you attack it from, it has swordguys to handle the attack. A mixed unit sacrifices this cohesion and specialisation (a squad of 25 bowguys will always be better bowguys than a squad of 10 swordguys and 15 bowguys, and probably do with different, more focused bowskills). Take this variation down to a unit-by-unit measure, the possibilities are endless - a "core" of mages surrounded by swordguys?

With a mixed unit, the possibilities are boundless, and with a good UI, potentially very user-friendly. A picture is worth a thousand words, so.... http://i29.tinypic.com/2amavk.jpg. I apologise in advance for the scruffy image, but hopefully it's enough to serve as a broad indication of how this might be achieved.
Not included are Large Units, which could be fantastic units (ogres that take up 4 squares each?), or the possibilty for creating formations out of individual units present at a location (such as combining two small units, and even incorporating a hero into the squad as well). This meshes well with the divided experience too, get a huge bunch of individual units, swarm someone with them, and combine the now experienced survivors together in a group. Perfect tactic for an evil warlord, right?

This would be similar to Refitting a unit and would take a minimal amount of time (1 turn to reformat?), and naturally certain heroes would be better leading squads than they were alone, and vice versa.

4b: Flanking.

Something fairly irrelevant when every unit is swordguys, flanking would be highly important to a mixed unit game, as would skirmishing, surrounding enemies....

Normally, two units fighting, the attacking unit would attack the front row, and the defending unit would defend using theirs. In a tile-type system, the unit would be assumed to be presenting its "front" to the square in front, and the two squares to either side. Any attacks to these parts would meet the front row units.

If a unit was attacked from the side, or the two units diagonally behind the unit, they would meet with the outside units on that side, with the backmost square hitting the back square, like so:

X X Y
Z O Y >
X X Y

Y hits front row. X hits side rows, Z hits back rows, > is the way the unit faces.

In the image, this means the front row of stealth guys (who we'll call "Waldo" now, or Wally for the British), will bear the brunt of all attacks from Y positions. If an enemy attacks from X, then they'd meet the side row - four Wallies and 1 Skirmisher Guy. Finally, if someone manages to get into spot Z, they unload on the poor, vulnerable 3 Cheerleader gals, plus 2 Wallies.

Note that in all these cases, the Attack Magic Guys get a shot, since they've got the appropriate ability, despite being protected on all sides, so long as they haven't used their ranged attack to shoot with the previous turn, and haven't been attacked twice in a round. If the Skirmisher guys had a second-row attack ability, like a spear, they could attack every time regardless of prior attacks.

When a row is removed, rather than moving things around, the attacks would carry on to the next row. If the back row was attacked from Z, and all 3 Cheerleaders were killed, then 2 Wallies and 3 Attack Mage Guys would take the hit.

Finally, if an entire row or column is wiped out - If in a flank attack from X managed to kill all four Wallies and the Skirmish guy, when the squad turned and attacked next turn, their front row is now only four Wallies against five swordguys, so rather than dealing with one attack each, the extra unit applies their attack power as a bonus, divided across the four Wallies.

Hopefully it becomes obvious why I keep mentioning skirmishing, mobility and powers that restrict them as being so important. The five skirmishers can slip out of formation and tear into the lightly defended flank of a unit, the Wallies might have the Counterstrike ability, making it painful to move around the unit to get into flanking position.

Note that there are such things as dragons in the world, where attack direction is still important. Y would encounter their breath weapon, their claws and their jaws. X would encounter their wing-swipe attack, Z would face their lashing tail. A special monster - the hydra, for example, would meet attacks to the front with its heads, getting attack bonuses with each head destroyed without losing any health from it. Hitting it from X or Z positions would attack its relatively defenceless tail, a Manticore you might regret attacking from Z on account of its poisonous tail...

All this on top of simple bonuses for cooperative attacks, attacking anything from behind and other features.

5: Ranged attacks, skills, and mixed infantry.

Again, I'll be using the squad in the image as my example.

The last aspect of all this is how ranged attacks work. With no skills involved, if two identical squads face each other, it's fair to assume that the six attack magic guys hitting a unit with their magic missiles will spread their damage amongst their target evenly, probably dealing more damage to Cheerleaders (low magic resistance), whilst the Wallies fare much better, and this would be the case if the unit involved was firing into melee (a ranged unit attacking an adjacent unit would use their ranged attack, but invite a melee counterattack), defending (those attack magic guys wouldn't just attack the front row, but the entire attacking enemy unit), or firing from a distance.

Each archer has a 1% bonus to hit per unit in a squad - A single guy may have a 51% chance of getting hit by those 25 archers at normal range (taking an average of about 13 archer's damage, ouch?), but a 25 unit squad will have a 75% chance - 19 archers, divided by 25, taking an average of 1.4 archer's damage each). Doing the maths, the optimal squad has the lowest bonus to be hit, but divides the damage amongst as many units as possible (I'd guess about 15 units, for 1 archer's damage worth per unit).

If we introduce possibilities that skills open up, Wallies can have "Big Shield" - allowing them to share their high ranged defence with their neighbours (with the upgrade "really big shield" which covers the whole group), whilst Skirmish guys can have "Bodyguard" - They take any shots that adjacent units would take (again, a Martyr advanced skill would let them cover a whole squad and then divide the resultant damage between them). Now these two are cooperating, Wallies share their defence bonus with their neighbours - this means only three of the team don't have Wally defence, one Cheerleader and two AMGs. They take damage as a result, but the damage on the adjacent AMG is absorbed by the three higher HP Skirmishers that are next to them (either diagonally or directly).

I'm sure it's getting old me saying it, but the possibilities really are endless for skill composition. "Legion" skill users gain bonuses with other units in their squad with legion abilities. "Defenders" could rearrange squad formation to always try to be on the frontline of any melee attack, "Leadership 5" give bonuses to all their squad members with "Leadership 4" or lower, squads consisting of several different, highly mobile, skirmishers could split into five mini-teams on command, surrounding and killing their target.

On the other hand, single units might have "Loner" skills, special abilities that provide better benefits based on a small unit size - An invisible unit has an easier time staying invisible alone, a berserker might attack friendly adjacent units in melee as well as the enemy (though for a lot), a "Heroic" fighter (or, indeed, a hero) might actually gain bonuses for being outnumbered.

This could still be used with a skirmisher as a "Payload" of a larger unit. Approach in the safety of a unit, separate from the unit at the appropriate time,

These five methods all serve to give unit sizes and composition consequence. as well as provide tactical variety and choice to unit based combat. I'm sure they could be built upon for polishing or improvement.

And thanks Adaban, I'm glad you liked it.

Keep in mind that a "black market" wouldn't really be "black". :/ nothing to regulate them and I doubt that it would be "illegal" or undercover by any means.

If you have a concept of a market, it should only be available randomly from turn-to-turn, prices should vary, and so should the quantities. There is a game that I played several years ago that had traveling merchants you could purchase more rare or scarse resources from, though I forget the name. Basically there were a few "caravan" units that roamed the map (and were oddly invincible) and they would travel from city to city. If they reached your city they would open up a dialog box at the beginning of your turn asking if you'd like to look at their stores.

It seems like one the major problems is keeping the flexibility and fun, but reducing the micro management. 

To mitigate this, in the unit designer, it would be great if you could specify a backup for each item.  If the primary equipment is not available, then the backup item automatically replaces it.

It is not necessary to specify the backup item during initial unit creation.  Typically, you would wait for the situation to occur, and then select the backup item once.   This way, it is no more work at the beginning, and in the case of a highly contested resource, you are not constantly adjusting your units.

The second problem is with unit diversity.  It is not good for a single unit type to vary.  It's complicated and confusing. Nor, would I want to individually upgrade units with backup items to the full unit, which is a tedious and annoying distraction.

To mitigate this, all units of that type would transform over time to the same abilities.  If you currently hold the resource, then when the unit gets resupplied, then it will get the primary item.  If you don't have the resource, then all units will get the backup item.  Basically, if you are able to supply the Twilight honey to your troops, then you will, otherwise they will get the backup item.  Likewise, if you no longer have mythral supplies, you can no longer repair and maintain swords and armor.

I'm not sure when the resupply would occur, but here are some possibilites

1.  Over time -- X number of turns (modified by warehouse tech, as proposed by other threads)

2.  After each battle -- X percent chance of equipment upgrade or downgrade per unit (Broken items for downgrade, presumed resupply for upgrades)

3.  Visit a city -- Logically, I like this for upgrade, but not for downgrades.  I mean honey could go bad, but why would armor suddenly break?

I like this because I still get to tweak my units, but I am not making the same decision over and over.  I make it once, and then it is handled.  I also like the concept of the units becoming uniform.  If you are a society that can provide twilight honey to the troops, then it will eventually propagate out to everyone in the field.

It also means you get to make strategic decisions when you lose a resource:

1.  If it is gone for good, then go to the unit designer to prepare plan b

2.  If it is gone, but you can get it back quickly, then specify a backup item for the short term, but you probably won't need it

3.  If it is going to be hotly contested, then you can specify a backup, and also upgrade your warehouse/resupply tech.

 i love the idea of customizing the look of your troops. id like to add a suggestion. not sure if its come up before but how about after x amount of units grouped together as a fighting force could there be a standard bearer on the battlefield. non combat of coarse i just think poeple would like a flag bearer representing their kingdom along side the army.