Damage cap

[DraQ]

So I've been thinking about damage scaling with strength (melee, bows), or speed or some attributes in general and came to a conclusion it doesn't really make sense.

Imagine you're stabbing someone with a dagger. "Damage" you deal will increase with more strength or speed but ony to a point. If you bury weapon up to the hilt in someone's body, then putting more strength won't result in extra damage.

The extra damage is not really straight damage anyway, just penetration. First, you don't deal any, then if you break the armour your damage increases only until full penetration is achieved. Even the penetration itself doesn't increase endlessly with increasing strength - a weapon may just be incapable of penetrating armour and putting in more strength will simply break it - relative durability or hardness should be a factor.

The max damage, measured with tissue disruption, done with thrusting weapon also depends on the target's bulk. If penetration of armour and body is full, then the damage dealt is either weapon's damage cap or creature's/bodypart's damage cap, whichever is lowest. Weapon's damage cap would correspond to it's length times base damage multiplier (corresponding to weapons width), while creature's/bodypart's damage cap would correspond to the distance weapon can travel inside before going out on the other side.

Depending on details of the system the cap may apply to raw damage that can be multiplied by crit hit multipliers or critical hits may depend on mechanics different from just damage (like hitting particular organs, % kill chance or whatever).

Now, thrusting weapons aren't the only ones in existence. With slashing attacks damage should scale higher (because it only caps when part of the body is severed and max damage generally increases with square of weapon's length), but the mechanism is the same.

Of course, damage cap shouldn't be the only factor in dealing damage, different weapons and attacks would also have different penetration modifiers - for example slashing wouldn't penetrate as well as stabbing.

From this perspective blunt weapons would be a really wide and not very sharp slashing weapons (lots of damage, abysmal penetration) but that would be off. Blunt weapons have ability to bypass armour by transferring kinetic energy to the target rather than disrupting it physically. But since you can bludgeon someone with a pommel or break their bones with a zweihander even if you can't cut through the mail, there should be no distinction here and all weapons should also deal damage via energy transfer term. It should depend on weapon's mass but also how much of a swing got soaked by armour and body (part soaked by armour would get partially dissipated) - for example imagine you're slicing someone with a heavy, but absurdly thin and sharp blade - despite high mass you wouldn't really deal any significant amount of blunt force trauma as it would just pass through without slowing.

Obviously, it would be easier to implement something like this starting from a physical engine, but the point I was trying to make is that scaling damage with attribute should be heavily capped and that most of it should be handled by scaling armour piercing modifier instead.

 

[Telengard]

Being a fan of Millennium's End Combat (armor converts damage from cutting or piercing to impact, damage increases by location hit not how high rolled, damage is trauma-based determined by type of wound not be ablatively reducing hp), I like the direction of your thoughts, with Strength affecting piercing ability instead, and all the rest.

The actual cap, that would depend on its intended purpose. The limits on damage via strength as I read your examples tend to be less about the amount of damage the person can do to another, but more about how much the target can take. If a person has 100 Str, then that person could theoretically do 100 hp of damage, but if the target's arm gets severed after 10 hp are dealt, that is it for damage - the person does 10, and that's all. However, that doesn't change the fact that the attacker could potentially have done 100, it just means that the other 90 were wasted effort. But if at a later date he attacks something that had 150 hp, his full strength comes into play.

A damage cap (or a damage multiplier) based on they type of attack, for one, has interesting possibilities.

 

[Johannes]

Imagine you're stabbing someone with a dagger. "Damage" you deal will increase with more strength or speed but ony to a point. If you bury weapon up to the hilt in someone's body, then putting more strength won't result in extra damage.

Well, that's not fully true.

 

[Norfleet]

Yeah, not true. You could bury the weapon up to your elbow, for instance, or yank it about forcefully to rip up some more innards. Ultimately, however, you're not really going to get the finer nuances of damage unless you go to extreme length and ditch "hitpoints" entirely, instead modelling damage based on perforated bodyparts and what exact effect this has on the guy you stabbed. As long as damage is always going to be abstracted to hitpoints, the system is always going to be fairly simplistic.

If a person has 100 Str, then that person could theoretically do 100 hp of damage, but if the target's arm gets severed after 10 hp are dealt, that is it for damage - the person does 10, and that's all. However, that doesn't change the fact that the attacker could potentially have done 100, it just means that the other 90 were wasted effort. But if at a later date he attacks something that had 150 hp, his full strength comes into play.

That's because you've subdivided a target into separate parts. You're no longer hitting "dude" but "arm of dude". The arm of dude is destroyed at 10 hitpoints. Dude now suffers whatever collateral damage occurs from having a bleeding stump where his arm used to be.

 

[Telengard]

On a stabbing wound, once you've gone through-and-through, all the excess damage happens to empty air behind the person. It's all wasted potential after that. (Which is why ammo that doesn't go through-and-through does more internal damage than that which doesn't. Something that bounces off the bone does more damage than something that goes straight through.) The hilt could do some extra during its passage, and after that it's over. To wrist, arm, elbow, shoulder, it's all the same. And the point still stands - no more damage gets done after a certain point, no matter the force behind the attack.

Of course, you can jerk the knife afterwards, but that's a separate, secondary cutting maneuver.

If we wanted to actually be really real, for most attacks you shouldn't be using your full strength, and you wouldn't be getting your strength bonuses anyways. In a proper attack, you would only use as much effort as is required. Not only is wasted effort tiring, weakening your overall strength, more importantly it draws your weapon farther out of alignment than it should be and needs to be.

 

[Excommunicator]

If you stab someone with a dagger, it's true that the piercing damage will hit a limit at which no more piercing damage will be inflicted, but a dagger with a hilt is going to have all that extra damage converted to blunt damage. The hilt is going to slam into them like a blunt weapon if you are strong enough, and if you are strong enough to push it past the hilt then it becomes slashing damage once again (until your hand reaches the wound, then the process begins again up until the shoulder). If you want realism, there should be never be a hard cap on the effects of strength. You can limit certain parts of it, but you will always have things like blunt damage and the amount of force that's going to contact the body (knocking them off their feet or concussing them etc), scaling upward to infinity.

Think back to the realism of it though: if you have enough strength to push a hilt through a person, then they'll need to be tied down or it won't happen. That amount of force will always knock a person from their feet, at which point most of it the energy will be wasted in turning the person into a projectile. I hate how movies and games treat characters as if their feet are set in concrete. Why the hell don't more games simulate knocking enemies flying with powerful enough strikes instead of stupidity like punching through their body or chopping fully armoured targets into pieces?


There is only so much gain to simulating things that far anyway. If you can do enough damage to push a weapon's hilt through their skin, bones and organs then it's over for them, they'll die almost instantly from any weapon in no time at all. There's no mechanical gain beyond making it look more impressive and you need to polish off the rest of the game before you can think about that sort of crap

 

[Awor Szurkrarz]

So I've been thinking about damage scaling with strength (melee, bows), or speed or some attributes in general and came to a conclusion it doesn't really make sense.

Imagine you're stabbing someone with a dagger. "Damage" you deal will increase with more strength or speed but ony to a point. If you bury weapon up to the hilt in someone's body, then putting more strength won't result in extra damage.

You really need to watch Rambo IV.

 

[Weierstraß]

Abstraction. Is this an RPG or a stab-physics simulation?

 

[Telengard]

Not all RPGs abstract to the point that DnD does. To detail how Millenium's End works (one of the few that even tries to tackle some of these issues):

You take an outline drawing of a body in a stance. There are 25 hit locations on the body. A target transparency is placed over it, with the center at the hit location you select. If you succeed at your attack, you hit your chosen location. If you miss, there's a spiral of numbers around the target's center. You find the amount you missed by on the spiral, and that's where your shot went, potentially missing, or potentially hitting another location. Target location being important because each location has a different multiplier to damage.

Guns do a fixed damage based on caliber and type of ammo. If it contacts armor, part of the piercing bullet damage is then reduced and part converted to impact damage. (Impact damage still being potentially deadly, just far less harsh than piercing or cutting damage.) The result is then multiplied by the target location damage modifier. Stacked with the body's resistance, this can mean several different critical traumas, as well as a direct reduction in skills related to that body location. Weapons like Katanas work similarly, but can be used to slash or stab, and have a different damage multiplier based on the type of attack used. Its damage is based on strength initially, strength normally allowing for a damage range of 2-6, which number is then multiplied by the weapon's attack type damage modifier, or 4.0 and 3.6 for the katana. Other than that, it's the same as bullets.

It's a very different RPG, with no ablative hp involved, and it ain't a system that works for heroic fantasy, but not all RPGs have to be heroic fantasy. And it can be highly useful when, say, making a Vietnam War RPG campaign. A game like that places the player in a completely different mindset.

Similarly, a damage cap could potentially form a game where tactical situations arise that just could not occur in a game abstracted to the point of the single hp line, and that are favoring strength- and big-sword-based damage.

 

[Murk]

EDIT: Those before me answered in similar or same ways, and I mostly concur with the above two posts. I, personally, am very fine with an abstracted system that is internally consistent and mostly balanced.

It is my belief that such discourses on overly realistic interpretations of physics, force, and "damage" are pointless in game systems that ultimately abstract them to cases of HP that represents in each unit, not necessarily that you are damaging then and there but rather, damage overall.

The vestigial attacks in the IE games are a silly apparition, but the explanation of them is about as apt to the idea of HP and damage in cRPGs as one is likely to encounter. Without things like location targeting, organ damage, "rending", and other factors there's no realistic way to measure damage.

Think this -- you stab someone with a knife then pull that knife upwards through their belly to their heart. Dead.

Now think this -- your fighter does a full attack round on an unarmed enemy with 100 hp and hits 3 out of 4 times for a total of 65 damage. Enemy is alive, yet if you had the time and chance to swing a greatsword at someone 4 times and connected 3 of those times, why not just stab them once and run the blade out one side to effectively decapitate them?

Because combat is heavily abstracted and things like "successful hits 3 out of 4" times are easily and adequately represented in a system and things like "you hit, you decapitate, insta-death" are represented with critical effects -- when realistically a good fighter should decapitate EVERY time (especially if he is able to attack 4 times in 6 seconds).

I believe base attributes, skills, and derived attributes are also such abstractions. Strength does not necessarily translate to "real world strength" but rather some kind of base attribute that can directly contribute to damage output. If you felt better having a general "Attack Power" derived attribute, then sure go for it, but ask yourself if the end result is worth the additional layer of complication.

So I guess what I'm really saying is if you want a real-world-physics system you would need to design it not using any current abstracted system, and you would run into the problems games like Red Orchestra run into -- it is novel and neat, but after a while, you just kind of want to play a game and not a stabbing and surface-area-cutting simulator.

 

[Excidium]

Is this what autism looks like

 

[Telengard]

HP systems have another effect - they heavily favor heroic actions. If a comrade is sniped out in the open, under a strict HP system it's actually not all that risky for everyone to just run out there a grab him, since it is only multiple wounds that really have a chance to take a person down. The sniper can't do enough raw damage to stop the group from doing whatever it wants. However, in a game where a shot to the chest is fine, but a shot to the head is deadly (even if it doesn't kill you), there is real risk in running out there, and players will actually then weigh those risks. They will also, under such a system, naturally use covering and enfilade fire, not because it applies an arbitrary bonus stat to the attack, but because it is incredibly advantageous under the basic rules of the system. - You don't get that experience under a strict HP game.

As for capping damage, in the same way, if you don't use one, then under the usual sort of combat rules, the game favors hulking strength and huge weapons - which depending on what era of history your game is set in, shouldn't necessarily be what is favored. A cap will do many things, but right off the bat, it will favor skill over raw power, since raw power then has a limit in what it can get you. Without changing anything else from the usual combat system, just adding a cap, you already have a better representation of the era of smallblade duels than you normally get out of most games. And a game system should ideally reflect both the ideas that the designer is trying to represent and the era the game is set in.

It does not need to be exacting in detail, but if you make a game set in the fencing era, and the greatsword is the best fencing weapon, you did something wrong.

 

[DraQ]

Yeah, not true. You could bury the weapon up to your elbow, for instance, or yank it about forcefully to rip up some more innards. Ultimately, however, you're not really going to get the finer nuances of damage unless you go to extreme length and ditch "hitpoints" entirely, instead modelling damage based on perforated bodyparts and what exact effect this has on the guy you stabbed. As long as damage is always going to be abstracted to hitpoints, the system is always going to be fairly simplistic.

Well, given that you aim to make a computer game, you need some numerical way of quantifying damage. It doesn't have to be full body HPs (because this is guaranteed to suck, unless you're making massively scaled strategy and using HP ranges on the order of single attack range), but having small ranges of organ, bodypart or subsystem HPs assisted by numerous additional effects and preferably separate structural and functional damage is fairly reasonable assumption.

Not all RPGs abstract to the point that DnD does. To detail how Millenium's End works (one of the few that even tries to tackle some of these issues):

*snip*

And this is what you can do while still in tabletop realm, now harness the power of a modern PC for your mechanics.

It's a very different RPG, with no ablative hp involved, and it ain't a system that works for heroic fantasy, but not all RPGs have to be heroic fantasy. And it can be highly useful when, say, making a Vietnam War RPG campaign. A game like that places the player in a completely different mindset.

Shit, I would like more RPG mechanics and settings putting you in such mindset.

Similarly, a damage cap could potentially form a game where tactical situations arise that just could not occur in a game abstracted to the point of the single hp line, and that are favoring strength- and big-sword-based damage.

Think this -- you stab someone with a knife then pull that knife upwards through their belly to their heart. Dead.

Now think this -- your fighter does a full attack round on an unarmed enemy with 100 hp and hits 3 out of 4 times for a total of 65 damage. Enemy is alive, yet if you had the time and chance to swing a greatsword at someone 4 times and connected 3 of those times, why not just stab them once and run the blade out one side to effectively decapitate them?

Indeed, why aren't our systems working like this? Maybe apart from giving the dude stabbed in the heart from 15s to sever minutes of consciousness (depending on various factors) and about extra minute of life afterwards (this can be very relevant if your setting has advanced technology or magic).

Why the fuck should I have to hack or skewer some unarmed, unarmoured fuck 4 times if I'm a proficient fighter? Is my weapon a fucking sword or a cluebat?

Because combat is heavily abstracted and things like "successful hits 3 out of 4" times are easily and adequately represented in a system and things like "you hit, you decapitate, insta-death" are represented with critical effects -- when realistically a good fighter should decapitate EVERY time (especially if he is able to attack 4 times in 6 seconds).

Not "heavily". "Shittily" is the word. Two similarly armed and proficient fighters should have trouble instakilling each other, especially when wearing armour, but instakilling unproficient, unarmed guy should be the norm. If your system doesn't account for that it can't be agood abstraction, because good abstraction conserves behaviour while dropping complexity.

I believe base attributes, skills, and derived attributes are also such abstractions. Strength does not necessarily translate to "real world strength" but rather some kind of base attribute that can directly contribute to damage output.

That's not an abstraction, that's desperate handwave.
Abstraction would be strength translating to real world strength, but instead of accounting for, say upper and lower body strength, and multiple other factors, approaximating it as single numeric value, while still having it produce realistic outcomes in the context of simulation.

but after a while, you just kind of want to play a game and not a stabbing and surface-area-cutting simulator.

Then I'll play Go.

Why would I not want to play stabbing and surface-area-cutting simulator if the game is about stabbing and cutting and I have a computer to handle all the bothersome stuff?

HP systems have another effect - they heavily favor heroic actions.

Making them non-heroic as a consequence. It's like "button - awesome" of PnP era, heroism for the masses.

under such a system, naturally use covering and enfilade fire, not because it applies an arbitrary bonus stat to the attack, but because it is incredibly advantageous under the basic rules of the system. - You don't get that experience under a strict HP game.

And this is the hallmark of a well designed system - you don't need to append it with superfluous shit in order to add interesting behaviour or balance things.

Simplicity of the rules, complexity of situations.

As for capping damage, in the same way, if you don't use one, then under the usual sort of combat rules, the game favors hulking strength and huge weapons - which depending on what era of history your game is set in, shouldn't necessarily be what is favored. A cap will do many things, but right off the bat, it will favor skill over raw power, since raw power then has a limit in what it can get you. Without changing anything else from the usual combat system, just adding a cap, you already have a better representation of the era of smallblade duels than you normally get out of most games. And a game system should ideally reflect both the ideas that the designer is trying to represent and the era the game is set in.

It does not need to be exacting in detail, but if you make a game set in the fencing era, and the greatsword is the best fencing weapon, you did something wrong.

Is this what autism looks like

Look at you, insect...

 

[Murk]

Indeed, why aren't our systems working like this? Maybe apart from giving the dude stabbed in the heart from 15s to sever minutes of consciousness (depending on various factors) and about extra minute of life afterwards (this can be very relevant if your setting has advanced technology or magic).

Why the fuck should I have to hack or skewer some unarmed, unarmoured fuck 4 times if I'm a proficient fighter? Is my weapon a fucking sword or a cluebat?

Good question, I guess some attempts to alleviate this have come in the form of aimed-attacks, more specific criticals, and eventually things like "damage resilience thresholds" which is a phrase I made up for a game that calculates something along the lines of "if a single non-critical basic attack inflicts over 50% damage on an enemy, the enemy must save-or-die". The threshold representing varying degrees of when that save or die is initiated.

Not "heavily". "Shittily" is the word. Two similarly armed and proficient fighters should have trouble instakilling each other, especially when wearing armour, but instakilling unproficient, unarmed guy should be the norm. If your system doesn't account for that it can't be agood abstraction, because good abstraction conserves behaviour while dropping complexity.

I can agree with this, though if combatting mooks turns into instakilling them then I'd also suggest a situation where weak enemies bump into you and beg for your gracious permission to leave unharmed (or perhaps offer to bribe you some set amount).

That's not an abstraction, that's desperate handwave.
Abstraction would be strength translating to real world strength, but instead of accounting for, say upper and lower body strength, and multiple other factors, approaximating it as single numeric value, while still having it produce realistic outcomes in the context of simulation.

Eh, at this point it's just how far one wants to take the abstraction. Like I said, if you want to call it 'attack power' instead of strength then I'm fine with it, but for most games having a Strength score is easy enough without sacrificing much in return.

Then I'll play Go.

Why would I not want to play stabbing and surface-area-cutting simulator if the game is about stabbing and cutting and I have a computer to handle all the bothersome stuff?

Hrmm... while clearly possible I fear how worthwhile such implementation will be. It will require a lot of work as it would be a thoroughly specific combat and feedback system for that combat, otherwise damage would seemingly be high and at other times be low.

But then, would there be a skill that determines the likelihood of how often you inflict one of these "large cut" types of attacks? If so, why not simplify it with something with a level or skill or stat comparison and allow for an appropriate increase in critical hit likelihood. After-all, critical hits are not "unlikely hits" but rather major ones.

I'm thinking something like Level 20 fighter vs level 1 goblin = 95% critical hit. Or, if you prefer to avoid a leveling issue like that have something like Str 22 fighter and Con 12 goblin = 60% critical hit. I just imagine these are much easier to implement and likely to give a similar or close enough to result than things like "surface area damaged, depth of damage, shock absorption through clothing and armor, "twisting" of blade, etc."

 

[Weierstraß]

HP systems have another effect - they heavily favor heroic actions. If a comrade is sniped out in the open, under a strict HP system it's actually not all that risky for everyone to just run out there a grab him, since it is only multiple wounds that really have a chance to take a person down. The sniper can't do enough raw damage to stop the group from doing whatever it wants. However, in a game where a shot to the chest is fine, but a shot to the head is deadly (even if it doesn't kill you), there is real risk in running out there, and players will actually then weigh those risks. They will also, under such a system, naturally use covering and enfilade fire, not because it applies an arbitrary bonus stat to the attack, but because it is incredibly advantageous under the basic rules of the system. - You don't get that experience under a strict HP game.

I don't see how that at all follows from using an HP system. The situation entirely relies on the relative damage from the sniper and player HP, and on the distribution of damage from the sniper over different probabilities.

 

[Telengard]

HP systems have another effect - they heavily favor heroic actions. If a comrade is sniped out in the open, under a strict HP system it's actually not all that risky for everyone to just run out there a grab him, since it is only multiple wounds that really have a chance to take a person down. The sniper can't do enough raw damage to stop the group from doing whatever it wants. However, in a game where a shot to the chest is fine, but a shot to the head is deadly (even if it doesn't kill you), there is real risk in running out there, and players will actually then weigh those risks. They will also, under such a system, naturally use covering and enfilade fire, not because it applies an arbitrary bonus stat to the attack, but because it is incredibly advantageous under the basic rules of the system. - You don't get that experience under a strict HP game.

I don't see how that at all follows from using an HP system. The situation entirely relies on the relative damage from the sniper and player HP, and on the distribution of damage from the sniper over different probabilities.

I'll run it through percentages.

Given a game where people have a low 20 hp, and a sniper rifle that has the potential to instakill with a damage of d20, if the sniper gets a hit, then there is a 5% chance of death. If instead the sniper rifle does d100 for damage, there is an 80% chance of death. If the sniper rifle does d1000, there is a 98% chance of death. And now we're at the chance for death for the game I mentioned, if you're shot in an unprotected part of the noggin. (Which is why, in that game, you protect your head at all times.)

Of course, you can step back and abstract this more with ablative hit points, as most games do, since a shot to other parts of the body isn't a kill, and the sniper won't necessarily hit you in the noggin (though chances are high if you're out in the open). And so, those games reduce the damage of the sniper rifle to better reflect shot to kill ratio for hits across all parts of the body. But the more you abstract, the less chance there is of the sniper being able to count on a kill. And if the sniper can't count on the kill, those he's sniping can count on him not getting the kill. Such as: a d10 damage with a x5 crit. Even though any one shot could kill, the shot to kill ratio is so low that the risk for everyone just running out there is minimal, since the sniper has to score a critical (instead of just a hit) and then roll at least a 4 for damage.

Or to put it in concrete terms with Jagged Alliance 2 (a game that also uses some of these ideas). You can run out into the open, get shot in the head and die with the hit location variable. The chances are fairly low, but it can happen. Abstract that over to a single core hp model with criticals, though, and the critical for all locations then get averaged out to a single value or simply erased (far more common). Either way, the risk of instant death from a single bullet for doing something stupidly heroic gets reduced.

 

[DraQ]

Good question, I guess some attempts to alleviate this have come in the form of aimed-attacks, more specific criticals, and eventually things like "damage resilience thresholds" which is a phrase I made up for a game that calculates something along the lines of "if a single non-critical basic attack inflicts over 50% damage on an enemy, the enemy must save-or-die". The threshold representing varying degrees of when that save or die is initiated.

Perhaps, but I still see traditional HP systems as kludge in cRPG context, especially if the presentation and controls allow more detailed mechanics to make difference.

If you're going to have detailed 3D models, physical engine governing how stuff collides and reacts to collisions, plus some material system, then why not go all the way and use those building blocks you already have in your health system?

Instead of "ruffian gets hit for 12HPs, inflicting critical hit and killing him according to whatever table or other piece of mechanics the game uses" you can now simply go "Ruffian gets hit in the neck :insert parameters blow, weapon and neck: resulting in decapitation. He's dead because he has no head".

Yes, all this stuff seems needlessly complicated, but you're going to use it anyway, and besides, it's just frontloading the work that makes such impression. You won't have to modify such system if you introduce a creature type or weapon vastly different from your original assumptions, you won't get absurd shit if you refrain from such modification. You'll just get system that works and ensures its correctness by staying close to the thing it seeks to simulate. It'll also be intuitive, allowing for better integrated feedback and general presentation.

Finally, it will integrate stuff happening with gameplay.

I can agree with this, though if combatting mooks turns into instakilling them then I'd also suggest a situation where weak enemies bump into you and beg for your gracious permission to leave unharmed (or perhaps offer to bribe you some set amount).

Hell, why not? The NPCs, including enemies, should behave realistically. If they are not meant to be suicidal kamikaze squad, they shouldn't behave like one.

Eh, at this point it's just how far one wants to take the abstraction. Like I said, if you want to call it 'attack power' instead of strength then I'm fine with it, but for most games having a Strength score is easy enough without sacrificing much in return.

The problem is that attack power is relative, circumstantial and doesn't mean one thing.

Trying to abstract it to a single value will result in one of two possible disasters - system turning into incoherent mess, or system becoming so simplistic that it will become uninteresting - "rock, paper, scissors" boring? Think "paper, cardboard, chainsaw".

Hrmm... while clearly possible I fear how worthwhile such implementation will be. It will require a lot of work as it would be a thoroughly specific combat and feedback system for that combat, otherwise damage would seemingly be high and at other times be low.

You already have feedback system for such combat. Just show what the fuck are you using, how are you swinging it and what happens as a result. If your combat calculations are done by physical engine, you can just show player what happens, without even using that much special effects.

Don't think Morrowind's "whiff! whiff!" air-sword.

Think this as our proof of concept (lulzy mod, but it shows relatively well what I have in mind):

The game in gif is Cortex Command, and while it has a HP system, it basically only represents blood loss or equivalent - a lot of damage happens outside of it and can override it completely. Basically, character is comprised of body parts. Each body part is a physical object with mass, collision, material and so on. On top of that there are attachements that are also physical objects and usually play the role of armour. Now, getting rid of attachments and bodyparts can be accomplished in several ways.

First, there is a wound system - if something hits the object hard enough for its "sharpness" parameter to overcome object's resistance the object gets wounded or dented (visibly, at the point of impact). An object can acommodate finite number of such wounds - too much and it gibs. For bodyparts such wounds will generally result in bleeding, gradually reducing HP score of an actor.

An object can also be subjected to force overcoming it's joint strength with which its attached to actor. Then it flies off - helmets can be ripped off your head, limbs fly off in bloody arcs, etc. (of course flying bodypart or piece of armour can also be a dangerous projectile).

Finally, if an object is subjected to too much static or dynamic force (for example when propped against something or attached with high joint strength) it will be gibbed.

Losing a bodypart will usually result in direct HP damage and bleeding, but the loss will affect an actor far more directly. For example the actor will become lighter (because limbs have mass), and won't be able to use lost bodyparts, for example an actor without hands won't be able to use weapons, an actor without legs, won't be able to walk - those aren't some extra pieces of logic - having no arms simply gives you no attachment point for weapon, while having no legs means they won't be pushing against the terrain. If your head or torso gets destroyed, then you die, regardless of your HPs - chunky salsa rule (and how many systems can you honestly say to not need chunky salsa rule invoked by a human).

The system is very simple, yet it produces far more complex outcomes than any system used in an RPG aside of maybe DF.

I have even seen "migraine gun" implemented in one of the earlier versions (prior to allowing scripting effects into the engine) that simply made people's heads explode by firing a beam of alternating negative and positive mass particles at resonant frequency of their heads. It didn't do any damage, it didn't even push or pull them visibly. It's simply that you aimed it at someone's head, held the trigger down for some time and suddenly the head exploded into ludicrous gibs.

Notice that there is no doubt why a character can't move or has suddenly died in such system. Everything is instantly clear because you can see it directly.
Leave textual feedback for things you can't really show. It will be better for both text (ability to put more text when it's actually indispensable) and gameplay mechanics (not swamping player with textual feedback for stuff he could actually just see).

Yeah, Cortex Command is a 2D game and is already computationally expensive, but that's because each pixel of the terrain is a potential particle and physical object - you don't need that for raw damage system, and doing it in 3D, without particle terrain would allow you to use the computing power of your graphics card for physics engine needs. Going 3d would actually make the underlying logic simpler, because it would eliminate the need for objects to overlap and phase through each other on certain occasions (like arms, torso and projectiles colliding with them). Yeah, it's been long in development hell, but the features I'm speaking of were in even in really early builds.

But then, would there be a skill that determines the likelihood of how often you inflict one of these "large cut" types of attacks? If so, why not simplify it with something with a level or skill or stat comparison and allow for an appropriate increase in critical hit likelihood. After-all, critical hits are not "unlikely hits" but rather major ones.

Why should there be a skill for doing that? You control when you want to swing full force, possibly exposing yourself and giving the target time to react and when do you want to execute weaker, but faster and more controllable attacks. You control where do you aim the attack and what type of attack do you perform. Skill can at most control parameters of your swing and stance (speed, recovery time, accuracy, weapon alignment error, weapon grip strength, how easy are you to push out of balance), availability of special attacks (but they'd better work as situational perks and generally similar to spells for casters - learned from sources or randomly available on level ups based on skill level, your intelligence and various stats) and parrying with and against given weapon type, parrying also being physical (and similarly parametrized), but most likely automatic (inverse kinematics) as long as you don't dodge or are otherwise busy and game logic determines it's worth the effort (the attack can hurt you but is possible to be parried - for example you might not need to parry if you're wearing full plate and an enemy tries to slash you with a katana, while you might not be able to parry a giant's club swung at you and parrying might get in the way of dodging it) to shift burden of controlling an obvious action from the player and free controls.


As for HPs, the main problem with simple flat HP scale is "I can take it".
Arrow to the gut? I can take five more.
Dagger to the chest? No problem, I'm not some puny mage, I can't be killed with a single thrust.

In a HP scale you either have enough HPs that at least some weapons have no chance killing certain characters outright, or weapons are too similar for the differences to matter (because they need to have damage range stretched between the strongest PC possible HPs and less than the weakest one has) and survival becomes a matter of chance.

 

[Murk]

Why should there be a skill for doing that? You control when you want to swing full force, possibly exposing yourself and giving the target time to react and when do you want to execute weaker, but faster and more controllable attacks. You control where do you aim the attack and what type of attack do you perform. Skill can at most control parameters of your swing and stance (speed, recovery time, accuracy, weapon alignment error, weapon grip strength, how easy are you to push out of balance), availability of special attacks (but they'd better work as situational perks and generally similar to spells for casters - learned from sources or randomly available on level ups based on skill level, your intelligence and various stats) and parrying with and against given weapon type, parrying also being physical (and similarly parametrized), but most likely automatic (inverse kinematics) as long as you don't dodge or are otherwise busy and game logic determines it's worth the effort (the attack can hurt you but is possible to be parried - for example you might not need to parry if you're wearing full plate and an enemy tries to slash you with a katana, while you might not be able to parry a giant's club swung at you and parrying might get in the way of dodging it) to shift burden of controlling an obvious action from the player and free controls.

I believe you are slowly leaving the realm of "feasible" when it comes to cRPGs. I just can't imagine what you are describing ever being used well, or... at all in a cRPG without it breaking down on some level. In cRPGs character-skill should always govern whether your character succeeds and fails, so with things like "you decapitate" it should be related to some characteristic, attribute, parameter, skill or otherwise "stat" that the in-game character has. If we want to talk about some hybrid RPG-action cutting simulator then sure, but that brings us back to the previous question of whether you want an RPG or a slashing simulation.

I personally think the current system can be easily remedied with something as basic as "character is vastly overpowered to enemy, insta-kill critical, move on" as opposed to something that takes into account whether the player decides to reach out and expose themselves per swing and carry through with the swing. It may be because you're talking in terms of concepts and not constructs but I donno if it'd even translate over.

Basically, give me an example of the above that you are proposing in a mechanical form that shows how such things would be handled.

As for HPs, the main problem with simple flat HP scale is "I can take it".
Arrow to the gut? I can take five more.
Dagger to the chest? No problem, I'm not some puny mage, I can't be killed with a single thrust.

In a HP scale you either have enough HPs that at least some weapons have no chance killing certain characters outright, or weapons are too similar for the differences to matter (because they need to have damage range stretched between the strongest PC possible HPs and less than the weakest one has) and survival becomes a matter of chance.

Arrows doing 1d6 damage is not a downside of HP systems themselves but that the particular game is using a lower than reality number. If you want arrows to be nearly always insta-kill then we should focus instead on the concepts of accuracy and resistance, not necessarily the above prescribed systems.

 

[DraQ]

I believe you are slowly leaving the realm of "feasible" when it comes to cRPGs. I just can't imagine what you are describing ever being used well, or... at all in a cRPG without it breaking down on some level.

Which part do you see as problematic and how?

I wasn't refering to some hypothetical superadvanced technology. I was refering to building blocks we have now and in most cases have had for nearly a decade. A decade is roughly synonymous with eternity in computer tech.

Whatever calculations I'm proposing are already being made in similar circumstances in modern games, except with no relation to damage system. Because the devs still think with their HP pools and damage values on weapons using more potent tech for cosmetic shit - partly because they are afraid of players being confused about their weapons not displaying damage numbers or their health bar gone.

In cRPGs character-skill should always govern whether your character succeeds and fails, so with things like "you decapitate" it should be related to some characteristic, attribute, parameter, skill or otherwise "stat" that the in-game character has.

It should and it would. Except not directly and depending on circumstances.

Weapon's point/cutting edge misaligned due to character's poor skill (random, skill-based variation of weapon's orientation)? Attack will slide off the armour instead of penetrating, or at least penetrate less. Games have been capable of simulating flat angle hits bouncing off for a long time.

Significant delay due to poor weapon skill? Target will dodge or parry, throwing character off balance or disarming him (poor grip) and finishing off.
Parrying can use crude logic for selecting which animation template to use for given attack direction, then proceed using inverse kinematics for intercept and physics for handling the result. Inverse kinematics has been used for animations since at least 1998. Skyrim, to not look far, does inverse kinematics with every step any of characters make.

Significant accuracy error? Blade may get stopped by helmet or shoulderpad (we assume no gorget to make our beheading feasible at least when performed by proficient fighter) yielding no damage or relatively light injuries. Games have been using multiple collision volumes and materials per object when calculating hits since at least 1998.

From my perspective year 2000 is 12 years late already and counting.

I personally think the current system can be easily remedied with something as basic as "character is vastly overpowered to enemy, insta-kill critical, move on" as opposed to something that takes into account whether the player decides to reach out and expose themselves per swing and carry through with the swing.

Except current system doesn't need to be remedied. It's an endlessly propagated mistake, epicycles on epicycles. It needs to be scrapped.

Allowing player to decide how much power to put into attack has been in longer than bethesda's games using such feature, it was staple of quite a few FPS games and even vertical scrollers.

The idea is dead simple - you don't get to attack as fast and leave longer vulnerability window.

Arrows doing 1d6 damage is not a downside of HP systems themselves but that the particular game is using a lower than reality number. If you want arrows to be nearly always insta-kill then we should focus instead on the concepts of accuracy and resistance, not necessarily the above prescribed systems.

Re-read my post:

Yes, you can fix low lethality problem by using comparatively shallow HP pools (meaning that all characters have some chance of being offed by a single hit), but then the system simply lacks depth and detail to make it anything but a russian roulette simulator. Characters will die a lot and randomly but that's it. There can be nothing you can do about it, because the mechanics doesn't go any deeper - they will die randomly because they will have random value subtracted from HP score that's less or equal than the value's range.

 

[villain of the story]

But since you can bludgeon someone with a pommel or break their bones with a zweihander even if you can't cut through the mail,

There is no such thing as " *even if* you can't cut through mail" as if one could expect to cut through mail. You just don't cut through mail. You may penetrate it but that isn't happening with a regular sword, especially not with a big ass two-hander. A two-hander would bend like a motherfucker (because all swords are made to bend under varying degrees of resistance) with the point pressed against the links and then snap and slide at some point due to movement of opponents. Possibly the only type of sword that might penetrate mail would be estoc and its derivatives.

there should be no distinction here and all weapons should also deal damage via energy transfer term. It should depend on weapon's mass but also how much of a swing got soaked by armour and body (part soaked by armour would get partially dissipated) - for example imagine you're slicing someone with a heavy, but absurdly thin and sharp blade - despite high mass you wouldn't really deal any significant amount of blunt force trauma as it would just pass through without slowing.

I think there is a need for another stat other than health (which in itself is often ridiculous anyway) and stamina. What happens when you receive a concussion impact? You don't actually get "worn out" the same way you do as you keep exerting physical energy. But you lose focus, ability to process clearly, you can feel quite fucked up though you could also physically perform relatively well.

What is that thing that determines how well you put up with blunt trauma? It's easy to translate it to other consequences, eg. penalties to your actions but how do you arrive at there? Is it to be derived from a combination of stats like (depending on the system) Strength and Endurance/Constitution? Then how well do you recover from it or resist against losing focus and whatever? End/Cons and Willpower?

Still, what is its name?

As for damage, instead of settling for linear damage ratings, they could be rated per functions a weapon was designed to fulfill to varying degrees.

For instance, a wider, broader blade ought to cause more damage when thrust due to larger surface penetrating the body and also transfer more kinetic energy when cutting also because of the form, than other blades. It would naturally have other comparative shortcomings, eg. less flexibility and control when fencing, especially with regard to winding the sword for counter-attacks, more fatigue from using it, reduced ability to penetrate (good luck pushing a broad blade through small openings in plate), reduced ability to half-sword and whatever.

Likewise, a spear of moderate length would transfer more kinetic energy while slashing* than a sword would, in addition to the thrusting and penetration, though you can't cut with it (obviously), less flexibility compared to smaller weapons and whatthefuckever else.
(*: the way I understand it ie. swinging a weapon to hit with just the tip of the blade in which case a blade attached to the end of a pole would store and transfer more kinetic energy while swinging and upon hitting.)

The things that factor into the severity of wounds (once a blade meets the flesh) can be quantified with a reasonably authentic degree of abstraction of physics (without running a simulation). Better, it might be possible to formulate all of it without introducing a dozen special conditions per weapon type.

Obviously, it would be easier to implement something like this starting from a physical engine, but the point I was trying to make is that scaling damage with attribute should be heavily capped and that most of it should be handled by scaling armour piercing modifier instead.

Unnecessary and a waste of resources. What exactly is it that a physics engine is supposed to achieve? Calculate kinetic energies, resistance of materials, bending? Then again, various qualities and parameters of an attack are supposed to be dictated by character skill and translating that into physical parameters without a substantial amount of abstraction (which negates the need for a physical simulation in the first place) is absurdly impractical.

 

[villain of the story]

Think this as our proof of concept (lulzy mod, but it shows relatively well what I have in mind):

The game in gif is Cortex Command, and while it has a HP system, it basically only represents blood loss or equivalent - a lot of damage happens outside of it and can override it completely. Basically, character is comprised of body parts. Each body part is a physical object with mass, collision, material and so on. On top of that there are attachements that are also physical objects and usually play the role of armour. Now, getting rid of attachments and bodyparts can be accomplished in several ways.

First, there is a wound system - if something hits the object hard enough for its "sharpness" parameter to overcome object's resistance the object gets wounded or dented (visibly, at the point of impact). An object can acommodate finite number of such wounds - too much and it gibs. For bodyparts such wounds will generally result in bleeding, gradually reducing HP score of an actor.

An object can also be subjected to force overcoming it's joint strength with which its attached to actor. Then it flies off - helmets can be ripped off your head, limbs fly off in bloody arcs, etc. (of course flying bodypart or piece of armour can also be a dangerous projectile).

Finally, if an object is subjected to too much static or dynamic force (for example when propped against something or attached with high joint strength) it will be gibbed.

Losing a bodypart will usually result in direct HP damage and bleeding, but the loss will affect an actor far more directly. For example the actor will become lighter (because limbs have mass), and won't be able to use lost bodyparts, for example an actor without hands won't be able to use weapons, an actor without legs, won't be able to walk - those aren't some extra pieces of logic - having no arms simply gives you no attachment point for weapon, while having no legs means they won't be pushing against the terrain. If your head or torso gets destroyed, then you die, regardless of your HPs - chunky salsa rule (and how many systems can you honestly say to not need chunky salsa rule invoked by a human).

The system is very simple, yet it produces far more complex outcomes than any system used in an RPG aside of maybe DF.

I have even seen "migraine gun" implemented in one of the earlier versions (prior to allowing scripting effects into the engine) that simply made people's heads explode by firing a beam of alternating negative and positive mass particles at resonant frequency of their heads. It didn't do any damage, it didn't even push or pull them visibly. It's simply that you aimed it at someone's head, held the trigger down for some time and suddenly the head exploded into ludicrous gibs.

Notice that there is no doubt why a character can't move or has suddenly died in such system. Everything is instantly clear because you can see it directly.
Leave textual feedback for things you can't really show. It will be better for both text (ability to put more text when it's actually indispensable) and gameplay mechanics (not swamping player with textual feedback for stuff he could actually just see).

Yeah, Cortex Command is a 2D game and is already computationally expensive, but that's because each pixel of the terrain is a potential particle and physical object - you don't need that for raw damage system, and doing it in 3D, without particle terrain would allow you to use the computing power of your graphics card for physics engine needs. Going 3d would actually make the underlying logic simpler, because it would eliminate the need for objects to overlap and phase through each other on certain occasions (like arms, torso and projectiles colliding with them). Yeah, it's been long in development hell, but the features I'm speaking of were in even in really early builds.

Pretty fucking cool, Cortex Command, but I don't see its application to martial arts in a serious manner. It's a simple arcade game with relatively complex mechanics underneath. You could apply the basics to martial combat in only a similarly very simplistic model because simulating all those effects with polygonal models would be a different ball game. Plus, martial combat vs. ranged combat. It's not as simple as pulling collision physics.

Remember the recent procedural soft body physics demonstration in CryEngine 3? Consider that it's completely new stuff and still doesn't have a practical application plus it hasn't even matured and it doesn't look too good. Well, what you are dreaming of is nothing short of that (except at a much smaller scale, obviously) because you need to simulate friction and the bending curve of materials in weapons, armour, damage and fuck knows what else.

Then again, maybe you are right, in which case, I wish you the best of luck with developing something of the sort

But then, would there be a skill that determines the likelihood of how often you inflict one of these "large cut" types of attacks? If so, why not simplify it with something with a level or skill or stat comparison and allow for an appropriate increase in critical hit likelihood. After-all, critical hits are not "unlikely hits" but rather major ones.

Why should there be a skill for doing that?

If there is no skill to do it, how is it a Role-Playing Game any more?

You control when you want to swing full force, possibly exposing yourself and giving the target time to react and when do you want to execute weaker, but faster and more controllable attacks. You control where do you aim the attack and what type of attack do you perform. Skill can at most control parameters of your swing and stance (speed, recovery time, accuracy, weapon alignment error, weapon grip strength, how easy are you to push out of balance),

Then it's no longer an RPG.

Also, is the system supposed to read your mind to understand what you are intending so it can apply character skills accordingly? For instance, if you aimed at your enemy's head, swung your weapon (via whatever input or interface) and your blow went towards its shoulder or arm, how is the system supposed to interpret that you had originally aimed at the head?

In essence, things you can "aim" at in martial combat is quite limited. Different systems have different sections of human body to aim and attack but they all meet at common grounds as they have studied and codified the fundamentals necessary to hurt and kill your opponent based on a system with limited resources: human body. Muscles, nerves, organs, arteries, skeleton, armour. And if you want non-existing creatures, you only need to apply the same fundamental and predetermine important aiming spots. You need these spots predetermined in the system so you can order your character to aim at these spots, which is the most basic creed of RPG: character skill over player skill.

Otherwise, for your physics simulation to have any meaning at all, RPG-or-no-RPG, you will also need to simulate the full extent of human anatomy. You will need to simulate whether a sword will hit an artery and cause bleeding, cut nerves and cause a limp, cut muscles and disable a limb or strike an organ and incapacitate the opponent when it penetrates the flesh. Too Much Fucking Work That Can Easily Be Abstracted Without Any Physics(tm). Because hey, if you simulate surface and armour physics, swing your sword with much mouse-wankery-mastery and penetrate your enemy, who is to tell you penetrated him at just the right angle to cause any significant damage?

What you want sounds more like a shitty action game.

and parrying with and against given weapon type, parrying also being physical (and similarly parametrized), but most likely automatic (inverse kinematics)

This is self-contradictory. You want full manually physical simulation to aim and strike at (and potentially miss) anywhere on an enemy but don't want the same kind of control for parrying?

for example you might not need to parry if you're wearing full plate and an enemy tries to slash you with a katana,

I would hope that in your quest to simulate physics, you wouldn't negate AI so that you wouldn't end up with brain-fucking-dead enemies who decide it is a good idea to slash at a target in plate. With a fucking katana.

 

[Norfleet]

I would hope that in your quest to simulate physics, you wouldn't negate AI so that you wouldn't end up with brain-fucking-dead enemies who decide it is a good idea to slash at a target in plate. With a fucking katana.

To be fair, it's more likely that the player would make this mistake than the AI would make this mistake. The AI has an internal understanding of your ruleset and can therefore easily and automatically pick the attack that has the highest expected value. The player, on the other hand, has played too many videogames.

 

[Excommunicator]

Going to agree with Mikayel and VotS. It sounds like DraQ doesn't actually want to have an RPG but a combat simulator which relies heavily on the player's twitch reflexes to control combat. A system relying on complex physics and energy transfer requires the player not only understand medieval combat intimately but also be capable of the difficult manoeuvres required to perform the right combat moves. Neither is conducive of a good RP environment.

In fact, even in a pure combat simulator there are certain inherent restrictions on the computer interface and controls scheme that will be huge obstacles in trying to make a satisfying experience. Field of view (the biggest one of all), controlling limbs separately, no standard surround sound, fast-movement induced vertigo, the list goes on.

Do you really want a true RPG or do you simply idealise RPG ambitions of simulation?

 

[DraQ]

Pretty fucking cool, Cortex Command, but I don't see its application to martial arts in a serious manner. It's a simple arcade game with relatively complex mechanics underneath. You could apply the basics to martial combat in only a similarly very simplistic model because simulating all those effects with polygonal models would be a different ball game.

Quite possibly easier one too.

Graphics cards these days pack far more computing power than CPUs, highly parallelized computing power suitable for physical calculations. While CC has to run all its calculations on CPU, a 3D game can be using PhysX or similar library to shift this burden onto far more powerful dedicated PPU.

Plus, martial combat vs. ranged combat. It's not as simple as pulling collision physics.

It isn't. But it also isn't as simple as decrementing a fucking integer and collision volumes responsible for simulating weapons, armour, body and vital areas are far closer to ideal than a shitty integer can ever be.

Remember the recent procedural soft body physics demonstration in CryEngine 3?

No and I don't particularly care for reasons detailed below.

Consider that it's completely new stuff and still doesn't have a practical application plus it hasn't even matured and it doesn't look too good. Well, what you are dreaming of is nothing short of that (except at a much smaller scale, obviously) because you need to simulate friction and the bending curve of materials in weapons, armour, damage and fuck knows what else.

Way to misinterpret and misrepresent my points.

You don't need to get that deep into physical simulation. There is hardly much to be gained from it.
OTOH there is definitely a lot to be gained from determining whether strike encountered armour (and what kind/how much), what organs or body parts could attack hit from given position, whether the attack angle was steep enough to not glance off armour, and from getting some basic physical variables from a weapon swing for use in (localised) damage calculation or checking how well a bodypart or attachable can handle whatever is happening to it.

It allows you to avoid explicitly handling a lot of stuff - most of which you won't even think about when making your mechanics, implementing stuff otherwise only enforceable by living GM like chunky salsa rule, reducing a lot of abstraction and thus freeing yourself from dilemma of realistic russian roulette VS tedious HP attrition by giving player much finer control and feedback, and finally, integrating a lot of feedback into presentation.


You don't even need good collision simulation for that, merely detection, though simulation also adds some benefits, especially for secondary effects and improvised tactics.

But then, would there be a skill that determines the likelihood of how often you inflict one of these "large cut" types of attacks? If so, why not simplify it with something with a level or skill or stat comparison and allow for an appropriate increase in critical hit likelihood. After-all, critical hits are not "unlikely hits" but rather major ones.

Why should there be a skill for doing that?

If there is no skill to do it, how is it a Role-Playing Game any more?

So if there is no skill determining how often do you choose to slash vs stab with your knife in FO, or how often do you choose targetted vs normal attack, then FO is not an RPG?

Your logic, kind sir, is impeccable.


You simply attempt some sort of full force thrust or swing when you see fit, and there is a risk/reward tradeoff, as such attack may penetrate the target better, but will also be slower, more exposing and difficult to recover from quickly for you.

You control when you want to swing full force, possibly exposing yourself and giving the target time to react and when do you want to execute weaker, but faster and more controllable attacks. You control where do you aim the attack and what type of attack do you perform. Skill can at most control parameters of your swing and stance (speed, recovery time, accuracy, weapon alignment error, weapon grip strength, how easy are you to push out of balance),

Then it's no longer an RPG.

Fallout is not an RPG because success of a ranged attack depends on lot of circumstantial shit like lighting, attacker's injuries, partial cover (both envrionment and characters) and distance, rather than just attackers ___ Guns score.

Also, is the system supposed to read your mind to understand what you are intending so it can apply character skills accordingly? For instance, if you aimed at your enemy's head, swung your weapon (via whatever input or interface) and your blow went towards its shoulder or arm, how is the system supposed to interpret that you had originally aimed at the head?

No, system is supposed to not care about your intentions.

What use is the knowledge that you had originally aimed at head? What difference does it make? Assuming that you have targeted head correctly (or had an interface do it for you, for example if we're speaking of a TB/asynchronous interrupt based party game) system applies skill-dependent variation to the direction of your swing (influencing probability of hitting your target and possible effects of missing it) and how straight are you holding your weapon (determining armour and body penetration), skill dependent latency and so on, and let the target do its thing while waiting for the attack to connect or miss (hit something else/finish without connecting) and then applies skill dependent effects - checks for weapon's pin strength to your hand to determine whether it has been knocked out of your hand with block, parry, or your own clumsiness, determines how long and severe swing recovery you'll be put into based on the result of your attack and so on.

In essence, things you can "aim" at in martial combat is quite limited. Different systems have different sections of human body to aim and attack but they all meet at common grounds as they have studied and codified the fundamentals necessary to hurt and kill your opponent based on a system with limited resources: human body. Muscles, nerves, organs, arteries, skeleton, armour. And if you want non-existing creatures, you only need to apply the same fundamental and predetermine important aiming spots. You need these spots predetermined in the system so you can order your character to aim at these spots, which is the most basic creed of RPG: character skill over player skill.

It's more important to know what have you hit than what have you aimed at.
Besides, it's also important to know what can you aim at and that is very circumstantial - depending on size, stance, physical structure, relative positioning, weapon used and so on.

Manually accounting for all the combinations would be fucking madness. Let's say you're battling a dragon. How much extra work will it take to specify all weapons that can be potentially effective (you won't reach anything important with a fucking dagger), and all bodyparts you can potentially reach depending on your position within spherical 360 degrees angle relative to the dragon AND both yours and dragon's stance? How many obvious flaws will such humongous litany of exceptions feature?
And then you repeat it for all creatures you have in game.
Have size affecting spells? Too bad, repeat it for all creatures at all sizes they can make.
Opt for simple 2D playfield without free positioning? Too bad, you'll have to check for adjacent climbable scenery, cover and who knows what else, and make a table some lovecraftian clusterfuck of derp allowing program to look up all the required info.

Oh, you can also make a simple little game where you can take down a dragon by repeatedly stabbing it in the toe with a dinky dagger 'till its over 9000 HPs run out.
Sounds fucking orgasmic.

Me? I'd rather make some volumes representing various areas and armour (quite a few rather reusable), attach them to base model and it's skeleton using a WYSIWYG editor, assign material variables and proceed to the next creature as they are made and animated, knowing that with those resources and basic collision checking as most of its ruleset the system will handle every encounter and situation you may or may not dream of far better than if you tried to do so yourself.


Also, the creed of RPG isn't "character skill over player's skill" it's simply implementing the former and preventing the latter from overriding it. Within those bounds everything goes.

Otherwise, for your physics simulation to have any meaning at all, RPG-or-no-RPG, you will also need to simulate the full extent of human anatomy. You will need to simulate whether a sword will hit an artery and cause bleeding, cut nerves and cause a limp, cut muscles and disable a limb or strike an organ and incapacitate the opponent when it penetrates the flesh.

Maybe I have to simulate atomic structure of the matter as well, hmm?

Why does it have to be all or nothing? I'll shift the abstraction down as long as it benefits the gameplay. The degree I described benefits the gameplay for the reasons I expounded upon. Going deeper would bring marginal improvement at heavy expenses. Not going as deep creates multiple issues that can be seen in cRPGs and other genres.

Too Much Fucking Work That Can Easily Be Abstracted Without Any Physics(tm).

Sure. You can abstract everything.

Technically you can even make a game consisting of chargen, and then making a roll at difficulty parametrized by the build you've made to decide the outcome of entire game. It would be awesome too - not only would it feature no pesky player skill, but it could be replayed seven times in under an hour.

Weighted coinflip - the best RPG ever.

What you want sounds more like a shitty action game.

As opposed to what? All the non-shitty ones? Action game and RPG doesn't need different abstractions for combat. They have similar scope measured in the amount of active combatants, unlike, say a strategy.
The main difference, apart from non-combat options, is that in an RPG you have to incorporate ways stats influence gameplay. By stuffing them somewhere between interface and mechanics for actions and into mechanics itself for reactive/passive stuff (otherwise it just gets weird - see TES). You can also benefit from less actiony interface, but that depends on the type of RPG you're making - in a single character one it would be spurious and even counterproductive.

This is self-contradictory. You want full manually physical simulation to aim and strike at (and potentially miss) anywhere on an enemy but don't want the same kind of control for parrying?

Attacks are active, parries are reactive.

You can attack anywhere at any time. You can only parry when you're being attacked and at enemy weapon/weapon arm.

There is much freedom involved in attack, but little freedom involved in parrying. Optimal attack may be not unambiguously defined in a well developed system, optimal parry, OTOH will largely depend on immediate situation. Other than that, in an RT, directly controlled game player only has so much keys on their keyboard and buttons on their mouse, that wasting them on an "obvious" action would be stupid, in a TB game your parries are not even during your turn.

Parries can work perfectly well as automatic default action - if you're not dodging and attack isn't of negligible risk should it hit you, then you parry.

 

[Damned Registrations]

Well, anything abstracted is apparently always abstracted in the worst way ever, so it's just hopeless anyways. Dagger attack you survived? Must have hit you in the heart. Dagger attack that killed you? Must have hit you in the toe.

May as well give up and go stab some people in real life, only that is realistic enough. Even if you fix this the fact that the price of a night at the inn is abstracted wrong will ruin games anyways. It'll be 5 gold instead of 1, even though it's a total shithole and the economy wouldn't allow that. Then Draq will have to ragequit and play something else. It's impossible to just change the fluff instead of the crunch.