Watchign part of Final Fantasy: Advent Children (specifically the Tifa vs Loz fight) I realized that if I ever get off my lazy 60 hour work week ass to write my own game up that one thing I desperately would love to have is a truely dynamic combat system.  I have seen the current 'role models' of Feng Shui and Exalted and I realize this of all things: They just aren't that cool in play. You have a limited body of techniques, and everyone will have that one key ability...

I have seen the alternative: narrated combat based on plot effects and conflict resolution and I realize these are hollow mockeries of what I want. There is no risk, no sense of accomplishment because you aren't playing out the fight, the give and take of combat. The awesome of a slow motion dodge, the feeling the walls shake from the force of impact.

I have only some few nebulous ideas to start with.  I hope to refine them.

First of all a great deal of the Onus of awesome dynamic combat is in the hands of the GM. No amount of simple advice can make awesome descriptions come naturally. No amount of rules forcing the awesome will accomplish anything.   All we can do is provide guidance to coax the awesome out.   I'm thinking minor rules, a dodged punch against a wall will have to damage the wall... something like that.  Every action should prompt a reaction, and in teh course of that build a narrative.  A blocked powerful blow should drive the defender back... thus even if the GM can't describe shit, the actions and reactions at least provide a framework that anyone can hang their own descriptions off of.

Allowing for maps gives us an opportunity to accomplish a lot. By giving out scads of mobility, a map can be used to generate plenty of dynamics in it's own right.  If your character doesn a flying over the head backflip kick to his opponent, encourage him to move his model in a flip over the enemy model and place him on the other side.  On the other hand, maps should remain optional as well.

Resource managament by a single pool of hit points is duller than watching paint dry.  On the other hand, resource management can very much give a sense of tension.  Warriors build up bodies of techniques, repeating a technique too often fatigues it, makes it weaker, less effective.  Thus each teachnique, each battle skill becomes its own resource to manage.  Do you risk dodging that attack for the third time, or take the powerful hit on your weaker block, knowing you'll take damage anyway, but at least mitigating it?

The most important thing is that combat has to move fast.  No big pools, no spending hours measuring distances or computing attacks. You say, you do. You stand their and quibble, you wind up playing reactively. Initiative, if it exists at all is rolled one time, at the start of the fight. Why? Because there are no rounds.  Fluid combat doesn't move on rounds. Faster fighters swing more often, not round by round, but cumulatively over the fight.  

More to follow.

One aspect I think that can be of interest is removing the feeling of being a slave to the dice.  There are two aspects to this: One remove the excessive randomization.  The second is using the idea that under almost all circumstances the heroes and villians will always hit their opponent. The only way to avoid this certainly is to defend, to oppose.  Bell curve single die types or small fixed pools with simple targets. A means of comparing relative successes is a must, as combat interactions should rarely be hit/miss only. That's not DYNAMIC AWESOME.

Exception: Very rarely you might have a circumstance where there is essentially an unopposed but nigh on impossible 'attack'... hitting a man with a bow at half a mile, at night, in a blizzard, striking a fly between the eyes between drops of rain without getting wet or some whacky shit like that. Obviously, these are not directly opposed by the enemy so much (well, flys have a pretty damn good dodge...)  Obviously some sort of mechanic should be in place for that, but otherwise, all combat type rolls would be opposed, with the default being a successful attack.  

There are an infinite number of techniques to be learned. Seriously. Aside from the staples of all combat (dodge, block, punch, kick) players can, with GM oversight invent new techniques, the GM can likewise determine new techniques as well.  This is sort of rough territory here.  THink the master swordsman, when confronted by an enemy he cannot beat, turns to his master... who in turns reveals that there is yet one last technique he has not taught.  Yet, this same swordsman can later find hidden scrolls, or invent his own school, his own techniques.  Each technique must be mastered and tracked in turn, the more you have the harder it is to master them all, but the harder it is to predict your actions, or to wear out your techniques. A fight between a master and a jack of all trades could go either way...

Semi-scripting combat.  Scripted combat is ass. It invariably breaks down to a very limited number of useable techniques  adn options, the very opposite of dynamic awesome. On the other hand, guessing an opponents next move is a staple of dynamic combats, in fact an entire genre of 'battle in teh mind prior to the battle in teh real world' sprang up in japan out of that sort of shit. Obviously we don't want repetetive techniqueing, that shit is predictably dull. If, at any time the opponent (on either side) wishes, they may write down their attack vs a specific defense (or vice versa) for the next exchange.  In fact, they can even write down several prescripted if this than that, and may 'use' that prescripted (only one at a time) whenever they like. If they guess right, they get a bonus, if they guess wrong they get a penalty.  It may be that they can even determine their own risk.  If the opponent, however, guesses their guess then the risk is doubled....

I had some weird concept of making stunting the default mechanism, that is the only way to actually do anything, but I AM trying to avoid too much "GM's call" situation, and I still have mixed feelings about stunting.

I know I'm digging up the dead here but what Spike has written here is very intriguing to me. I've always felt that it took a great deal of work on the part of the GM and players to create vibrant, exciting fight scenes. I would usually try to add a bunch of environmental and situational effects to a combat scenario. Say, fighting in an elevator, where stray punches and kicks hit buttons or certain moves can't be done. (Think of the struggle in Kill Bill Vol. 2 in the trailer, when a sword couldn't be unsheathed because of the tight space.)

But I started giving the idea of dynamic combat some deep thought. And for some reason, the Duel mechances from L5R jumped into my head. Now, I never played that game. And when I read through the rules, I had a hard time understanding how duels worked. But essentially (I think) the two sides would gamble raises (adding to the difficulty) until someone relented. Then the last bidder started rolling against the difficulty. If he succeeded, he rolled for damage and assuming his opponent survived, regular combat would commence. If he failed, then his opponent would then strike at the difficulty that he had last raised. Same thing, until they move backwards through the bidding. At some point, with enough misses, regular combat begins.

I explain that because I think a similar mechanic can be used to create a sense of dynamic combat.

I'll use D20 as a basis for the explanation of the mechanic but I think it could work in GURPS or even WOD. And it could be the basis of a combat system, but I dunno how that might work. I'll explain the ful of my idea in the following post.

In D20, you can limit the "duel" to similar combatants. Say, swashbuckling agile fighters of similar skill or level. Or evenly match gladiators. I'll use martial artists for my example.

The two combatants roll for initiative. Bob wins and will go first. To start combat, he simply explains his attack.

Bob: "I throw a quick jab to the head."

Now, Frank has to roll to defend himself against the attack. It's assumed the attack will land unless Frank does something to stop it from hitting him in the face. So, he rolls a D20 against an average number: 10. He rolls an 11.

Frank: "I slip the jab by sidestepping and counter with a wicked right-hand hook."

Now, it's Bob's turn to roll. He has to do something or the hook lands. But instead of rolling against 10, the furiousness of the combat (still rather low here) has increased the difficulty, he has to roll against 11. He rolls a 15.

Bob: "I duck the hook and twist for a leg sweep."

Again, Frank has to do something to avoid being taken down by the leg attack. But, once more, the difficulty has raised. He has to roll against a 12. He manages to roll just that, a 12.

Frank: "I backflip away from the legsweep but maintain my momentum for a dropkick."

And now Bob has to avoid two booted feet headed for his chest. The difficulty raises one, once again. He has to roll a 13. Instead, he rolls a 7. So the attack lands successfully and Frank will simply roll damage as normal.

You can also add other elements to the fight. Using a freeform style addition that gives a bonus to the roll for particularly inventive counters or whatnot. Or perhaps characters can purchase manuvers that create penalties for defending. But those might slow the combat down. The idea is that the two combatants would be rolling fast and furious, thinking quickly on how they react to attacks, defend them and then counter. So, being able to pop off moves verbally and then get the dice moving should be rewarded.

A ceiling could be put in place, too. Say, once the difficulty threshold reaches a point that neither combatant would be able to defend (with no modifiers, that would be 20) you get the cinematic concept of the two backing away to reappraise their foe before closing to resume fighting. At the point that damage has been rolled or you reach the ceiling, the next round of combat begins, so the other players would then begin their combat resolutions. Feats could be taken that allow characters to be better at dueling. Say, a "Combination" feat where one character has the ability to (once he has succeeded in an attack) add another attack right after that. A jab that's followed by a haymaker, for example, would require the attackee to roll to defend twice (after failing, he would roll against a higher difficulty to defend the follow-up attack) or take double doses of damage.

Another possible addition could be the idea of "limit breaks." These would have a mechanic of adding the number of ripostes, bobs and waves and such to add to the damage once someone does connect on a blow. Say, if you do start a ten, you add how far you are from the starting number or divde by two, whatever seems most balance. So, if Bob connects with an uppercut on Frank at the difficulty of 18, he would add 8 to his damage roll.

Have you looked at ORE?  As implemented in Reign, it gives some great tactical possibilities without dragging down in grid positioning details or handing out excessive narrative authority to "invent" the results.  I think with some nudging it might be able to deliver what you want.

I never consider one of my threads well and truely dead until the Forum its on dies an ignoble death.

That said, I've been trying to implement my ideas in another thread, mixing this stuff up with some other ideas I had to.  

If it was easy, some other dude would be selling this stuff and I could focus more on my real job...