Entertaining combat

[Erebus]

Combat is a very important part in the vast majority of CRPGs, but it's not always as satisfactory, far from it. What do you think makes combat entertaining or not ? In my opinion :

- The tactical aspect

I'm currently replaying Champions of Krynn, one of the Gold Box games and my first CRPG ever. It wasn't such a memorable game, but it had some nice things... that I was very disappointed not to find in my second CRPG (Bard's Tale 3). The most important one was the ability to move around each of your PCs as you wanted during battles. It really gives combat much more dimension ! But in jRPGs and games that keep first person perspective (Eye of the Beholder, Ishar, M&M...), movement and placement have little to no importance. And some games which allow your PCs to move around don't do a very good job because of poor pathfinding, cramped spaces, artificial barriers, etc.
The tactical aspect also includes the number of choices that are available to each PC (of course, said choices have to be significantly different from each other, which isn't always the case).
The number of PCs also plays an important role. A party of 5 or more definitely seems more entertaining to me. I found The Witcher entertaining enough, but there were times when having a single PC felt pretty frustrating. For similar reasons, I don't like having companions managed by the AI (Fallout, Arcanum...), and not just because the AI is usually pretty dumb.

- Is this battle necessary or in any way useful ?

Combat being nine tenth of many CRPGs, the above question rarely seems to cross the minds of the people who make them. In jRPGs and games like M&M, most areas are crammed with countless foes, whose only goal in existence seems to be your death (they very seldom seem to bother the rest of the world at all). This leads to very repetitive and absurd battles.
Even in CRPGs that don't go quite so far, the number of enemies you'll have to kill is often way too high. Bandits and wolves can't find an easier target than your well-armed party. Criminal guilds and evil cults have ridiculously high numbers of mooks to send against you. Countless evil humanoids have nothing better to do than loiter around places through which you'll have to pass. Etc. These unrealistically numerous and needless battles end up being chores.
Of course, there aren't many CRPGs in which each battle has a logical reason for existing. Mysteries of Westgate is the best example I can come up with. Torment and Fallout are good examples, but they have some unnecessary combat-heavy areas here and there.

- Interesting enemies

A developed personality can be a nice thing for an enemy to have, but it's not enough to make him interesting during combat. Being impressive is nice, but not sufficient. A truly interesting enemy, in my opinion, is one that keeps you on your toes. He's unique and challenging, uses surprising tactics, has several different ways of attacking you, etc.
In fact, I personally tend to prefer groups of enemies, as long as said enemies are diverse and complementary. The BG series has some greatly entertaining fights against evil adventuring parties and other mixed groups. The end fight of SoZ pits you against a dozen of challenging opponents with very different skills and powers. Fighting a group of enemies that are all dangerous but will act very differently requires you to think throughout the whole fight.


Of all the CRPGs I've played, I'd say BG2 and ToB did the best job of having entertaining combat. I remember many of the fights very fondly. Torment and MotB are two of my favorite CRPGs because of their excellent stories, atmospheres and characters, but it's a bit frustrating that they don't have more memorable battles.


I'm waiting for your thoughts !

 

[BethesdaLove]

*silent audience*
*somewhere in the back* YOU SUCK!
/thread

 

[Erebus]

Yeah, well, you can lead a camel to water but you can't make him drink.

 

[Castanova]

BG2 and ToB did the best job of having entertaining combat

Right on brutha, and I agree with most of your post. I think one bullet point you missed out on is "Interesting Character Building." Combat is your opportunity to try out the characters you've built and see how they work out against a challenge. That's one of the reasons I don't like pre-set companion characters that you have no control over in terms of development.

 

[Phelot]

Yeah, it was satisfying for me in BG finding the other parties and offing them for being douche bags. Normally, I would wonder why they even gave them dialogue, but it's much more satisfying to have them babble and then kill em all! Like that one party in the area were Drizzt is. It's a fighter, his mage girl friend, and a hobgoblin I believe. Man, I always love taking them out.

Anyways, on topic I agree it's fun fighting these little encounters much more then having to deal with hordes of generic enemies. Honestly, I wouldn't mind a game were that's ALL the combat you have to deal with.

 

[Mastermind]

What do you think makes combat entertaining or not ?

It depends.

I either want deep strategic options or decent hack & slash. In both cases, the more real options I have, the better. I think a lot of rpgs have unnecessarily dumbed down combat.

 

[Erebus]

Another thing is that I prefer battles that are winnable with a large variety of tactics. Opponents that are immune to most kinds of attacks or who require very specific tactics to be defeated are often annoying.

In BG2, for instance, I didn't like fighting Kangaxx much. Either you were protected against his special attack and the fight was disappointingly easy or you weren't and it was impossible.

 

[Kaucukovnik]

I like an idea of multiple conditions that affect combat abilities. The point of combat wouldn't be to reduce your opponent's health to zero, but get into a position so that you can deliver one deciding strike.

You would also need to master different areas of combat. Taking down armored knights with a dagger on regular basis is not very likely, same as hitting swift opponents with huge, heavy weapon.

An often overlooked aspect is confidence and evaluation of the opponent. You could pretend to be weak and then strike with unexpected skill and force. Or the other way around, making your opponents insecure and nervous. More experienced foes would be less likely to be fooled.
Your reputation should matter a lot as well. Nobody yields to a murderous maniac who is expected to kill him anyways, and many won't even start a fight with you and rather run away.

Chasing enemies who flee shouldn't be a chore. You should be rather relieved that no battle took place at all. Most game mods that make enemies flee end up being not very popular, because players don't want to miss the loot and/or experience. This needs to be offset, probably by the need to conserve your strength in order to remain efficient until the next opportunity to rest and recover.

Also, being outnumbered should be pretty much fatal in most cases. Giving more importance to positioning and movement, clever use of choke points like doors...



Regarding modern/futuristic settings and firearms, Jagged Alliance 2 got it right for the most part (not that I have any real experience to compare). And the 1.13 mod keeps polishing the game even further.

 

[DraQ]

- The tactical aspect

I would broaden that so as to include all kinds of options and ways to influence combat. Having sufficient wiggle-room for the search for necessary tactics to be non-trivial is one thing, but any sort of meaningful influence increases your participation in combat and is therefore desirable.

The reason why a party of ~6 is usually more interesting in combat is that you are controlling 6x more characters as well as often their spatial relations. This is shitload more than controlling single, similarly realized character, meaning more influence, more tactics and less similarity between encounters.

On the other hand, when done excessively poorly (I'm looking at you, BG), controlling one character is far more enjoyable, as the extra options and influence boil down to micromanaging your brave party so it doesn't collectively off itself.

- Is this battle necessary or in any way useful ?

Fucking this.

Excessive amounts of filler combat hurt verisimilitude without adding anything of interest. To add insult to injury, the enemies are usually suicidal, even if they are a pack of wolves interested in meal or some thugs looking for some quick dough.

Individual combat encounters should be interesting, different and reasonably motivated.

- Interesting enemies

What you look for is diversity of combat encounters - not only through the means of the diversity and unpredictability of the enemies, but also thanks to different environments.

To sum this all up, what we need is ability to meaningfully influence combat (else why would it be interactive?), the amount of options and diversity of scenarios making search for good tactics not a mindless task (too much option to just brute-force it, too different scenarios for single universal tactics to exist), less filler and more sense.

That would be 1.

2. ???
3. Profit!

I like an idea of multiple conditions that affect combat abilities. The point of combat wouldn't be to reduce your opponent's health to zero, but get into a position so that you can deliver one deciding strike.

This, whittling HPs down is neither realistic nor interesting mechanics. It sucks no matter the point of view you adopt.
However, since in cRPGs combat encounters tend to involve single combatants and small to moderate parties rather than clashes of vast armies, I find mechanics based on a handful boolean flags still unsatisfactory.

For about a fucking decade we've had access to advanced physical engines simulating behaviour of non-rigid objects like bodies in real time. Yet, this ability has never been utilized in any purpose save for the purely cosmetic ones. For almost two decades we've had precise collision detection and arbitrary numbers and shapes of collision hulls. A decade is considered more or less an eternity in gaming industry, so someone please fucking tell me why the developers are so bloody fucking oblivious to the potential at their very fingertips?


What I want is a system where hits and their results are determined by physical engine, where "chunky salsa" rule is an integral part and natural consequence of the game mechanics, where armour consists of materials and actual collision hulls, and with stats tweaking the actual performance of combatants rather than simplistic attack rolls or outrageous damage scalers. On top of that I want the part of actual combatant AI responsible for combat actions to be reminescent of good beat-em-up AI and able to utilize its relative strength and weaknesses effectively (so the noob PC, for example, will get quickly parried, thrown off balance, knocked to the ground and coup de grace'd by AI with much higher melee stats).

There could even be both direct and indirect control modes, in the latter direct control would be assumed by game's combat AI, with player only issuing commands regarding movement, target priority, and special goals (like taking an NPC alive or aimed attacks).

You would also need to master different areas of combat. Taking down armored knights with a dagger on regular basis is not very likely, same as hitting swift opponents with huge, heavy weapon.

Indeed.
Main disadvantage of dagger is that it's fucking short, approaching a swordsman with dagger in a hand, counting on being able to stab him in one of his weak spots sets you up for one hell of lethal disappointment.

An often overlooked aspect is confidence and evaluation of the opponent. You could pretend to be weak and then strike with unexpected skill and force. Or the other way around, making your opponents insecure and nervous. More experienced foes would be less likely to be fooled.
Your reputation should matter a lot as well. Nobody yields to a murderous maniac who is expected to kill him anyways, and many won't even start a fight with you and rather run away.

Good idea. The former could use acting/bluff skills and empower otherwise weak characters like bards and thieves.

Chasing enemies who flee shouldn't be a chore. You should be rather relieved that no battle took place at all. Most game mods that make enemies flee end up being not very popular, because players don't want to miss the loot and/or experience. This needs to be offset, probably by the need to conserve your strength in order to remain efficient until the next opportunity to rest and recover.

Fuck, yeah. But traditional XP-based types of mechanics kind of throw a spanner in our works here.

Also, being outnumbered should be pretty much fatal in most cases. Giving more importance to positioning and movement, clever use of choke points like doors...

This would also be a powerful balancing tool, cutting the big damn heroes back down to much more believeable level, and telling the dilemma between player impunity and unreasonably powerful guards to fuck off and die.

 

[Suchy]

1) No repetitive filler combat!
2) Diverse enemies that require different tactical approach.
3) Character skills define your chances, even in case of twitch realtime.
If the first 3 are met, then single/party based or realtime/TB combat only adds different flavor.
Personally I prefer party based TB, but I definitely don't mind well made realtime (Mount & Blade).

 

[denizsi]

What I want is a system where hits and their results are determined by physical engine

So you want a consoltard game? Combat governed by float-point intricacies of physics simulation is the last thing we need to acheive reasonable realism in a RPG. Physics simulation, collision detection, these are just fluff gimmicks on their own. If they aren't dictated by stat outcomes, it's all shit and when they are dictated by stat outcomes, they will only serve to make the scene look realistic, not some "swings miss through enemy" bullshit as in Morrowind.

Taking down armored knights with a dagger on regular basis is not very likely, same as hitting swift opponents with huge, heavy weapon.

What exactly are these huge heavy weapons? Are we talking about big-as-shit fantasy weapons?

Also, the current belief is that taking down armored opponents often involved grappling and locking them down to the ground and killing with a dagger through the exposed openings in the armor. That's supposedly the primary use of daggers as carried by armored fighters. Just saying.

The entire concept of "rogue with dagger takes on Knight in full plate" bullshit needs to be purged, if you ask me. Occasionally make situations to make it work? Yes. Make it a fundamental part of combat where you got faggots with kitchen knives taking down men in plates left and right? No FFS! A thief is a thief, a rogue is a rogue, a man in fucking plate is a man in fucking plate. They should make their own contribution to the plot and to the elimination of men in shinies in their own roguish ways; not be a part of the Full-Frontal Tin-Can-Disposal-Squad of Doom.

Anyway, how do you like my ideas on combat here and here? I'm also applying all of that to a real-time model.

 

[denizsi]

Severe injuries on the player or at least on party characters should also be acceptable, though most sheeple would likely simply reload. I've read about this medieval fighter who lost his main hand in a battle. The guy was so bad ass, he had an iron mock-up hand with adjustable grip forged and he kept on fighting, still using the same arm as his main weapon-grip and lived on pretty fucking long by the standards of a medival fighter. If that isn't bad ass, I don't know what is. Now imagine things like that happening in your game and receiving feedback about it from the game world's inhabitants.

Sadly I don't remember his name but a quick Google on medieval prosthetic should immediately point to him.

3. No pausing during combat or accessing your inventory. The only things you should be able to use are your weapons, and whatever you have on your belt/other generic thing that gives easy access to an item.

Who the fuck fights with a bag of inventory anyway? Dropping your entire main load on the ground when the combat starts should be the damn standard!

 

[Cassidy]

If Fallout allowed for you to directly control all party members and that "but they're not professional soldiers" excuse be damned, if it had interrupts and if combat was better balanced to not make eye shots so overpowered and make grenades actually be useful, it would be the most entertaining combat system ever.

 

[denizsi]

You've just described JA2.

 

[Kaucukovnik]

What I want is a system where hits and their results are determined by physical engine

So you want a consoltard game? Combat governed by float-point intricacies of physics simulation is the last thing we need to acheive reasonable realism in a RPG. Physics simulation, collision detection, these are just fluff gimmicks on their own.

What if the direction, speed and force of the blow is directed by stats? Same for the opponent's ability to counteract. Physics engine could take care of the collision (solving graphical aspect at the same time), depending on mass and material of weapons and armor, and resulting bonuses, penalties and damage would be its output given back to the stat-calculating part of combat engine.

Taking down armored knights with a dagger on regular basis is not very likely, same as hitting swift opponents with huge, heavy weapon.

What exactly are these huge heavy weapons? Are we talking about big-as-shit fantasy weapons?

Also, the current belief is that taking down armored opponents often involved grappling and locking them down to the ground and killing with a dagger through the exposed openings in the armor. That's supposedly the primary use of daggers as carried by armored fighters. Just saying.

http://www.youtube.com/watch?v=bGKz58KsAB0
Sadly, there is no version with subtitles I'm aware of. You should get the idea anyways.

Yes, I'm aware that there is warhammer or warhammer. Honestly, I like over-the-top weapons, if they are part of some consistent theme, and symbolism is clearly more important than realism in given setting.

I'm quite curious against what realistic weapon would a dagger have a (realistic) advantage. The only ideas I have are polearms at close range or any big weapons in narrow corridors etc.

Anyway, how do you like my ideas on combat here and here? I'm also applying all of that to a real-time model.

I guess traditional skills like "long blade" or "axe" would be replaced with ones appropriate to mentioned aspects of combat. I'd welcome that.

 

[Awor Szurkrarz]

I'm quite curious against what realistic weapon would a dagger have a (realistic) advantage. The only ideas I have are polearms at close range or any big weapons in narrow corridors etc.

Dagger gives you an adventage in grappling range because you can repeat stabs very fast. Also, if someone attacks you from a close range and you need to draw a weapon very fast.

 

[soggie]

I'm quite curious against what realistic weapon would a dagger have a (realistic) advantage. The only ideas I have are polearms at close range or any big weapons in narrow corridors etc.

Dagger gives you an adventage in grappling range because you can repeat stabs very fast. Also, if someone attacks you from a close range and you need to draw a weapon very fast.

That is, of course, if you CAN get into that range in the first place. And even then, bringing a dagger to bear in grapple range is not as easy as it sounds. A longer bladed dagger, like a short sword, works equally well in that range too, but with a longer blade, it's easier to wound the opponent.

 

[Awor Szurkrarz]

From what I've read, an optimal length of a dagger used in combat was it's wielder's forearm. It sounds like a pretty long weapon.

 

[denizsi]

More or less. Any longer then that, it's usually the civilian sword, or "arming sword", or "messer", until it's longer enough to be called a longsword, disregarding single/double edgedness.

What I want is a system where hits and their results are determined by physical engine

So you want a consoltard game? Combat governed by float-point intricacies of physics simulation is the last thing we need to acheive reasonable realism in a RPG. Physics simulation, collision detection, these are just fluff gimmicks on their own.

What if the direction, speed and force of the blow is directed by stats?

Then why do you need a physical simulation at all? If it's going to be governed by stats, adding a layer of realistic simulation will most likely produce lots of unpredictable outcomes as per the physics, which will fall out of the ground your combat system covers where the outcome can not be worked into the stats system. Besides, show me one benefit to going full simulation. Accounting for material differences, resistances, amount of force absorbed by armor or sword and trauma inflicted on the target himself, all these can be handled to various levels of grades on a reasonable realism-gamism scale using simple animations and maybe physics based animations without resorting to a physics simulation, so would using a simulation not be wasteful on top of harder to implement?

Same for the opponent's ability to counteract. Physics engine could take care of the collision (solving graphical aspect at the same time), depending on mass and material of weapons and armor, and resulting bonuses, penalties and damage would be its output given back to the stat-calculating part of combat engine.

Maybe if anyone at all wants a purist medieval warfare simulation where they won't even understand why half the things that happen happen at all. Especially in the case of fantasy weapons, reliance on physics simulation would most likely be pure shit since they have no basis in reality at all.

Anyway, the most critical point is that I doubt anyone can dissect any form of physical combat, contemporary or historical, down to all of its component to be able to merge them back to be accounted for in a physics simulation. Closest examples are goofy-looking shitty "kinetic fight" games. Collision, mass, momentum, material resistances and what not, I don't see a single benefit to using all of these in a realistic physics simulation and I'm one to prefer simulationism in games over gamism in general.

Taking down armored knights with a dagger on regular basis is not very likely, same as hitting swift opponents with huge, heavy weapon.

What exactly are these huge heavy weapons? Are we talking about big-as-shit fantasy weapons?

Also, the current belief is that taking down armored opponents often involved grappling and locking them down to the ground and killing with a dagger through the exposed openings in the armor. That's supposedly the primary use of daggers as carried by armored fighters. Just saying.

http://www.youtube.com/watch?v=bGKz58KsAB0
Sadly, there is no version with subtitles I'm aware of. You should get the idea anyways.

Heh, actually there are huge two handed swords as a half-way weapon between longswords and spears, with sideway handles on the crossguard, to be wielded like a chainsaw and believed to have employed techniques from both weapons.

I'm quite curious against what realistic weapon would a dagger have a (realistic) advantage. The only ideas I have are polearms at close range or any big weapons in narrow corridors etc.

That's a close guess, I guess but polearm vs. dagger is still just as risky for the dagger-wielder as it is for the polearm wielder. In an open space, a skilled polearm user might even have a slight advantage over the dagger as polearms aren't exactly hulking weapons that clog down the wielder. In fact, nothing is supposed to be so.

On the other hand, when faced against a dagger, a polearm wielder could simply drop his polearm and draw his own dagger or sword as per common sense, I guess. Ultimately, it kind of comes down to what I said earlier: daggers aren't built for full-frontal assault against anyone.

Anyway, how do you like my ideas on combat here and here? I'm also applying all of that to a real-time model.

I guess traditional skills like "long blade" or "axe" would be replaced with ones appropriate to mentioned aspects of combat. I'd welcome that.

Kind of both, actually. The latter isn't something common folk or cutthroat outlaws could be holistically trained in without a proper background, so individual skills are still there with some overlapping between each other.

 

[DraQ]

What I want is a system where hits and their results are determined by physical engine

So you want a consoltard game? Combat governed by float-point intricacies of physics simulation is the last thing we need to acheive reasonable realism in a RPG. Physics simulation, collision detection, these are just fluff gimmicks on their own.

What if the direction, speed and force of the blow is directed by stats?

Then why do you need a physical simulation at all? If it's going to be governed by stats, adding a layer of realistic simulation will most likely produce lots of unpredictable outcomes as per the physics, which will fall out of the ground your combat system covers where the outcome can not be worked into the stats system.

Wait, what.

You seem to have everything ass backwards - why would you work outcomes into stats system? You should alter stuff like precision, speed penalty, balance and recovery time after swinging based on stats, assign masses, collision hulls, sharpness and materials to objects, then craft combat system out of those. You don't need to conform to some combat system, you create your own, then tweak stuff if something unexpected or lulzy pops up.

Besides, show me one benefit to going full simulation. Accounting for material differences, resistances, amount of force absorbed by armor or sword and trauma inflicted on the target himself, all these can be handled to various levels of grades on a reasonable realism-gamism scale using simple animations and maybe physics based animations without resorting to a physics simulation, so would using a simulation not be wasteful on top of harder to implement?

1. Integration of mechanics and presentation - traditionally one of the areas where RPGs suck *hard*. Implementing WYSIWYG in RPGs would probably be one of the best things to ever grace the genre, maybe even reviving the genre. First, this kind of mechanics would be highly intuitive, without having to be simple, or even easy to simplify. Second, this kind of mechanics would naturally convey itself through what player sees, and vice versa - for example, if armour forms a collision hull equivalent to its mesh, you don't have to account for things like open and closed helms or visors when coding, as the mechanics will automatically account for things like shooting enemy in the eyes, face or for armour coverage in general.
Which leads us to...

2. Content creation. While coding this kind of engine will indeed be a harder task (but by no means impossible), the creation of actual content will be greatly simplified - for example basic defensive properties of an armour piece will be defined by shapes and materials of its components. You have to create models anyway, and picking materials from the list is not much work either. RPGs tend to feature a shitload of content. As a bonus, such system will prevent the existence of high AC chainmail bikinis, which is always good, as this kind of armour is jarringly retarded. Of course chainmail bikinis are hardly the only act of jarring retardation an RPG dev team may accidentally commit, luckily physics-based mechanics can help, by enforcing...

3. Internal logic. There is an awful lot of stuff that while pretty straightforward and commonsensical when you think about it, is really hard to come up with mainly because of how obvious it is. Even if you could take into account all this stuff, you'd still have to implement every bit of it individually. For example, some parts of a giant monster may be impossible to reach for a aground based melee combatant, those will be different parts in case of a humanoid giant, than those in case of some giant quadruped, lastly said monster may expose different parts more or less during its attacks and other actions. Apply some basic collision detection? Problem? - solved. Similarly, you can replace BSB HP attrition with system based on volumes representing locations and organs, plus some system determining probability of failure on hit - may even be comparatively small (to avoid attrition) HP scale - add some general probability modifier to adjust difficulty between heroic and frighteningly realistic. The main advantage of physical engine is that it derives higher level rules directly from low level fundamental ones, rather than have the developer guess the high level ones using the ancient ritual of asspulling. But those are merely basics, what we really want is...

4. Flexibility. Once upon a time, there were games where nerds sat around a table rolling polyhedral dice. Those games featured powerful wetware narrative engine known as GM, typically implemented in one of the nerds, plus a relatively skeletal set of rules tacked on, for retrospectively determining if some action was very stupid and rewarding its perpetrator with an opportunity to reroll character if it indeed was. Then, cRPGs came and often seemed very similar, until a horrible revelation struck - they had no narrative engine of any sort. Realization of this horrible truth was a massive sanity hit for many, some (hi mondblut!) never fully recovered and maintain their tenuous grasp on reality by vehemently denying cRPGs right to incorporate any meaningful dynamic narrative at all.
Luckily, there are also aspects where computers don't suck. While lousy storytellers, computers seem to enjoy grinding numbers. Unlike your typical nerd, even one seeking relief in this ultimately mindless activity (hi mondblut!), they can also do it frighteningly fast. Finally, through a fortunate coincidence, there is a lot of stuff that can be simulated via number crunching. The problem is that for an event or object A to influence an event or object B in our simulation, they must be interconnected by some comprehensive mechanics. In a PnP game, powered by a nerd running a narrative engine, you can declare an attack on a guard using wooden chair - the engine knows what the chair is, what the guard is and how said chair, or indeed any sufficiently hefty object may serve as an improvised weapon, and may even devise so more or less accurate stats for it if ruleset provides none. In cRPG, however, all the computer knows is that the chair is some abstract object with a bunch of polys representing it attached, and that guard is another abstract object - unless we tell the computer that the chair is indeed a weapon with predefined stats it will just go "Bwuh? " on us, causing much dismay by failing to accommodate our cunning plan.
What will happen, though, if we take another approach, and instead of defining stats of all possible weapons armours, projectiles and such, we provide all movable ingame objects with approximate physical stats and let them all be weapons, shields, cover, projectiles, ballast or whatever else the game may demand depending on context? This flexibility will not only provide much needed space for player's creative endeavours, but also free our content creators and quest designers from constraints of arbitrary limitations, like having to pick from a narrow set of possible actions.

Maybe if anyone at all wants a purist medieval warfare simulation where they won't even understand why half the things that happen happen at all. Especially in the case of fantasy weapons, reliance on physics simulation would most likely be pure shit since they have no basis in reality at all.

Quite the contrary, they are usually recognizable as certain weapon types, sometimes endowed with some extra qualities, like super sharpness, boosting user's abilities, elemental effects, etc. - those can be added to the simulation. Oh, they often are retardedly ornate, but as long as basic shape is more or less correct, this shouldn't matter at the level of detail we are discussing.

Anyway, the most critical point is that I doubt anyone can dissect any form of physical combat, contemporary or historical, down to all of its component to be able to merge them back to be accounted for in a physics simulation. Closest examples are goofy-looking shitty "kinetic fight" games. Collision, mass, momentum, material resistances and what not, I don't see a single benefit to using all of these in a realistic physics simulation and I'm one to prefer simulationism in games over gamism in general.

You can always make hybrid solutions for stuff you can't simulate on base level.

In any case it should still be much better than tables of arbitrary, asspulled values that have been the norm since the dawn of PnP.

Also, what "shitty kinetic fight games"? My role model here is Cortex Command, except in 3D, with no detailed pixel physics or terrain simulation, more attention to character movement, environmental/elemental effects and an RPG built around it. The game is rather primitive, but still very impressive in its use of destructible attachables, physical mayhem or adaptive animations. And it's a textbook indie.

I don't want to relive the 'glorious' past of cRPGs, I want the genre to stride boldly into the future, except I want it to be the promised future that dawned before us back when computer gaming was taking its baby steps, when games like Frontier were taken to be forerunners of complex and detailed virtual worlds waiting for enough megahurtz and megabytes to contain them springing into existence, not the future full of bloom, unholy amalgams of simplified stats and dumbed down action, shaped by greedy suits forming comitees.

 

[denizsi]

You want to introduce too many parameters. Even something as mundane as swinging your sword in a specific direction will require a shitload of premade data for it to work, including a limited physical framework so your anime should pads or impossibly huge anime swords won't fly around uncontrollably.

I talked about working the outcome of a physical simulation into the stats because once the starting parameters are set and the weapons sent flying, the system will need to make sense of every collision and you can't possibly account for every collision and expect to make it work realistically with all materials, resistances and forces in a cohesive union.

Ok, I made a diagonal cut at this guy, who made a counter-attack so our swords bind somewhere around the middle. Is there a point to running a physical simulation on our swords so there's a 0.000000000000000001 increment towards degradation? Where's the beef? Yes, I swung, he swung, our blades met in the middle, and the worlds moves on. Or the exact spot where our swords binded is 0.000000043546457 points away from me past the mass center of my sword and 0.0000002314453 points closer to my opponent from mass center of his sword so now according to these complex calculations here, my opponent has a slight leverage advantage of 0.0954345 points. So, he decides to use this average by pulling a half-swording move and starts to rotating his grip over and around the tip of my sword to force my sword down, <engine makes the physics calculation> and he succeeds and now he attempts to reach with his left hand towards my grip, crossguard or ricasso while applying 0.045375638457 points of force while I try to pull myself back with 0.75348645987 points and ha! I broke out of his lock.

How very exciting. Instead, we can simply grade the where the swords bind on a small abstract scale based on the input from physical collision, like 1 to 10 because both of these gentlemen are swordsmen and their bind is sure to be within a certain threshold around the mass centers of their blades and for rare cases where either can make a critical mistakes with that, we can simply introduce another abstract situation to account for it and so on and so forth. The key difference is, unless a critical failure was involved, the brief physical simulation here will be actively forced to correspond to any point within the threshold of our abstract scale, while still looking just as good as the real simulation.

What will happen, though, if we take another approach, and instead of defining stats of all possible weapons armours, projectiles and such, we provide all movable ingame objects with approximate physical stats and let them all be weapons, shields, cover, projectiles, ballast or whatever else the game may demand depending on context? This flexibility will not only provide much needed space for player's creative endeavours, but also free our content creators and quest designers from constraints of arbitrary limitations, like having to pick from a narrow set of possible actions.

No, instead the designer will have to worry about materials, mass centers of objects and intricate components since we are doing physics based on material properties. Every game object will have to be digitally crafted in a way true to the actual craft where you have intricate individual parts of an armor and even the rivets on it since you want details or individual handle, blade and crossguards for swords, for instance because you need the mass centers of the overall object and the individual parts both. You swing a weapon based on the mass center of the overall item but whenever your weapon collides with anything, energy is transfered through individual pieces based on their individual properties so you want to account for every crook and nanny, this is how you do it.

You can not simply make a weapon model, assign single material/mass properties to it and call it a day. The physical reaction when the blades of swords bind, when a blade meets the opponent's crossguard or when one of the combatants hit the other with the pommel of his weapon, are different.

Perhaps there's something I'm missing here, I don't get this physical simulation fetish at all. I %100 agree with providing detailed environments that allows for emergent uses in procedural ways using smart physics and further abstractions whenever either is more feasible. I agree with adopting a high level of simulationism that could in turn allow for even a higher quality of gamism, but I just don't think that actual physical simulations are that necessary at all.