Cones

[Kitty.cat]

I was talking with BTH today about the super-soldiers in INF. The ones that are kinda weak in the endurance and stamina levels, but must have spent their entire lives in shooting ranges. Basically, we were commenting how INF has the basis for a great shoot-out type of game, but the thing holding it back is how every gun is perfectly accurate. Well, I know RealAim has helped this by adding in artificial "human error" (AKA twitching you can't entirely cease) into the spectrum but there have been a lot of complains about how big of a pansy the INF soldier becomes with that much nervous energy. So, I want to throw this idea out for comments/changes before I write up a real proposal to the IMT.

What if, rather than having so much twitching, we keep the aim as it is, but also add a gun-respective cone effect to each gun as it should have in real life. For those of you who don't know what I'm talking about, if you study ballistics, guns don't have a perfectly straight path when they fly (in relation to the barrel, that is.) They kind of go through this cone of effect before they hit (or miss) their target. The better the rifling on the gun, the smaller cone the gun has and the more accurate it is at certain distances. For example, pistols (save the Mk23) are generally not that precise at 100m+ distances while high powered rifles have effective ranges up to a mile or so. I think the concept of firefights would add a very interesting aspect to gameplay if it could be implimented in a way the classic INFer could still appreciate.

Edit: I hunted down a diagram of the "wobble" of a flying bullet.

Edit 2: I think I should clarify for those of you who are confused. Guns don't really "cone" in real life, but it's the best way we could really program it into UT for simulated ballistics. If we wanted to get REALLY technical we could even include the fact that a bullet falls (*gasp!*) over time. Y'know, gravity. But for now, I think a cone effect would be pretty cool, as long as we did it tastefully and realistically.

Edit 3: Not all bullets fly exactly like that diagram. It seemed like a "duh" to me but I already got a comment about it. That diagram is just an EXAMPLE I found off a ballistic forensics site.

 

[Nukeproof]

This -weapon dependant - cone effect does already exist in vanilla 2.9 (and before I assume). Go shootingrange and shoot some, you'll notice

 

[Kitty.cat]

I have noticed. But being able to snipe with a beretta leaves something more to be desired. If you don't know what I'm talking about, watch Dillinger on a long range map

 

[Keganator]

Falling ballistics is implemented ingame. However, the hitboxes are such that it usually doesn't matter.

 

[Kitty.cat]

(Because he won't post, himself)

 BTH says: (11:20:13 AM)
btw, nuke is right, with the robar, in the 500metre range, ive had my sight on the target and shot, but never actually hit it, the dust from the shot appears higher and slightly to the right, where its too high to compensate

 

[yurch]

You're talking about MOA accuracy of the weapons. Inf has a small bit of this already.

Human error is far larger, though. Especially for a stockless weapon like a pistol.

 

[Vega-don]

(Because he won't post, himself)

 BTH says: (11:20:13 AM)
btw, nuke is right, with the robar, in the 500metre range, ive had my sight on the target and shot, but never actually hit it, the dust from the shot appears higher and slightly to the right, where its too high to compensate

the ballistic isnt the same in the shooting range than on other maps , i dont know why. especialy bullet drop.

 

[Logan6]

Yep. There's already MOA in the game. Its because the hitboxes are so large that you can hit every time. Well, almost. Try hitting opposing enemy in Road to Kandahar at full range with the Robar. Its fun.

Weapons with .5 MOA are very accurate. Especially when you take time to aim and control you breathing. Im not a fan of RAV3 cause I think it has way too much twitch, but I agree ( from experience ) that you cannot completely hold a scoped rifle dead still for that long ( as in game ). Toad had a mutator out that got close to a real scope but still just wasn't right.

 

[Mercie]

I'd rather get new weapons than have the IMT be bothered with long distance sniper in such a case.

 

[randomas]

Erm how many maps actually played online have distances that warrant that kind of tweaking? The "grid effect" alone is good reason enough not to ... Then again if you wan't to force the pistols into useless mode at over 30 m go ahead!

The best sniping effect I remember is Deus-ex's woble ... I loved it!

 

[keihaswarrior]

MOA accuracy and Cone fire are different things. One is realistic, the other is lame lame lame lame.

 

[geogob]

MOA accuracy and Cone fire are different things. One is realistic, the other is lame lame lame lame.

The accuracy of a rifle is a measure of the standard deviation of the bullet from the ideal trajectory. In game "cone fire" is an implementation of this standard deviation. I do not see what's lame... it's a very good first order implementation.

Of course, implementing a 5 degree standard deviation when the actual accuracy is 1 MOA is indeed lame.


When talking about "cone fire" people tend to mix things up. You need to distinguish the error caused by the rifle design/quality/etc, which defines the rifle accuracy, with the "handling" induced error. If a rifle, on a ballistic test stand can achieve 1 MOA standard deviation from the target, it doesn't mean that a soldier in the field (considering fatigue, stress, position, etc.) can achieve 1 MOA standard deviation.

You really need to see the so called cone of fire as a first degree implementation of process describing the random aiming error. There is really nothing lame in this. That is when it is not exagerated.

edit: and for the love of god... rifles have a caracteristic accuracy, not a MOA or MOA accuracy. MOA = unit of angle...

edit: well. basically what yurch said. Nice 1 line to say all I say here

 

[(un)l33t_1]

the AKM is far too accurate. 'nuff said

 

[(un)l33t_1]

the AKM is far too accurate. 'nuff said

 

[cracwhore]

I beg the mods - again - please sticky my 'Forum Tutorial' (featuring 'Ed' the Edit button).

 

[geogob]

I beg the mods - again - please sticky my 'Forum Tutorial' (featuring 'Ed' the Edit button).

I don't remember it featuring the forum bug where you get double posts... ... ...

 

[Kitty.cat]

The accuracy of a rifle is a measure of the standard deviation of the bullet from the ideal trajectory. In game "cone fire" is an implementation of this standard deviation. I do not see what's lame... it's a very good first order implementation.

Thank you. It's exactly how I feel. And I also feel it wouldn't be all that hard to impliment into INF. As long as we didn't exaggerate it, like you said. But even assuming the soldier is in fresh physical condition, there will still be some degree of gun inaccuracy.

And by the way, the IMT works on quite a few things other than just guns, you know. And some of those things, *cough*RT*clears throat*, are more important. I'd like to see good shootout quality with good hitboxes and good realistic gunfire before I saw more guns. The guns are what you play with once you get that realism out there

 

[yurch]

When talking about "cone fire" people tend to mix things up.

When talking about games, "cone fire" is the classic term for forced inaccuracy due to handling, made part of the common language by it's ever-so-infamous implementation in Counterstrike. I refer to the inaccuracy by mechanical physical limitations of the weapon as MOA for this reason, although I'm aware that it's a measurement.

 

[Kitty.cat]

So then I may as well ask Yurch, since I've already caught his attention. Is what I'm asking A: Reasonable and B: Easy to impliment

 

[Nukeproof]

It is.

Instead of determining one value for random missalingment you generate multiple of smaller range and just add them to final missalignment. This way you get a Gauss'ian normal distribution.

(i.e. instead of one dice [1-6] you flip 5 coins [0-1] sum it up and add +1... still Range of 1-6 but values 3 and 4 way more likely)

But it'll take a statistic expert to tell the difference in game....