Ballistics inaccurate - much more than reality?

[zOldBulldog]

In Authentic, I just took a shot with a Mosin Nagaint, aiming at the top of an armored vehicle driver slot, from about 30m. At that distance the bullet drop should be virtually nothing. And yet it hit below the slot.

Puzzled, I tried it a few times. I ended up having to shoot over the slot (not just inside at the upper end) in order to hit inside. This is just plain wrong. No rifle ever built has any noticeable bullet drop at just 30m. It is especially wrong in a mode called "Authentic".

If anybody wants to test this... it is in the Abrunza Monastery map, following the road to the right from the spawn. There is a rock at the turn, bush before, and armored vehicle just after the turn. You can take the shot from the edge of the bush aiming left of the rock.

[TJD]

Set the Elevation to 100m. Let it be your default for all shots <200m.

[HalloweenWeed (US)]

There is also something called the "spread." The spread increases as difficulty increases. The spread is increased by: Barrel warpage (sometimes by uneven heating/cooling), dirt/residue in the barrel, bullet tumble (virtually non-existent at that range), and bullet asymmetry. This is why snipers are known to carefully check their bullets for abnormalities, and sometimes carefully file them off. Even the thickness of a bullet jacket varying around the axis can cause the bullet to travel in a spiral just a bit, due to imbalance (the materials are different densities). Bullet tumble also causes the bullet to move in a spiral.

[zOldBulldog]

Originally posted by HalloweenWeed (US):

There is also something called the "spread." The spread increases as difficulty increases. The spread is increased by: Barrel warpage (sometimes by uneven heating/cooling), dirt/residue in the barrel, bullet tumble (virtually non-existent at that range), and bullet asymmetry. This is why snipers are known to carefully check their bullets for abnormalities, and sometimes carefully file them off. Even the thickness of a bullet jacket varying around the axis can cause the bullet to travel in a spiral just a bit, due to imbalance (the materials are different densities). Bullet tumble also causes the bullet to move in a spiral.

I understand those things, and it could be a neat rationalization. But what I saw is MUCH more extreme than any of those factors can account for.

Any real life gun that has well over 10cm of bullet drop (or wind deviation) at 30m would either be sent for serious repair or most likely to be melted down.

[HalloweenWeed (US)]

While knocked down (bleeding out), recently I shot an enemy solder three times in the face at <1m with a pistol and he didn't even flinch. The game is not perfect, but regardless it's my favorite.

[Jurassic Fart 1]

I stopped trying to pay attention to ballistic mechanics in games a LONG time ago, being IRL qualified with a number of small arms. I have YET to play a game that gives a passable representation of ballistics. There's no way to judge wind, or measure temperature and humidity, and in this game specifically, suppressed-fire weapons behave identically to non-silenced weapons, which is REALLY wrong.

It's too irritating and frustrating trying to figure out what was actually inside the heads of the coders that wrote the subroutines governing ballistic mechanics, especially considering that, in all likelihood, most of them have ZERO training and experience in small arms proficiency.

[TJD]

Well, I wouldn't be so hard on Rebellion. They didn't set out to create a ballistics simulator. Of course the ballistics aren't accurate. It's a game, but within those limits it does a very credible and I think clever job of representing the weapons of the time. And yes, fwiw, I know whereof I speak. I am personally familiar with most of these weapons. But pulling an actual trigger and clicking a mouse button are two very different realms of experience, and I think it's a mistake to expect them to overlap very closely.

[zOldBulldog]

Originally posted by TJD:

Well, I wouldn't be so hard on Rebellion. They didn't set out to create a ballistics simulator. Of course the ballistics aren't accurate. It's a game, but within those limits it does a very credible and I think clever job of representing the weapons of the time. And yes, fwiw, I know whereof I speak. I am personally familiar with most of these weapons. But pulling an actual trigger and clicking a mouse button are two very different realms of experience, and I think it's a mistake to expect them to overlap very closely.

Sorry, but I must disagree about being tolerant of bad ballistics, especially on modes that Rebellion calls "Authentic". The moment you use that name you set a high expectation.

And a high expectation isn't even needed. The formulas for ballistics are quite simple and any 2nd year college student in sciences knows them. If not, I recommend buying a copy of the Feynman Lectures on Physics (hopefully I remembered the title correctly, and I believe Mechanics are covered in volume 1).

Perhaps Rebellion set a much higher gravity and wind impact than reality to compensate for the maps being so small. Of course, having cross wind indoors or in a street between buildings is ridiculous and just plain lazy.

I seriously hope they do better in their next game. It is not even close to hard to do.

[TJD]

Originally posted by zOldBulldog:

, I recommend buying a copy of the Feynman Lectures on Physics (hopefully I remembered the title correctly, and I believe Mechanics are covered in volume 1).

lol

[zOldBulldog]

Originally posted by TJD:

Originally posted by zOldBulldog:

, I recommend buying a copy of the Feynman Lectures on Physics (hopefully I remembered the title correctly, and I believe Mechanics are covered in volume 1).

lol

http://www.feynmanlectures.caltech.edu

Memory still good after all these years :)

[TJD]

Not suggesting your memory is inaccurate. You miss the point entirely.

[Jurassic Fart 1]

For the truly hardcore critics of Sniper Elite ballistics mechanics, a different kind of "game" is needed: a combat simulator accurate enough to begin training marksmen.

Back at the turn of the century, Bohemia Interactive released Operation Flashpoint: Cold War Crisis. As development continued, BI became threatened by looming bankruptcy, but they were saved by the US Marines, who were attracted by their then-ubiquitous and code-heavy demos. One thing led to another, and the end result was a USMC combat simulator.

It's not "easy" making a decent simulator complex enough and cheap enough to garner lucrative profits from a significant share of the gaming market. Make a game accurate and realistic enough, and you've priced yourself out of making the bucks. You haven't made a game, you've made a simulator, one that only military services and PMCs might be interested in.

Sometimes I think that only people with actual military training and experience AND a strong background in application development and publication can adequately comprehend the challenges presented by such a proposed task.

It is NOT easy.

[zOldBulldog]

As a programmer with origibal training in physics I completely understand the physics behind ballistics and what it takes to code it.

What I am saying is that the formulas are very simple.

Any college student that has taken the 1st calculus based physics course understands those formulas.

It is very easy to calculate bullet drop based on muzzle velocity and gravity. The impact of wind of a certain velocity and of a certain density is only slightly more complex but still quite easy to program, even easier as you can make some simplifications.

Avoiding applying wind indoors or cross winds between buildings is a simple thing too.

Frankly the only "hard" that I can think of is of the kind I saw in a recent random survey to find out how many young people knew how to use a can opener... and over 50% percent failed. That is not a matter of being hard, it is a matter of simple ignorance and easily overcone.

That is why I can only assume that the unrealistic physics come from other priorities, I simply cannot imagine Rebellion not being able to hire someone with the know-how.

[Terminator Tony]

Assault rifles are zeroed at 300 metres I think. But try to shoot anything in Sniper 4 over 30 metres away. Forget headshots, be glad if you hit anything at all. Just take a SMG like this. https://www.youtube.com/watch?v=LxHeCyqbHk8 That gun was used in WWII. Stable, little recoil, even in full auto. Sniper Elite 4s gun mechanics are so way off it might as well be a cartoon game. Major letdown.

[HalloweenWeed (US)]

Originally posted by zOldBulldog:

As a programmer with origibal training in physics I completely understand the physics behind ballistics and what it takes to code it.

What I am saying is that the formulas are very simple.

Any college student that has taken the 1st calculus based physics course understands those formulas.

It is very easy to calculate bullet drop based on muzzle velocity and gravity. The impact of wind of a certain velocity and of a certain density is only slightly more complex but still quite easy to program, even easier as you can make some simplifications.

Avoiding applying wind indoors or cross winds between buildings is a simple thing too.

Frankly the only "hard" that I can think of is of the kind I saw in a recent random survey to find out how many young people knew how to use a can opener... and over 50% percent failed. That is not a matter of being hard, it is a matter of simple ignorance and easily overcone.

That is why I can only assume that the unrealistic physics come from other priorities, I simply cannot imagine Rebellion not being able to hire someone with the know-how.

Yes, agreed. That said, it may not be a matter of being "able to program it," it may be a performance issue. When you have 20 bullets in the air simultaneously, what does it take to calculate and render that, along with the already burdened graphics/physics load on the CPU & GPU & interfaces between them. And then there's the MP interface with the server and/or other players. Given that some players are going to have minimum requirement rigs, it may be an issue of how much of a performance hit this causes. And as time goes on, the computers get faster, so perhaps on the next iteration they could do this (I hope).

Also, there's the issue of the "bullet trace lines." In real-life, they don't do that. Perhaps in a loadout for a machine gun every 5th bullet may be a tracer, but not for a sniper's rifle. Tracers show the enemy where you're at too (particularly when not full daylight). So the trace lines that SOMETIMES show (don't understand the logic there) are unrealistic, however I am glad they are there, I don't EVER use aim assist.

[HalloweenWeed (US)]

Originally posted by Terminator Tony:

try to shoot anything in Sniper 4 over 30 metres away. Forget headshots, be glad if you hit anything at all.

I always play on Authentic, and even if i didn't, I have aim assist turned off. And 70% of the time when I shoot at 1 to 150m I hit a target area the size of the slot in an armored car turret on the first shot (you can ask the players that I play with, they would confirm this), without even setting my scope range. This is what a year of practice gets you.