It doesn't provide choice at all, and it doesn't allow you to freely choose to only use certain units unless you want to deliberately shoot yourself in the foot. The point is to have meaningful choices - "do I train all my exceptional units or train all these rejects?" does not present a meaningful choice because one is inherently superior to the other.

The beginning of a game is always immune to this problem, because economies, army sizes, technology, research etc are never at a point yet where it's relevant. The problem only arises once your economy is in place, your resources secured and so on. But once the problem arises in a system that doesn't provide meaningful choice, then there is only one real option for someone who doesn't want to handicap himself.

No, you completely missed my point. With a limited number of exceptionals, once the dust has settled the superior choice will be to use up all your exceptionals, then all of the next rung below them, and so on down the chain.

If there are good reasons to continue building inferior troops (possibly due to training costs, upkeep, etc) even if you haven't exhausted your pool of 'exceptionals' then your mechanism doesn't really accomplish anything because the problem is already solved. If there is no good reason to build inferior troops until you have exhausted all your superior ones, then your system results in just as little choice as we started with: the only difference is armies, instead of being composed of your best archers, your best infantry and your best cavalry, will be composed of as many of those as you could possibly train, followed by as many of the next rung of 'combat potential' as you could possibly train, etc. It breaks the monolithic army, but it doesn't add choice or strategy. It forcibly prevents a player from trying to field a small but elite army (hard cap on the # of 'elite' troops), and there is a clear superior strategy.

Your suggestion doesn't solve the problem, it just pretends to; by which I mean on the surface it appears to solve it, but in reality it just changes the nature of it.

That's my point, in disguise. If you can have so many exceptionals that you don't need anything else, and there is no reason to train inferior troops unless you have exhausted your supply of superior troops, then it means the superior strategy is to just train exceptionals all day long. If you do have a limited number of exceptionals, but there is still no reason to train inferior troops unless you have exhausted your supply of superior ones, then the 'low hanging fruit' will just the highest rung still available. If you've exhausted your exceptionals, you will move onto your superiors; if you've exhausted your superiors, you'll move onto your mediocre, etc. But trying to field, for example, an army consisting of almost entirely mediocre units with a small handful of exceptional ones would be a losing strategy, because we've already determined that training the best that is still available before moving on down the ladder is the way to go...

On the other hand if there is a reason to train inferior troops even if you haven't exhausted your pool of exceptional troops, then all of a sudden there are meaningful player choices. Do I focus on an elite core supported by a large number of inferior troops? Maybe. Do I focus on a relatively small but entirely elite army? Maybe. Capping the number or exceptional troops in this situation accomplishes nothing because there is already reason to train a mix of troop qualities.

Sure, that would provide some minor strategic choices, but it would also be a real headache... Also, it assumes that someone with great combat potential is also going to make a fine carpenter, or blacksmith, or whatever. I'd rather do without any sort of citizen-ranking system, unless they implement governors and things like that, but those would be special individuals rather than numbers on a spreadsheet.

What i would like to know if the devs consider steamrolling battles using hordes of the same unit type to be a problem?

Take for example the historical Battle of Agincourt, where an army of 5000 British knights,men at arms and archers defeated 30 000+ heavily armoured mounted French knights and their men at arms.

 The powerful British archers and careful positioning and use of terrain allowed the British to pull off a victory against a vastly superior foe.In fact, historically, the French army's heavier armour actually worked against them by slowing them down and was one of the reasons for their defeat.

Now lets say we are faced with exactly the same battle in WoM and we keep replaying this same battle scenario. Would the 30 000 armoured knights keep steamrolling the 5000 British mixed army no matter what you did? If yes, then do we think such battle mechanics as fun, no fun, or it is not important?

at moondoggie, I think it would be wise to enable well-played (and perhaps lucky) smaller army to defeat a poorly played (and perhaps unlucky) large army.

Also, poorly-organizing a super-large army, even with victory, could quickly become a Pyrrhic Victory. Where-as leading-organizing a large army well could be a viable strategy for victory against a smaller elite army. This is what I would like to see.

say, a double strength elite units army, but outnumbered 1:10, could kill many times more their number than simply twice the number (as implied by being double strength) if the large army plays poor tactics. And if the small army also plays very well tactics, up to say 90% of this army could be destroyed ... effectively doubling the usefulness of the elite troops if played well

Meanwhile, elites poorly fielded against such an army could kill below twice their number of enemy troops (such as being surrounded/flanked/ ect ... and perhaps even kill as low as only their number (against half-strength troops, which would mean a 50% loss of strategic effectiveness)

Ergo Tactical Advantage can emphasize and evolve the strategic usage of certain units .... the battle-field being a determining factor.

Well to me it seems like this will simly be solved by 3 things

First, availbility of resources.  i.e. Iron ore is far more rare than leather.

Second, time to craft and equip. i.e. The time it takes to make plate mail is far longer than the time it takes to make studded leather.

•  Because of the resource limit and the extra time it takes those 'uber units' to be made they essentially become rare.
•  It would make more sense to mobilize a greater proportion of you population in leather than it would be to attempt to wait until you have enough plate to equip your entire army in it. 
  
• What this would mean is that while you would still have a contigent of 'uber dudes' wearing plate in your army but most of your army would not be uber due to resource and time constraints.

Third, these 'uber units' are not really all that uber, they may be very good at one thing but quiet simply, their strengths can be accounted for in tactics, like frogboy said, some shitty mounted archers will destroy some 'uber footmen' as long as they do not engage in melee combat.

• Your uber units will not universally be good at everything and should be able to be countered in a relativley inexpesinve fashion.  This means that you need to protect your valuable units until it is their time to shine.
  

Basiclly all of this is contigent on the differention is strength of units being somewhat realistic in comparison to each other.  If we are talking about 1 knight killing 1000 peseants than we have left a realm that we can balance for.

Holy crap Tasunke, what world do you live in? A knight killing 100-200 peasants that have him surrounded? Nah uh. Maybe 10, if he is really luck and skilled. Nor would he be able to kill 20 or 30 peasants at a time, nor would he be able to continue to kill round after round of 20-30 peasants. Killing that many units would be exhausting, and he'd be absolutely spent long before he came anywhere near defeating 1000 peasants. A lone knight would not really be able to defeat very many (determined) peasants at all, in my book.

Unlimited exceptionals, leading to armies of solely/mostly exceptionals, is to you better game design?!?!

We'll have to agree to disagree on this.

You're focusing on the wrong type of 'choice' -- you're focusing on the obvious and unarguable 'best to use best units when all else is equal'.  That's a no-brainer.  The real choice starts after that, and that's what I'm addressing, and to maximize choices past that point is why limiting exceptionals is proposed.

You're arguing the obvious 'apple' when I'm talking about the 'oranges' that come next.

As an aside -- and moving beyond your 'apples' -- you're also making assumptions, as we don't know how things will be designed so you can't say your above scenario is best.  It may be that lessers can be fielded in sufficient numbers and timely manner, and exceptionals in such limited numbers, that your scenario isn't true.  It may be that a 'bottom heavy' army in some situations it better than your 'top heavy' (hopefully this will be true).  We just don't know this yet.

One thing we do know is that the more exceptionals that are available, the more preferentially using them ('top heavy' armies) will be the 'best' choice (once they're available) so the fewer viable choices there'll be.  That's why limiting their number increases choice.

You're also ignoring the basic idea that exceptionals are by definition limited in number.  That this is a 'fantasy' game doesn't preclude using such 'realities' -- especially as it increases choices.  Even 'fantasy' has rules, and if they improve game play then using them is a good thing.

As the rest of your response is predicated upon this I'll not address it further, except for...

If you are limited in the number of knights and are still training them, you don't stop training footmen until knights are done  -- unless you have unlimited knights which obviates the need for said footmen.  Your unlimited exceptionals is what causes what you decry.

Say one's empire can support 1000 troops (as troops numbers depend on population/etc.).  Let's use your 10% number (easier math  and let's assume that the game mechanism would be that each recruit has a 10% chance of being exceptional so the number would vary slightly according to chance ) -- there could be at most ~100 exceptionals.  You'd have to increase you base population to get more (and perhaps there'd be a game mechanism to increase/decrease the odds a slight bit, so one could choose to pursue a top heavy or bottom heavy approach).

How you'd get them is up for discussion (special resources, special training, surviving a number of battles, etc.).  They'd be relatively expensive to get and maintain and take relatively long to field (thus balancing risk/cost vs reward).  Their numbers would be small enough that by themselves they'd not be game winners (unless one is very good at tactics&strategy and has designed their empire to maximize their effect vs their particular enemies and one can survive long enough to field them in sufficient numbers, etc.).

How their 'exceptionalism' is defined (endurance, strength, ability to utilize certain gear or spells or the like, etc.) is open for discussion (and could be determined by the players, subject to their race/available resources/whims/etc.).

I'm just looking for a way to encourage using a mix of units, adding choice.  If we don't want armies comprised of only the 'best' units then they have to be limited in number by some way.  Training time, cost, etc. only works in the early game.  In time, that's no limit.  So, some other way to limit their number is necessary, and thinking of units as having 'potential', and some units having exceptional potential (which is realistic, as no amount of resources/training/time/experience will turn me into a Michael Jordan or Muhammad Ali) is one such possibility.

There could also be 'inferior' recruits -- they can't be trained beyond 'peasant cannon fodder' skill.  Giving them better training/items/etc. would be fruitless.

As an aside -- is there no 'preview' option for those of us (like me) who make too many typos/mistakes? 

Lets not forget army composition. Cheap/weak units can always be used as a buffer zone to protect your more valuable units from getting hit. I've done that often in games and I'm sure it'll be a tactic I use in Elemental. If I have a caster unit who is weak on defense I'll surround him with a bunch of "expendable" units so he can cast safely without worrying about getting killed by anything but direct damage.

Well, I would prefer to see organic incentives to build a diversity of unit caliburs.  Every army doesn't have to be an even mix, but when people do make armies of almost completely elite units or almost completely lowbie units, it should be because they have some kind of strategic advantage on the map or that they have pursued to their advantage.

I don't mind forced incentives if they work, but you can disguise them if you want: a elite unit gives a bonus to normal units (command bonus), but if there is more than elite unit you get penalties instead on your units (contradicting orders). The posibility of having 2 elite units exists, but it's discouraged.