Yeah, 1-N is freaking awful for single dice rolls.  It wouldn't be quite so bad if calculations were more like 1-N for each of the twelve guys in a squad, but even that is pretty bad when you consider that a single bad defense roll can kill your sovereign and end your game right there.

A lot of people are asking for special abilities.  I don't think the game needs them.  I think the game needs better balance with weapons, armor and unit training and a smarter AI, and it'll be fine.

Fortunately or unfortunately for me, depending on how you look at it, I purchased the LE months ago to get in on the beta (and then wasn't able to participate much at all past a few crash reports in 1Z.. I know, it's my fault the game shipped like it did), so I'll just comment on your complaints:

Sad but true, this is completely right.  It's not quite the Civ1 spearmen vs. battleship issue, but it's damn close.  By turn 50 or so, or as soon as more than 1 AI has researched the basic adventure techs and unleashed level 3+ monsters on the world, Sovereigns stop being even moderately badass and are relegated to being that guy who summons the fire giants (early game) and teleports the army around (late game).  It's exacerbated by the horrible healing and mana regen mechanics: anything with enough HP to be useful and survive those random damage spikes is also going to take dozens if not hundreds of tuns to heal back up because no matter what you only regain 1hp in friendly territory and 5hp in a city (including large squads.. apparently they only get one bed per squad, so can only heal one guy at a time).  And since mana regen never goes higher than 1/turn without a high powered building, forget using magical healing to get around it, since you're just trading in HP for mana.

I will say that if it weren't for all the massive issues affecting the military, this might not be so bad.  I agree that a variety of troop abilities is absolutely necessary, but I think if the combat system weren't fundamentally flawed I could at least have fun with it 'til they fixed the issues.  You can manage some decent specialization now (ex. I have a 'scout' unit that is only ever built singlet, never in squads, because they have every sight-bonus granting item I can give them, and I build them to accompany my sovereigns and independent armies), but real specialist units would be nice.

Actually, according to at least one post I've read by devs, squads of troops do actually roll an attack for each trooper vs. the target.  It's not terribly noticable amidst the general randomness of the combat syste, though, so it's hard to say how accurate that is (yes, a dev said it, but given the amount of confusion in the XML files and the comments of "I think Bob checked that in, but I'm not sure", that's no guarantee).  It does seem to jibe roughly with what I've seen.. large squads of troops with basic weapons seem much less effective than their ATK stat would suggest.  Unfortunately its undeniable that squads are treated as one big ball of flesh when it comes to healing (see above, one bed per squad) which makes them insanely frustrating to use in the later game, or even after gaining a few levels.

Again, something that I'd actually mark as a bonus.  But again, that's coming from the perspective of someone who's already bought the game and is struggling to justify the purchase to themselves.  This wouldn't be quite as important if it weren't for #5 below, but all spells being X mana for Y damage really makes this lack stand out.

Personally I'd put this at #2, if not #1.  It's the most worrying because unlike the combat system, the fix goes beyond just tweaking some calculations.  Rewriting the spell list, redoing a lot of spell effects (and fixing broken ones) and adding unique secondary effects is just the first step, and that's likely to take a while.  Ideally we'd see the ability to purchase different levels of each spellbook to give a mix, and perhaps a way to make your Sovereign a more powerful magic user other than going out and squashing spiders and ambushing poor bandits on the road.

Hi,

I too consider the 1-N combat damage system a horrible design decision, and it is one of the things that's holding me back from buying the game right now. It gives more weight to luck and weakens tactical and strategical playing.

Balance is also a big concern to me. I like games where there is no optimal way to play, no "best path", but where instead you have multiple options of how to try to win, and have to be flexible and adapt to the current situation.

Another big factor why I'm not buying the game yet is the weak AI. I don't care for multiplayer, and would like to face a challenge when playing single player *without the AI having resort to cheating too much*. The later Civ 4 AI (Blake's betterAI) was fine in that regard, and has raised the bar for what I expect from a 4X game. An ideal AI would have to play under the same rules as a human and still surprise me and pose a challenge. I know that this is not possible to achieve, so smaller boni are okay. Big boni like no fog of war are not, though.

-Kylearan

Agree with OP's points, totally 100% agree in the case of #1. That's the single biggest problem with the game right now. The system is so wildly unpredictable that any reasonable attempt at strategy or balance just doesn't work. Make a bunch of guys with maces and you destroy everything. And if they die? Just make more. Champions are much harder to build up, more fragile, and far harder to replace.

What I'd add to the list is multiplayer. Listed on the box feature, still disabled, and now we've learned it won't be enabled this week either.

Great summary of the most important issues. 100% agreement, especially with #1.

According to Frogboy, the squad members already roll damage separately, even though this isn't shown in the UI. However, it seems that defense is just a sum of all members of the defending unit, which is bad. Combined with #1 I would call it crippling.

Bugs, late game crawl and crashes, then balancing units out.

Now I have not had as much problems with the Champions in the game. what I do is Hire a lot of them and as I get them My Soverign empowers them with magic. So basically my Soverign is there to make spell casters. and these spell casters build up mana and become power houses when with an army.

It is a very effective stratagy. However the Champions are quite usless if they are just melee units without spellcasting.

Yes, yes, and yes.

The combat system is my #1 as well, and not just the system itself, but also the tactical combat engine.  Fundamental changes need to be made to make tactical combat fun, rather than an excercise in making sure the auto-resolve doesn't kill your weak guys that are tagging along for XP. 

First, I find the "tactical" combat almost devoid of tactics.  Magic users cast, archers fire, melee guys move up and swing.  Positioning and unit formation mean nothing.  Combined arms and army composition mean very little.  Morale has little to no effect. 

My suggestion: the tactical combat engine, at a minimum, should include the concepts of engagement and flanking.  Units in melee with each other are engaged - breaking that engagement gives the opponent a free shot.  For flanking, if a unit is engaged, and attacked by a second unit (not the one it is engaged with) it suffers extra damage, or is easier to hit.  Archers engaged with a melee unit cannot shoot anyone else.  Similar to these concepts is the idea of support - that a unit with a friendly unit next to it is more effective.  Adding these things in a relatively simple way would make the tactical combat much, much more enjoyable.

Of course, one of the reasons combat is so simple is probably so that the AI can handle it.  This will require a major AI upgrade, and I'd encourage Stardock to think about changing the Auto-Resolve in the process.  I suggest that Auto-Resolve ought to be based on something other than the AI playing both sides really fast in the tactical combat engine.  I'd recommend something based on a combat rating ratio (assuming they modify the way Combat Ratings are calculated so it really does reflect the combat power of the stack), with some random chance of attrition.  Make it roughly equivalent, but not the same.  The reason is that I don't want to take the time to play out every battle, but if I can see that the AI is doing stupid stuff with my troops, that's going to be frustrating.  I don't want to have to choose between being bored or frustrated.  If you can make an AI that does everything exactly the way I would have, great! - except that now there's no reason for me to do tactical combat at all.  So I just don't see a good solution if the AI is just playing out tactical combat for me. 

I'm not impressed by the random-terrain-gives-bonuses-and-penalties "feature", which requires you to mouseover the whole battlefield in order to find the important squares (if there are any) on the battlefield.  Add in that the terrain modifier isn't really that significant, and it's a layer of the game that could be interesting, but isn't.  I like the idea that terrain influences the battle, but I'd rather see just a few squares in each battlefield that give a significant bonus.  These squares should be easy to identify, maybe by a graphic, or maybe a toggled overlay.  It would be even better if the location of the squares was dictated by the terrain in global map.  So if I attack someone on a hill on the big map, they get a few "high ground" squares on their side of the battlefield map.

I think the Morale system has promise, but it needs to be made a lot more significant in the game.  Morale should definitely affect the effectiveness of groups, and should have more levels than just "rout if it's low enough" - losing movement points and/or moving randomly are both possibilities.  This, of course, becomes a lot more significant if positioning actually matters.

Lastly, from an eye candy perspective, I'd really encourage Stardock to look at the scale of the figures on the battlefield.  It really looks terrible when a single figure is defending one corner of his square, and then has to run to another corner to defend against a second attacker.  Figures should be large enough that they don't need to move within the square, but can attack and defend from the center of it.  Of course, the consequence here is that larger squads and monsters (i.e., dragons) will have to be larger than a single square.  I recognize this is a lot of extra work on the core code, but I think it would be worth it in the long run.

If the tactical combat engine is interesting, then the units can be fairly simple, and you still have interesting combat.  Even so, I think a single Attack and Defense rating is too simple (and, as the OP said, the 1-N model is just terrible).  I'd really like to see the "to hit" and damage rolls separated.  I think the minimum here is to have Attack and Defense used to calculate probability to hit, and then have a separate "Damage" stat, which could be modified by the Att/Def ratio.  I think I'd prefer to add an Armor stat as well, that mitigates damage.  Then Att/Def are used to calculate whether the strike hit, and Damage/Armor are used to calculate damage to the unit.  I think 4 stats would really start to give the units some differentiation and personality.  Implicit in all of this is that HPs are balanced better with the combat model, so that trade-offs in the stats mean something - if the combat model strongly favors any single strategy (glass cannons currently), then it isn't much fun.  The model should implicitly lead to a system of expensive generalists and cheaper specialists who are strong in specific situations, but weak in others.

Aside from the combat system, I'd probably put a meaningful magic system as a close second priority for me.  Special unit abilities become less important if the tactical combat itself is interesting.  I'd still like to see special abilities, but they're not the centerpiece of combat - the combat itself should be the centerpiece of combat.  The OP mentioned balance as well, and that's pretty much a no-brainer.  But, really, I think Stardock hasn't even finished the engine yet (or at least, I hope they haven't), so let's get that done first because it's hard to balance the units when the combat system itself is in flux.  Meaningful damage types I could do without, actually.  I think the whole slashing/bashing/piercing thing is probably more complexity than the system really needs, and doesn't necessarily offer that much.  Yes, I would like to be able to build a counter unit to a specific enemy unit that's hurting me, but I think a good combat system would offer that without multiple damage types.  Magic resistances, OTOH, would be very nice to see.  That would help differentiate the magic types in the game, which would be a good thing, but the magic system needs a lot more help than just that to be really interesting.   

This is the most important thing to me (I call it "diversity").

Are you the same Dwiggs that created an AoWSM mod?

Yeah I don't disagree with this.  I didn't specifically state it, but 1-N being a problem because N gets so large is absolutely what I was getting at.  And I agree 100% that N is related to hitpoints.  A 1-10 damage system where units have 50 hp is generally going to be fine, because it will take on average 10 hits to kill which means that the overall combat result becomes gaussian in distribution (like rolling multiple dice.)    However when 1-N is used and N can be higher than average HP (like elemental) then this is just absurd.

The fixed defense is also very important.  Warhammer went with it and so did D&D.   These guys know their stuff and decided that one random roll per combat outcome is quite enough - adding a second for defense almost always introduces too much randomness.  Two very popular systems right there that did things the way they did for very good reasons. Elemental could learn from this.

Yes, I was the creator of DwigMod for SM. Eventually, The_stranger took over on his own and developed it even further. I must thank you for your wonderful unit recoloring tool, which I remember making extensive use of while creating new units. AoW:SM was a great TBS experience, and I'm hoping Elemental can head down the same road with some work.

Exactly this! Right now, if we define a "skirmish" as the exchange of attacks between two units until one dies, the skirmish's outcome tends to be determined by a relatively small number of dice rolls. I think (based on dev comments I've read, although I could be wrong) that this may be the byproduct of a conscious design effort to speed up tactical combat.

However, the problem with this approach is that it tends to make combat extremely random, especially as N (the set of possible discrete random results) increases. If N is large and a skirmish is decided by just a few dice rolls, battles become a toss-up (strongly in favor of whoever strikes first if units start getting one-shotted, which is horrible for the current AI and the tactical combat experience in general).

The more rolls that become involved, the smoother the distribution becomes, and the more predictable combat becomes. While "predictable" sometimes tends to carry negative connotations as a word, it is a good thing to a degree in strategy gaming. Having some level of predictability is what lets the system "make sense"; a player expects a 6 ATT / 6 DEF unit to be much sturdier and almost as strong as an 8 ATT / 0 DEF unit. The more rolls that a skirmish entails, the more predictable it becomes; precisely how smooth the curve should be is a design choice, but I'd argue that hovering on the "few rolls/very unpredictable" end of the spectrum is generally a very poor choice, as it trivializes so many decisions.

For comparison's sake, Master of Magic had a highly predictable system - there were a lot of dicerolls per skirmish. Each individual point of ATT had a 30% base chance to deal 1 damage; then, each individual point of DEF had a base 30% chance to reduce damage by 1. It was an elegant system with great room for mechanic expansion (+ to hit, armor piercing, etc.), and it was very good for predictability in that combat results felt like they "made sense". (Even with this many rolls per skirmish, battle results certainly weren't predetermined, as my Load Game button can attest...)

While reducing the ratio of damage output to HP might slightly increase the length of tactical combats, I think it's a far more important issue than pacing - at the current ratio, strategy becomes fairly trivial and results become uncomfortably random. This is what I mean when I complain that we need a solid redesign of the damage/combat system. (If tactical combat length is an issue, there are many other avenues to explore which aren't so detrimental to game balance - speeding up attack and movement animations, tweaking army positioning, reducing spell animation time or allowing the player to act through it, addressing kiting, and so on.)