Okay, mechanics idea. This governs entirely physical damage, I won't be talking about magic here except tangentially. I'm going to have math in here, but also describe how the formulas will affect things qualitatively and why I am putting things forward they way I am. Also I am going to wall of text here. The important stuff is in bold.
I mentioned already that weapons should specialise in overcoming defence and/or having a high multiplier (via swings). Here's how I see it working:
First of all, we create a stat called Agility. (it can have another name, or could even overlap with another stat; the concept is what matters) Agility represents two things: your ability to swing multiple times, and your ability to avoid being swung at multiple times. Basically how fast you are in close combat (rather than how fast you get turns). In theory this can be split up into offensive agility and defensive agility but I consider that unnecessary because we can (and will!) modify offensive agility with weapons.
When you attack an enemy, your agility is compared to theirs. If your agility is less than the enemy's you swing once. If it's greater than or equal, twice. From there on, you can additional swings if you beat the enemy's agility badly:
3 swings if your Agl > 125% of target Agl
4 swings if your Agl > 150% of target Agl
5 swings if your Agl > 175% of target Agl
And so on, possibly until some cap to prevent 100 swings if your agile fighter faces an earlygame slime or something.
Now, to prevent 1 swing (i.e. the only attack a low Agl fighter is going to get, most likely) from being too weak, all the swings except the first do half damage compared to the first. It makes sense, those added blows you land don't hit with the same force as the opener.
Weapons of course modify attacking agility. They work as multipliers, which exaggerates the existing Agl differences of PCs. Because of this, low Agl PCs will usually do better with low-swing weapons, while high Agl PCs will usually do better with high-swing weapons, which makes sense - the slow, lumbering tank is going to get better money out of a warhammer while the speed demon will get more out of knives, and so on.
To illustrate this, here's an example of how this will work. You have three PCs, with Agl stats as follows: Noemi with 29 Agl, Mirek with 21 Agl, and Isolde with 15 Agl (example may not reflect the real stats of these PCs, obviously). And you have three weapon choices:
-A warhammer that does 180 damage with one swing, and an Agl multiplier of 70%
-A sword that does 150 damage with one swing, and an Agl multiplier of 100%
-A dagger that does 100 damage with one swing, and an Agl multiplier of 140%
We're facing off against a goomba with 20 Agility. Using the mechanics as I've set forward...
Isolde's options:
-Warhammer for 15x0.7 = 10 Agl, swings once for 180
-Sword for 15x1 = 15 Agl, swings once for 150
-Dagger for 15x1.4 = 21 Agl, which beats the Goomba's and swings twice for 100 + 50 = 150
Warhammer is the best option.
Mirek's options:
-Warhammer for 21x0.7 = 14 Agl, swings once for 180
-Sword for 21x1 = 21 Agl, swings twice for 150 + 75 = 225
-Dagger for 21x1.4 = 29, swings three times (beats Goomba Agl x 1.25) for 100 + 50 + 50 = 200
Sword is the best option.
Noemi's options:
-Warhammer for 29x0.7 = 20, swings twice for 180 + 90 = 270
-Sword for 29x1, swings three times for 150 + 75 + 75 = 300
-Dagger for 29x1.4 = 40 Agl, swings six times (beats Goomba Agl x 2) for 100 + 50 + 50 + 50 + 50 + 50 = 350
Dagger is the best option.
Hopefully this illustrates the system at work. Of course, to avoid Noemi destroying Isolde for damage, we give the PCs different strength stats, which will multiply damage dealt. If Isolde has double the Strength multiplier of Noemi she'll do more damage, for instance. More on that later.
Noowww, if this were all to the system, then each PC would just choose the obvious weapon after doing the math above and we'd be done. This would be lame. The point of the weapons isn't just who uses them well (especially since each PC will only have limited options, though hopefully more than one!), but it's how well each weapon, and hence each fighting style, works against different enemies.
-High power weapons do well against high Def. Knicking a heavilly armoured enemy repeatedly with a dagger probably won't do much, but the hammer won't be slowed down as much.
-Weapons which only swing once aren't bothered by enemies with high Agl. Oh no, my hammer goes from swinging once to swinging once. Meanwhile, the 6 swings Noemi got against a goomba? That's dropping to 3 if she faces a SUPER GOOMBA with 30 Agl, which takes out just under half her damage.
-Okay, both of the above favour the heavy weapons. Does this mean the heavy weapons are always better against stronger enemies? No. Think about it: would you want to hit a dodgy enemy with a hammer? Probably not. Now, a knife fighter? She might have a better shot. So, generally speaking, high power, low-swing weapons will tend to have lower accuracy. Accuracy won't matter much against an unevasive target, but it sure will against the evasive. I don't have an accuracy formula for this but FFX is an extreme example of this at work.
-The exception to the above rule will be ranged weapons, which will likely tend to be high-power, low-swing, but high-accuracy. That crossbow is only hitting once, ever (Agl multiplier of 0!) but hits things fine.
So, back to illustrations, since I love them. Consider the following enemy types, and observe which weapons might work best.
-An average soldier, like the goomba above. Averagish on all ways defensive. Each character wants the type of weapon which fits her stats best. We already discussed this situation.
-A heavilly armoured, slow-moving tank in plate mail. Bad evade, bad Agl, awesome Def. Daggers are practically tinking (even if they hit 8 times), while the hammer? Sure, it's taking a damage reduction from the armour Def, but not nearly as much as one... and it may even swing twice, unlike normal, which means it isn't taking a damage cut at all most likely. Characters will want to switch to high-power weapons to do well here.
-A balanced defensive fighter, like your stereotypical paladin with chain mail and a shield. The shield gives him solid evade (hoses hammers), and he has solid Def (hoses daggers). If neither stat is extreme, the sword may sneak up the middle and be the best. Or we're falling back on magic here.
-A combat mage armed with a buckler, with poor Def, good evade, and mediocre Agl. Dagger wins, should be obvious by now why. Go go light, accurate weapons.
-A lightly-armoured, lightning quick whirling dervish. Lots of Agl, lots of evade, bad Def. What weapon works well here? Not the hammer, good luck hitting. Not the knife... it'll hit, but with that Agl, probably not more than once or twice, which destroys its damage. Solution? Indiana Jones had the right idea. Just shoot the bastard. High accuracy, isn't losing any swings, leading to nice, reliable damage. He can't dodge or dance out of the way of that.
So, wrapping this up, damage formula time. Now, I already mentioned Strength is going to be a multiplier. I haven't yet justified why, so here goes: in order to not have a completely overpowered attaker, Strength and Agility will generally need to not both be too high on the same PC... in fact, for attackers of equal competence, the two need to be in opposition. Thus, since high Agl characters like knives, and high Agl characters tend to have lower Strength, it follows that knife users tend to have lower Strength. However, if we ADD the weapon power of knives to said Strength, then it's going to be pretty bad at overcoming defence compared to the same knife in the hands of someone with high Strength, which will lead to the high Strength character being better with knives against higher Def enemies. This is bad. To avoid this situation which I find wholly unintuitive, we make Strength a multiplier. Strength doesn't help you break defence, it just helps you do more damage. Even someone muscular isn't going to be damaging a giant armoured crab much with a slim knife. Now, if it were a masterwork, resonance-forged knife, then maybe... but that's represented via higher weapon power, again emphasising that weapon power overcomes defence. Thus!
Damage formula = Strength * (Weapon Power*A - Defence*B) * (Number of Swings + 1) / 2 * C
We can call Strength something else if we like, same with Agility. A, B, and C are arbitrary constants to make the formula work. They ALSO may represent buffs, skills, and other effects. Stuff like elemental defence probably goes where C is in the formula, A might be something like a proficiency bonus with a certain weapon, etc.
Passive skills can have a number of effects, here. Some examples off the top of my head:
-Quick shot: Allows a character a second hit with a bow.
-Defensive mobility: Provides a 25% bonus to Agility when attempting to avoid enemy attacks
-Freak strength: 25% bonus to weapon power (allowing the character to better overcome Def because he hits just that hard... maybe he can take down a giant enemy crab with a knife)
-Stunning blow: The shock of your attack multiplies the target's Def by 25%, making its armour nearly useless
Active skills, on the other hand? May increase the multiplier of your damage (C) or they may help you overcome Def better (A). They may or may not change your number of swings (down to 1, or a boost, or anything in between). Depends on the skill. We really have a lot of flexibility for different attack skills with this system... and that's without bringing magic and half-magic attacks into things!
You get the picture. Anyway, this is what I'm proposing.
Finally, criticals. This is somewhat unrelated, so it gets to be a footnote, but I thought of it as well. Criticals should overwhelm (ignore?) defence against human targets since their armour has weak points, but possibly not creatures lack weaknesses (golems, possibly, maybe some rare monsters). Suikoden 3 basically did this, but it is a good idea and we should rip it off. Either way they should probably have a damage boost to final damage of 1.5 or something (boosted by some skills).
