Yurch's antics reach a new level!

[Tiffy]

RAv2 has a variable amount of damage each round does. This is to stop the I've got an AKMSU and can kill him with 2 shots, hes got an M16 and takes 3 shots to kill me so I'll always win an exchange of fire assuming we both hit with all round and no head hits are involved,thing.

You can't be sure any more. Yurch didn't go as far as some wanted though. I'd like to see any weapon have a chance of a one shot incapacitation and not have that dependant on hitting the top 15% of the model. That would bring the damage (and the players behaviour) more in-line with real life......

Still you can't plesae all the people, all the time

 

[keihaswarrior]

Random elements should be kept to a minimum. There is no good reason to make things more random. Bullets wounds in real life are not random, they are based on physics and biology and can be predicted. INF lacks the code to do these predictions, so we shouldn't try to fool ourselves into thinking that random damage will somehow more closely simulate real life -- Because it doesn't.
It is random, and therefore does not simulate anything at all, just chaos.

I think the random damage right now is only OK. Further versions of RA should focus on better hit detection and making the bullet velocity and damage correlate.

 

[Keganator]

IIRC, there was a small random chance to instigate this random damage, too.

 

[Tiffy]

Until there is a way to detect where a model is hit, random damage adjustments simulate very nicely the differing effects of shooting someone. Also as Terminal Ballistics is still very much in its infancy, predicting exactly what will happen when someone is shot isn't possible. Effects can be calculated based on a probability of there happening, but as soon as that word 'probability' gets into the equation your done to random effects.

 

[(SDS)benmcl]

I agree that random events should be kept to a minimum but right now it is the only way to simulate the type of damage that a bullet does to the body. I agree that you could survive a head shot but you not be in any condition to do anything and there for black screen and wait for next round or respawn.

Hits to the torso or even arms and legs can cause bullets to do weird things...

Actually as I type this I realize more and more that if a player that is "dead" is actauly considered ineffective and not necessarly dead that I have to agree more and more with KW. Sure if a player is hit with a shot that they can easily survive but they are no longer effective.

Better hit detection would of course help but it also depends what players consider INF is simulating.

Edit: Thank you Tiffy. You explained it better, clearer and in english much better than I did. 12:30 pm and I should head to bed.

 

[keihaswarrior]

It wouldn't be very hard to calculate the distance the bullet traveled before hitting the player. From there simply input the distance into an equation with the ballistics coefficient to get the bullet's speed. The damage the bullet does is directly proportional to the energy (speed and mass) of the bullet.

For hit detection, UT can detect head, torso, legs, arms. That is sufficient until UT2k3 where you can have hitboxs for all the major bones, arteries, and vital organs.

 

[Shrap]

I'm looking forward your mutie keihas, when will it be ready ?

 

[keihaswarrior]

As soon as I figure out Unreal script and someone gives me all the hard data for the weapons.

--I think yurch could manage to have the bullet velocity be proportional to damage. (especially since the ballistics are already in 2.9).

 

[RAZZ]

Getting a proper collision setup for the body alone would do wonders I believe.

For terminal ballistics, Im with tiffy on this one.
Personaly, I believe most people simply wont care about it during the fight.
Random is simple and quick, and for now its indestinguishable.

 

[Freon]

For hit detection, UT can detect head, torso, legs, arms. That is sufficient until UT2k3 where you can have hitboxs for all the major bones, arteries, and vital organs.

UT CAN'T detect head, torso or anything else. collisions are based on a cylindre. depending on the animation played, there is no way you can tell what part was hit. all we can do is check the upper part of the cylindre for head shots, and it's quite fucked up already

 

[Cleeus[JgKdo]]

UT CAN'T detect head, torso or anything else

Freon, that's not 100% right.
UT can, although it means to rape the engine The men from that zombie mutator gave their zombies an additional collision cylinder for the head which is much more accurate that the default headshot detection. But the effort is very high.
For the arms, legs and torso detection, most people who tried to realise this used a calculation of the hit-angle and coordinates and then tried to guess where the bullet went. Unless someone finds a very smart way to calculate the "virtual" collision inside the cylinder, this is also very inaccurate and may lead to strange effects like shooting the arms from the side but hitting the torso instead or the other way around.

 

[Tiffy]

There's no point in trying to take bullet drag and hence loss of velocity into account. A 5.56mm NATO round will deliver over 80j out to a range of over 1km. 80j is the amount of energy weapons are designed to to incapacitate a target and if you can hit a man sized target at 1km range with an M16 - have you thought of entering the olympics?

 

[yurch]

It wouldn't be very hard to calculate the distance the bullet traveled before hitting the player. From there simply input the distance into an equation with the ballistics coefficient to get the bullet's speed. The damage the bullet does is directly proportional to the energy (speed and mass) of the bullet.

Or you could just use the speed that the bullet is going in, which is already provided by the inf/ra ballistics.

Damage = Default.Damage * velocity / MaxSpeed * velocity / MaxSpeed * 1.35;

RAv2 uses a scalar 'damage' for the bullet's speed - it adds a 35% increase to all bullet damage and scales it down as the bullet loses energy, IE as a relation of the square of velocity. It's only really noticeable for quicker slowing rounds like shotgun pellets.

It still won't save you, just increases your chances of survival when this is executed.

if ( FRand() < 0.12 )
Victim.Health -= 101;
if ( FRand() < 0.6 )
Victim.Health += ( ActualDamage * ( FRand() * 0.9 )) - (ActualDamage * ( FRand() * 0.8 )) - (( 100 - Victim.Health) * 0.25);

The "random" damage has a 12% chance of instant 'death' for a player for any bullet hit regardless of damage, (note - players don't have +100 health, unlike the 'cows') and a 60% chance of modification from thier ordinary damage, which is on a 'random' scale that is skewed towards higher amounts of damage.

Note that these settings are just the tweaked out versions that got the least complaints. I want to add more locational factors later, if not just using how close the shot was to center mass as a multiplier.

 

[keihaswarrior]

Note that these settings are just the tweaked out versions that got the least complaints. I want to add more locational factors later, if not just using how close the shot was to center mass as a multiplier.

That sounds like a good idea to me. If you were able to make headshots instant death and other shots would scale the damage based on distance from center.

-One other thing to try and do would be to make the collision somehow smaller for the head (if that is somehow possible). Since right now, bullets that pass over your shoulder will register as headshots.

 

[Freon]

Freon, that's not 100% right.
The men from that zombie mutator gave their zombies an additional collision cylinder for the head

hmm i already tried it. but i don't think it worked that well, on some animation, the zombie bent over, and then the head collision was ****ed up. that's why i say you need to know the head position for each frame, and the inf models have waaay to many frames for that

 

[Ice]

Bump.

Felt this thread slipping. Where is yurch, and will he do any sort of mutators for INF after 2.9?

 

[Keganator]

I'm sure he will. The question is, will he do any before 2.9?

 

[RAZZ]

The thing about shrinking the head cylinder made me think of something controversial to consider.
Why not shrink the entire collision cylinder to the width of the head?
Probly wont work depending on the body position, and could cause more harm than good. But in a way, its accurate.
My thoughts is this covers center mass and vunerable spots (drawing a cylinder around your head, and extending it down to about knee height) while leaving out most of the questionable areas.

If a hit is scored, its likely to be in a lethal spot.

 

[keihaswarrior]

There's no point in trying to take bullet drag and hence loss of velocity into account. A 5.56mm NATO round will deliver over 80j out to a range of over 1km. 80j is the amount of energy weapons are designed to to incapacitate a target and if you can hit a man sized target at 1km range with an M16 - have you thought of entering the olympics?

First of all, the 5.56 round loses considerable wounding power past 200m because it does not fragment much at all past 200m.
Secondly, that is only concerning the assault rifles, for the smgs and pistols there would be a very noticable drop in wounding power. Therefore, it would be a good thing to add into ra.

 

[spm1138]

RAZZ, that sounds like something that could very easily become exploitable.

Even if you're only exposing your arm to fire you could still get hit in an artery or something.

I think the removal of the health-bar altogether and having n chance of being wounded, n chance of being killed and n chance of being fine when you get shot with a given weapon at a given range would be a better way to go.

The human body is too complex a thing to reduce to a health bar.