Have you ever been bitten by a grasshopper?  I'm a gardener, so I know that when they do bite they can hurt   Even a single lucky grasshopper can bite you in the eye.   If 1,000,000 spears are thrust at a dragon, are we to assume that not a single one ever hits the dragon in the eye, strikes a molting scale, jabs it in the throat as it rakes soldiers into its jaws, or jabs it where its tallons meet its scales (thus opening up a wound and, hence a larger vulnerability)?  But this is beside the point, and like I mentioned earlier, the issue becomes mired in people's personal resolve with how uuber they want dragons to be.  I would appreciate it if at least 1 person in this thread would address my point on the grounds of what would make meaningful game mechanics rather than what gratifies their intuition, anyway. 

Comparing the strongest unit in the game to the weakest is really beside the issue of my concern.  The important thing to consider is the full continuum of unit strengths from weakest to strongest and deciding what part of the spectrum can hurt what.  If less trained, less equipped soldiers are trounced too easily, players are forced, in order to be effective, to adopt a linear strategy of Build the Best Units and Only the Best Units as they become available.  We have seen this the strategy of the day in many, many strategy games (as is the case in MoM).

By allowing even the weakest unit to have a chance (the key word is chance here, and that chance can be very small) it keeps other strategic options revolving around quantity, rather than quality, much broader.  Programming this into the combat scheme wouldn't be hard. 

Here's a hypothetical scenario based on a hypothetical combat system.  Let's say, for instance, that a mighty Groglock (completely made up monster) has 20 defense rating, and a poorly trained archer has 5.  The archer must exceed 20 defense to do any damage.  When an archer rolls, they have a 1 in 5 chance of getting a 5.  If they succeed, they roll again, adding the second roll to the first.  If they score another 5, they roll an additional time (and so forth.)  Now, the actual chance of that archer hitting the Groglock would be 1 in 625, and that only makes a dent.  An archer one shotting a Groglock might only happen once every 1,000 games.  You would need, naturally, many many archers to bring him down it's likely that, during most typical games, it would make more sense for the player to have used the resources needed to make an army of archers to instead train elite knights to bring down the Groglock, but my point here is that the strategic options in the game are broader and more diverse if you allow players who play their cards right to use these kinds of strategies effectively. 

To eliminate a meaningful strategy from the game on the grounds that "spears don't kill aircraft carriers in real life" (10 spears jammed into an engine turbine?) can only make the game smaller.  If you have any criticism of the idea presented above, please address it on the grounds of meaningful game mechanics and not ancedote.

In D&D Second edition (and all the others too) there are many types of creatures that can only be hit with magically enchanted weapons. As Frogboy already stated, without the ability to actually hurt something it won't matter how many people are fighting it, if they can't hit it they can't hit it. I like this. I like this a Lot.

This is really going to force the player to think about what kinds of troops they are sending into battle.

Thus begins the slide toward the Monolithic Army.  This mechanic limits player thinking.  If you are always encouraged to build the biggest and baddest units, why even bother with the small ones?

Once again, using ancedotes to argue against legitimate game mechanics doesn't produce good game mechanics.  Whether peasants could concievably kill God or spears can destroy an aircraft carrier is irrelevant when you are making a balanced game. 

Killing all of the Gods worshipers. Or one of the peasants have his (extremelly) lucky day (blessed be Avandra) and get Godsbane from Cyric and manage to stab the God a la Tasslehoff.

Btw, the Fire Mages could consume the whole oxygen around the dragon to suffocate him (he still needs to breathe, right?) as long as they managed somehow to limit his mobility (traps, terrain...) and keep the flames long enough (being 10k, that's the easy part compared to the previously mentioned one).

First off, I explicitly stated that my battle system post was hypothetical to illustrate a contest between a lowbie unit and a high end unit.  I was positing a possible combat system, but made a crude version to illustrate combat outcome.  You are absolutly correct that 5000 archers might match up to a Groglock, and that's precisely what I want to see for game balance reasons.  I want to see 5000 archers be able to make a dent in a Groglock and, given enough archers, even kill it.  It's not unknown that lowly defenders have killed great beasts in fantasy--- as a matter of fact, it's a staple of fantasy (think of the Hobbit and the defenders of Esgaroth).  We can come up with ancedotes all day long about grasshoppers, aircraft carriers, and counters to those ancedotes.  Personally, I don't see why the awe of a terrible and massive creature is reduced just because 5,000 archers can pitifully witle away at it (while the beast vanquishes scores of them with its breath weapon and rakes them into the air with its claws).  But as I mentioned before, this is beside the point of game balance.  I want someone to respond to this specific question:

"How does allowing low end units the chance of damaging a high end unit damage the balance of the game?"

I still haven't recieved an answer to this question beyond a response that distills down to, "Because it's uncool."

Now, this question should be stipulated by the fact that high end units should still grant vast and priceless strategic rewards.  In otherwords, although this is a simplification, X resources worth of dragon added to a large army should be a vastly better than a purchase of archers with that same X resources.  However, if I'm playing a game and my strategy has yielded a lot of extra citizens and a lot of extra food to support them, I want there to be a legitimate strategy to use them in a productive way without that strategy being dwarfed by a low hit cap ceiling, even in the late game.

If you want to make a balanced game, you don't make it based on what does and doesn't offend your sensibilities about dragons and Groglocks.  Like I've said before, trivializing masses of lowbie soldiers by neutering them against high end enemies reduces strategic options and forces other options.  If someone has an army of oober units and an opponent has spent the same number of resources on any army of lowbies, you can expect that army of lowbies to essentially never stand and fight.  Allow the lowbies to do some damage (abliet, still not be as effective as their opposing army) and retreat becomes an option, rather than a mandate.  And all other arguments aside, which battle would you sit up in your chair more?  A battle with an impervious high end unit where you might as well click autoresolve?  Or a battle you might actually have to manage?

all he is saying is that defense will have a 10% minimum (maybe attack too?) ... so that if a unit is out-classed in stats on the scale of 10:1, then scratches will be impossible. I think, maybe a critical hit could do some effectiveness (two max rolls in a row, or something) as in 1 damage, but no more than that. And thats talking critical hits.

One thing to keep in mind though ... is that if its truly an ancient dragon, elephant/rhino hide will be the LEAST of your worries, think closer to 10x that, plus a general magical aura.

Now, those types of units are EXTREMELY RARE.

If I were to, say, make a dragon-heavy mod with lots of lesser drakes, wyverns, more tribes of lizardpeople, ect, ONLY the most rare ancients will have such an advantage, and others will be able to be easily defeated by large enough numbers ... but will of course have the power of the lizard on their side

Yeah, that's totally cool.  I guess I'm just putting emphasis on chance-to-hit-ceilings.  I have no problem with, say, the top 3 units in the game being invulnerable to the bottom 3 weakness.  The key though is that you don't have too many units that have impunity, otherwise you end up with strategic loopholes, abundant blindspots, and predictable, monolithic armies of powerful units.  After all, if there are too many units that can take no damage from low to low-mid range units, no one will ever develop strategies around building the weak units and there will BE no lackeys for your mightier units to wade through during an Epic battle (unless the AI is unintelligent or seeks to please the player by building them anyway).   

For one, Smaug wasn't defeated by a horde of lowly defenders, he was defeated by Bard, a singular archer who was literally told of a missing scale in Smaug's armor by a little birdy. Sure, in this situation a horde of enough archers could get lucky and land an arrow in said weak spot, but maybe dragons in Elemental don't have such imperfections. And I can guarantee you that Glaurung, or Ancalagon the Black, for example, would never have been destroyed by even never-ended hordes of crappy little archers.

And I have nothing against 5000 archers being able to make a dent in a Groglock, so long as said archers are powerful enough to actually scatch it in the first place. But under no circumstances should any one of those archers ever do more than a menial amount of damage do said Groglock.

Also, I think Brad's method of determining whether or not a creature can be hurt by another is pretty good. If your Groglock has a defense score of 20, then he will roll between a 2 and a 20 (but never a 1). Therefore, any creature with an attack rating of at least a 2 will have a chance to hit. A creature with 2 ATK then would have a ~1/40 chance of hitting (could vary depending on how rolls are weighted etc) and doing minor damage. A unit with 1 ATK would have no effect. The result is that even if you have a unit with crazy high defense - to take Brad's example again, say 50 DEF for a dragon... It wouldn't take another uber powerful unit to be able to hurt it. It would just take decent units. Your crappy conscripted archers with 4 ATK will have no effect, but your slightly better-trained archers equipped with bows with actual tension, with 6 ATK would. And this mechanic probably wouldn't even be evident at all between mundane troops - the weakest regular troops would still probably be able to swamp the strongest regular troops.

Ever seen a Spearman defeat a Tank in Civ? That is your answer. Completely obsolete, underpowered units should not  be able to defeat overwhelmingly superior units due to the RNG. A game that based on randomness is not fun, it's random. You start a fight and pray something unpredictably bizzare doesn't happen to screw you over.

A system like you propose carries the same RNG risk. Sheer randomness is bad for a serious strategy game, because fights should be determined by tactics and army compositions and not by the RNG.

yea, RNG shouldn't be an all-powerful god, enabling that one peasant to kill that one Sovereign or Dragon during that world-wide tournament game. Or something.

Oi, once again, I don't think we're properly understanding what I mean by randomness (I'm sorry if I sound rude, but I'm a biologist and have to confront people's statements about biotechnology on a regular basis based upon gut feelings and prejudices rather than technical discourse, so my patience is already a bit exhausted.)  I'm not suggesting that there should be randomness in the game that swings so violently that nothing is predictable.  Not even close.  As a matter of fact, let's remove the word "random" entirely since it seems to be a loaded pejorative that sets a lot of people off.  Let's instead call it "range of probability."  If you are aware that your soldiers have an X chance of hitting a Groglock, and that 1 out of those thousand will end up hitting it per volley, there is really very little randomness at all, but instead an analysis of the range of probability with which you are acting.  I am in no stretch of the imagination suggesting that every other battle, your mighty creatures will be insta killed by the lucky shot of a peasant.  

I am proposing an exponentially diminishing, but potentially infinite ceiling for hit probability which does leave the possibility open that, at some point in the billions of games of elemental that will be played that a peasant will manage to drop a mighty end game monster in one shot.  But seriously guys. one end game monster being single handedly slain in the history of elemental isn't game breaking.  Not even close.  The true effect that you'll see a billion times more frequently is that units will always have utility, even if it rapidly diminishes in the face of stronger opponents.  This will reduce the supremacy of the classic mad-rush-for-endgame-units strategy that can make games of this type bland much quicker and it will produce a plethora of other viable strategies with which to pursue in the game.  It remedies way, way more problems than it causes (even though it would seem to offend plenty of wishes for oober doober monsters, despite having almost exclusively a psychology rather than literal effect on their performance or strategic value).

And I'm still waiting for an answer to my bolded question beyond authoratative, but unsubstantiated, declarations of "it just doesn't seem right to me".