CMRT: Observations from the Firing Range

[Migo441]

The Soviet SMGs II thread (link here) started by Poesel piqued my interest.  The initial question was the perceived supremacy of Soviet SMG troops:  Is a real phenomenon?  Is it automatic weapons generally?  At what range do regular rifles gain an advantage over the SMGs?  Etc...

 

I ran a few tests and showed my initial results in the linked thread.  A few members made suggestions about how to structure valuable tests.

 

- c3k said to limit variables. 

- Sgt Joch said to place the targets on pavement to eliminate micro-cover.

- Poesel said to only test one thing at a time.  (i.e., if you're interested in testing accuracy, don't worry about ROF)

 

I took their advice and went back to the drawing board.  I built a range with five lanes for each of five range groupings: 60, 120, 180, 240, and 300m. (1)  The lanes are 11 tiles wide and divided by high walls.  The bulk of the terrain was the default editor "open ground" but I lowered the terrain between the shooters and targets by two meters to reduce / eliminate any LOS blockage figuring that none of the grass types would be six feet tall!  The firing lines are placed on a step at elevation setting 22.  All shooters are Regular.

 

 

The floor of the range is elevation 20 and the targets, to match the shooters, are placed at elevation 22 in the middle of three block wide stretch of pavement which slopes from elevation 21 to 23.  To emphasize the point, shooters and targets are at the same elevation. (2)

 

 

For the targets, I used five 2-man sniper teams: one in the middle of the range directly across from the shooter and two more arrayed on either side with an action spot gap separating each team.  All targets are Fanatic and given short firing arcs to keep the shooters alive. (3) Yes, it was more tedium to use five two-man teams in the place of a potential single 9 or 10 man squad.  The method to the maddness was that I wanted to focus the test on simple accuracy.  I worried that an entire squad placed in a single action spot would provide a target density that would disporoportionally benefit automatic weapons. (4)  As JasonC put it in the SMG II thread:

 

c3k - sure, one of the reasons I wanted realistic examples is that SMGs might be favored by massed targets, lack of cover, and movement, as all things that can be benefited by spray and pray and hits on targets *other* than the intended one, especially at close range.  Whereas a longer range shot at a stationary and small / single unit target in good cover should bring actual accuracy to the forefront.

 

So five two-man targets it was.  Point-of-emphasis:  I am (I hope it is obvious) NOT claiming this represents typical firing conditions.  These targets tend to lay prone and are therfore smaller than a standing or kneeling target.  On the other hand, they are stationary and marooned on a stretch of pavement without so much as a blade of grass to shelter behind.

 

I marched to the range and fired a combined 261,761 rounds from eight different weapon types.  Does "over a quarter million rounds" sound more impressive?  I tracked rounds fired per range per type along with the casualties caused. (5)  This is what I found:

 

 

 

Now remember, this isn't a promise or prediction that you should expect to fire N bullets at range X with weapon Y to achieve Z casualties in CMRT battle conditions.  I'm simply reporting my results for the given sample size under the admittedly artificial conditions I laid out.  The interesting thing to me is to see how the accuracy of a given weapon degrades with range and/or how different weapons compare at a given range.  All tests were performed on the exact same range and I have the stats per lane so we could see if a certain lane seemed to perform poorly over multiple weapons for example.

 

For those who prefer the raw tabular format:

 

 

Note that I'm not advancing 'Composite' accurracy here as a meaningful metric, but I just threw that in there so that I had something to sort by.  I'm not sure how interesting or useful anybody will find any of this.  My two big reactions were as follows:

 

1.)   My first-hand experience with shooting and ballistics is dominated by Basic Training at Ft. Leonard Wood in the Summer of '90.  I qualified Sharpshooter (the mid-tier), hit a few pop-ups at 300m, and by now have surely forgotten most of what I learned although I could probably still field strip an M-16 by rote.

 

So I'm nobody's idea of a shooting expert but I was expecting the accuracy to degrade expotientially with range.  But maybe that's not how it works or maybe that's only true at longer ranges and 300m and below is within quite reasonable ranges for these weapons and (mostly) iron sights.  I simply don't know.

 

In particular, it's striking how the accuracy at 180m is much the same as the accuracy at 120m for several weapons.  The range jumps by 50%, but the accuracy (judged by average rounds per casualty) degrades by only 21.44%, 3.08%, 3.53%, and 4.12% for the LMG42, DPM, PPSh/PPS-43, and MP-40 respectively.  Not sure what's going on there or if anything is going on.  Statistical noise?  I dunno.

 

The related point is how the accuracy of the weapon types relate to each other.  The snipers are best, then the bolt-actions, then the LMG, and finally the sub-machine guns.  Each type is grouped together in its expected place along the accuracy continuum which feels good.  

 

At 180m, it takes ~ 84 Mosin-Nagant rounds or ~ 126 PPS-43 rounds to cause a casualty in testing...  So the PPS-43 is 50% less accurate but it's trivial for a PPS-43 shooter to exceed the rate-of-fire of the Mosin-Nagant by more than 50%.  I know I wasn't supposed to consider ROF, but you see the point here.  Just rough figures: Maybe a determined and steady Mosin-Nagant shooter gets off 10 rounds in a minute.  In the same timeframe, as soon as a shooter behind a PPS-43 exceeds 15 shots he's causing more casualties on average at 180m (according to the test results).

 

2.)  What's going on with the German Snipers?  Why are they performing so much better than the Soviet Snipers? 

 

I believe I found the answer to this one hiding in plain sight.  Yes, the German snipers performed better than the Soviet snipers but my working theory is because the German sniper test shooters ended up with a freakishly large proportion of designated 'Marksmen' rather than the basic 'Soldiers.'

 

For the sniper shooters of both sides, I used Sniper Teams at 50% headcount in attempt to limit the teams to just guys with scoped rifles and avoid the SMG-toting buddies.  However, even at the 50%, I found a couple German teams still had the MP-40 guy so I placed those at the 300m range knowing they wouldn't fire.  (See thread about 200m hard range cut-off for SMGs here.)  So those teams would have had improved spotting abilities relative to the singletons but I didn't worry about that as I only cared about the resulting aimed shots and not if the snipers had spotting help.

 

As I placed the snipers of both sides I idly noted (mentally) that some were Marksman and some were Soldier but I didn't think much of it.  However, I later had the impulse to mark which lanes contained Marksman and it was then that the Germans' relative overperformance made sense.

 

 

The lanes with the m notation off to the side contained Marksman snipers and the ones without contained Soldier snipers.  The # symbols designate K98 armament; the balance were armed with G43.  (Digression, were G43 the predominant Heer sniper rifles?)   You see I ended up with only four non-Marksman among 24 active lanes. (240m Lane 1 is the NULL lane referenced in footnote 3.)  As you see, the accuracy of the non-Marksman snipers is significantly worse.  As for the Soviets:

 

 

You see they only ended up with two Marksman and they are the best two of their grouping although the effect is less discernable at 60m.  Ironically, I put those two in the 60m grouping on purpose as I noticed they were carrying only 55 rounds per man compared with 150 rounds for the bulk of the Soviet snipers.  My thought was that I wanted them at a closer range so they wouldn't run out of ammo and, at that time, I didn't make the connection between the Marksman designation and the lower ammo count.

 

Although the sample size is small, we see that the few non-Marksman Germans perform comparably with the non-Marksman Soviets at a given range.  So that riddle is tenatively solved.  The German Sniper "results" should then be accompanied with a big asterisk at the moment and are subject to revision.  But I thought I'd still show what I have for the moment to demonstrate the seeming weight of the Marksman-factor.  Plus, the Snipers are such outliers in the scheme of these results.  Even the regular Soldier Soviet snipers are 6 and 11 times more accurate than the Mosin-Nagant at 240m and 300m respectively.  

 

(1) More specifically, the number of action spots between the deployed shooters and targets were as follows: 7 - 56m, 15 - 120m, 22 - 176m, 30 - 240m, 37 - 296m.  (Yes, I know you guys can do arithmetic.)

 

(2) I know it's a more challenging shot when there is an elevation difference between shooter and target.  That's not from experience, that's from reading Stephen Hunter novels.

 

(3) There was one exception.  In Lane 1 at 240m in the German Sniper test, I missed a cover arc on one of the Soviet sniper targets.  The result was a shooter that got shot...  So I threw out that particular lane and then ran another iteration with Lanes 2-5.  Every other test scenario featured 4 iterations per lane wheras that one ended up with 0 iterations for Lane 1 and 5 iterations for lanes 2-5.

 

(4) Naturally, in CMRT battles, there are occasions when automatic weapon fire DOES hit massed troops and multiple casualties are the realistic result.  But again, I don't want to confuse that with the question of accuracy.

 

(5) As with my earlier tests, I disregarded light wounds.  The casualties shown in my results are KIA and serious wounds only.

[TheVulture]

Nice systematic work.

 

You'd expect the accuracy to decrease consistently with range for a single shot weapon, certainly, but I'm not sure that automatic weapons will follow the same pattern. Each burst spreads over an area that increases with distance. I'd intuitively expect to see some effect in the accuracy curves showing up around the ranges where the area a burst is spread over is comparable to the area of the target. So the idea of the LMG accuracy plateauing around 180m might potentially be explained by that. But that's wild-ass speculation on my part, rather than something I've worked out the maths for.

 

For the single shot rifles, it might just be a question of statistics. It seems you are looking at around 50 casualties per weapon at each range band, give or take a bit of hand waving. As a rough ballpark estimate, you are probably looking at a statistical uncertainty of +/- 7 kills for the number of rounds fired, or about 15% So e.g. the 84.43 rounds per kill for the Mosin Nagant at 180m is +/- 12.6 - and you'd expect a 1/3 chance that the true accuracy is outside that range. So it is certainly possible that the Nagant curve is consistent with the K98 curve within statistical limitations. (And the null hypothesis would presumably be that both single shot bolt action rifles would have the same sort of change in accuracy with distance, even if the absolute accuracy might be different).

[MOS:96B2P]

 

 

Nice test Migo.  I really like the graph.  It would seem the sweet spot for rifle/LMG units (For the setting of Target arcs) is about 220 meters to 240 meters.  This keeps them out of the range of SMGs that have a hard coded range limit of 200 meters (see below link) however the rifles and LMGs are still reasonably effective in the 220 - 240 meter range.   

 

http://community.battlefront.com/topic/120738-hard-coded-smg-range-limit/

 

Now I wonder where on the graph the US M1 Garand, BAR, Thompson and the CW weapons would fall.  

 

Thanks for posting this.  Did I mention that's a cool graph?      

 

 

[Muzzleflash1990]

I do not understand some of these result.

 

Why do the SMG flatout? I mean why are you able to hit as often at almost double the range (180m vs 120)?

 

The K98 skyrockets, whereas the snipers stay almost constant. That suggest the reason for KAR98 is the difficulty in actually seeing and lining up the target. But then why is mosin nagant then significantly more accurate? If it is not in difficulty of seeing the target, then 300 shots at 300m, against a target the size of a football, from a unsuppressed shooter, that sounds rather poor?

 

Also, I can understand why you might be able to hit more often with a LMG, but not how you hit with more shots? Surely the shots with rifle are all about same precision and at least as accurate as the first shot from a LMG, whereas for LMG bursts the remaining shots will be more dispersed.

Edited September 8, 2015 by Muzzleflash1990

[TheVulture]

I imagine the LMG advantage comes from two factors:

 

* Firstly, many LMGs are stabalised in some way - maybe with a bipod for example. This makes it more accurate intrinsically than a rifle braced only on a shoulder (a rifle braced on a window ledge or some other stable platform would be a different matter entirely).

 

* The LMGs are usually belt or drum fed, and so can fire off multiple bursts without losing the aim. A bolt action rifle on the other hand has its aim disturbed by each reload, even if the shooter is able to perform the action without lowering the weapon from the aim point. So LMGs might be expected to improve in accuracy for shots after the first burst, giving it an advantage (except in situations where the first shot at a target is very likely to hit).

[Migo441]

@Vulture, thanks for chiming in with your thoughts and math!

 

MOS:96B2P says: "Now I wonder where on the graph the US M1 Garand, BAR, Thompson and the CW weapons would fall"

 

LOL, I was afraid someone would ask that.  In all seriousness, I expect I'll digest the feedback here and head back to the range at some point.  The testing bug is like a chronic disease, once you catch it, you can only manage the symptoms.  If a guy has a spreadsheet, some free time, and nothing to lose...  Well, you see what happens.

 

@Muzzleflash, yes, I was also puzzled about the observed non-degradation of accuracy from 120m to 180m.  TheVulture had a proposed explanation though which, if I understand correctly, is that the increased number of bullets from each burst is making up for the decreased accuracy per bullet.  We talked about this in the SMG II thread too.

 

As for the Mosin-Nagant appearing "significantly more accurate" (your words) than the K98, I want to be cautious here.  I make no claim that the game is favoring the Mosin-Nagant over the K98.  Yes, at the 300m distance, it does appear significant but the relative paucity of casualties at that range make the results more tenuous.

 

I will say this.  300m from 4-5 guys with bolt action rifles -- unexpectedly -- doesn't appear to be a very dangerous place to be.  I saw rifle teams fire for five minutes at that range against targets in the open and not cause a single casualty.  But again, we need to remind ourselves that, although we are kinda-sorta constructing a firing range, this DOES NOT represent firing range conditions, i.e., CM has a built-in "combat conditions" modifier that will decrease accuracy even in a pristine setting.

 

Do players have anecdotal CM stories about notable successes with non-scoped rifles out around 300m?  Or beyond?!

[Bulletpoint]

 

 

Do players have anecdotal CM stories about notable successes with non-scoped rifles out around 300m?  Or beyond?!

 

 

My experience from playing about 2 years is that rifles are not much use beyond 300 metres. That's when the machineguns start to come in handy.

 

Rule of thumb: Bipods from 300-500m, MMGs from 500-800 and HMGs from 800-1500m.

 

(of course, long-range weapons will also be effective at shorter ranges, but then they have more competition, so better to keep them at ranges where they can outclass the rest)

 

At ranges of about 800m+ you need binoculars in the team, if you dont have them your machineguns won't spot much, and suppression at long range is super weak unless you have lots of guns firing.

 

Not based on testing, just on my gameplay experiences.

 

Thanks for testing, seems like good work.

Edited September 8, 2015 by Bulletpoint

[Muzzleflash1990]

Just to be clear, i'm not not questioning the lethality of a rifle, or even multiple, vs. a LMG at 300m, only the per-round hit probability.

 

Sorry, maybe I'm dim, but what I don't get is the argument about SMG (and the LMG), that firing more rounds in a burst somehow is the reason for the increased, and also flat, rounds-per-hit ratio. Like, "the increased number of bullet from each burst making up for the decreased accuracy per bullet", seems contradictionary to the idea of SMG and LMG having better per-bullet-accuracy. For the flatness of the SMG case, if you are able to hit with every 175 bullet at 180m, the target being closer at 120 you should be able to do even better? That is the lines in the graph can at best be linear, at worst, curve upwards?

 

About K98 and Mosin Nagant, you are right. Since it requires so many bullets, without knowing how many casualties occurred, like at least 30, it is impossible to conclude, the K98 is more inaccurate at 300m.

 

Oh, btw, thanks for taking your time to do all this

Edited September 8, 2015 by Muzzleflash1990

[Migo441]

@Muzzleflash, you're right...  I'm confusing the issue.  Yes, we could have SMGs and Rifles "balance out" in terms of casualties inflicted at a given range where the SMG fires three times as fast as the Rifle and the Rife has three times the accuracy of the SMG.  But you're saying how or why would the automatic have a better PER BULLET accuracy? (as we see with the LMGs compared to rifles past 240m)

 

And yes, the flatness from 120m to 180m is curious.  One takeaway is that a commander should think twice about closing with the enemy over that distance, his troops will be more vulnerable as they move and they won't be any more effective once they arrive.

 

But -- IMPORTANT NOTE -- please take these results/observations with a grain of salt.  I don't expect they will unlock some heretofore unknown secret to CM success.  Not sure how much relevance the range data brings to the actual battlefield, but I do hope it can shed some light on things and maybe help us ask different, better, more specific questions.

[Hister]

Awesome graph! Sweet results. 

[JasonC]

First on the discrepancies.  120 and 180 being equal looks suspicious and suggests a coding thing, not a minute of arc aiming thing.

K98 better than Mosin at 120 and worse at 300 is screwy and probably just low sample size at the longer range. In real life, there is no appreciable difference at 100 meters but the K98 is significantly more accurate than the average Mosin at 200-300 yards. A typical K98 is a 2 MOA rifle, a typical Mosin is more like 4 MOA.  This is a matter of production tolerances and quality control.  There needn't be any large difference between them in game, but if there is one is should favor the K98 and favor it at longer range, not shorter.

The LMGs being more accurate *per round* than the bolt rifles at 240 yards is just silly.  A short burst from an LMG may indeed be better than a single rifle shot at such distances, but not per round.  Sure the LMGs are heavier and have bipods, but prone or supported firing positions match what bipods do, for a bolt rifle.  And the automatic is spraying and has muzzle climb and such.  The right relationship should be that a 6 to 9 round LMG burst is maybe 2-3 times more likely to hit than a single rifle shot, but each bullet is 2-4 times *less* likely to hit.

 

There is a similar comment to be made about the SMGs.  They aren't giving up nearly enough compared to the bolt rifles, and will get back way more than the reported accuracy difference from rate of fire, at any range where the men will fire them.  If that relationship actually held, they wouldn't stop firing at 200 meters, because they'd be keeping up with the rifles, using quantity instead of quality.  In reality, even with their quantity they are behind at that point and it is a waste of their ammo and total firepower potential to throw lots of low accuracy shots at 200 meter ranges.  The point being, the ratio of an SMG accuracy to a rifle's accuracy should have fallen below the practical ROF difference by that 200 meter range window.  At ~100 meters they made have the same relationship described above for LMGs - firing more rounds, less hits per round, more hits per second, in roughly an even "geometric ratio" split.

 

The ratio between the scoped and unscoped rifles is excessive for the shorter ranges.  A scope certainly helps at 120 yards, but not by a factor of 10.  It makes a much bigger difference at 300 yards - but the scope rifle still sees accuracy drop significantly at that distance, and shouldn't be hitting one shot in 4 at 120 and one shot in 5 at 300, as the marksmen-only German data shows.

Even with the scope, the 300 meter shot is dramatically harder than the 120.  At 120, the inherent weapon dispersion is an inch and a half and the bullet drop is less than 2 inches and for most common zeros less than 1.  At 300 meters the inherent weapon dispersion (side to side etc) is 3-5 inches and and  bullet drop is up to a foot and a half - even benchrested.  And every minute of misalignment in the barrel that the shooter imparts has 2.5 times the effect on the point of impact downrange.

 

Basically, effective accuracy should always decrease with range faster than linearly.  It can start as nearly a linear degradation from a good starting figure; it accelerates once bullet drop becomes a significant issue to faster than linear.  It is pulling away from linear already from the slowing of the round with air resistance, and drop adds a need to correctly estimate range and holdover, with imparts an additional source of shooter-input misalignment of sights to target etc.

 

Overall, the stats show are very favorable to scoped rifles for marksmen at least, are very unkind to bolt rifles, are overly generous to full LMGs at the longer ranges, and are too kind to the SMGs in the outer half of their range envelopes.

[Sgt Joch]

Great work Migo.

 

it does suggest areas to be looked into:

 

- The accuracy of the PPS-43 over 100 meters seems to be a bit high, it should probably be closer to the MP40. Did you also test the PPSH-41?

 

- The accuracy of the K98 and Moisan seems a bit low, especially in the 150-250 meter range. They should be  more accurate than a LMG in that range.

 

LMGs/MMGs/HMGs were substantially reworked after CMBN came out to increase ROF/accuracy at short range, so I doubt those numbers will change much. We should use them as a base line for the other weapons.

 

Did you check on the leader rating? a +1, +2 leader will provide a small accuracy bonus. That could have an impact.

 

a "Marksman" designation also provides an accuracy bonus. Snipers seem to be performing well.

Edited September 8, 2015 by Sgt Joch

[General Jack Ripper]

 

 

 

 

 

 

Now that's some good testing. I hope you had a good sound mod. 

 

In particular, it's striking how the accuracy at 180m is much the same as the accuracy at 120m for several weapons.  The range jumps by 50%, but the accuracy (judged by average rounds per casualty) degrades by only 21.44%, 3.08%, 3.53%, and 4.12% for the LMG42, DPM, PPSh/PPS-43, and MP-40 respectively.  Not sure what's going on there or if anything is going on.  Statistical noise?  I dunno.

 

Most WW2 era weapons were battle sighted between 100-200 meters, so you could engage a target anywhere in that range without having to adjust your sights. I don't know if that's the exact case here, but it's a possibility.

 

The related point is how the accuracy of the weapon types relate to each other.  The snipers are best, then the bolt-actions, then the LMG, and finally the sub-machine guns.  Each type is grouped together in its expected place along the accuracy continuum which feels good. 

 

Until you get past 240 meters, where the LMG's have a marked improvement over the Bolt-Action rifles in terms of performance. I was actually surprised by that result until I realized with an LMG, you can adjust your fire from the previous burst, where as a bolt-action rifle doesn't have that option, being required to work the bolt and reacquire the target for every shot.

 

2.)  What's going on with the German Snipers?  Why are they performing so much better than the Soviet Snipers? 

 

I believe I found the answer to this one hiding in plain sight.  Yes, the German snipers performed better than the Soviet snipers but my working theory is because the German sniper test shooters ended up with a freakishly large proportion of designated 'Marksmen' rather than the basic 'Soldiers.'

 

Having the actual 'Marksman' soldier type makes a big difference, but you also need to factor in the relative quality of the optics mounted on the rifle. The game simulates differences in optics quality.

 

Excellent work with this testing.

[Sgt Joch]

 

The ratio between the scoped and unscoped rifles is excessive for the shorter ranges.  A scope certainly helps at 120 yards, but not by a factor of 10.  It makes a much bigger difference at 300 yards - but the scope rifle still sees accuracy drop significantly at that distance, and shouldn't be hitting one shot in 4 at 120 and one shot in 5 at 300, as the marksmen-only German data shows.

 

 

 

 

While I would agree that in real life, the difference between a sniper and and average rifleman would probably not be that high, we have to remember that CM is trying to simulate combat conditions. Normally a sniper/marksman has the time to more accurately line up his shot while the average rifleman will usually only have quick opportunity shots. These are all elements which are factored in

[c3k]

Migo,

 

Nice, but, as we all know, it takes HALF a million rounds for any test to be considered even halfway meaningful.

 

Seriously: this looks very good. I haven't re-read it, but your numbers and graphs are a solid reference for any work moving forward from here. Very nice. Thanks for doing this.

 

Ken

[Vanir Ausf B]

Did you check on the leader rating? a +1, +2 leader will provide a small accuracy bonus. That could have an impact.

 

This is news to me. Has this been tested?

 

a "Marksman" designation also provides an accuracy bonus.

 

A rather massive one, as I recall from the Black Sea beta testing.

 

[Sgt Joch]

This is news to me. Has this been tested?

 

 

its in the manual. Leadership bonuses help "direct fire to be more effective". I have never tested it myself.

Edited September 8, 2015 by Sgt Joch

[c3k]

Geometric vs. Arithmetic loss of accuracy with range: It makes sense. Bullets drop in a ballistic trajectory which is geometric. E.g. (spurious example), a round may drop 2" at 200 yards, 6" at 300 yards, and 16" at 400 yards. JasonC talks about MOA (angular error, therefore "arithmetic" wrt range), but it is beyond that. In combat situations without optics, the "regular" CM troops should show a significant drop in accuracy at about the same rate he bullets' trajectories drop.

 

It is hard to calculate a 41" drop vs. a 31" drop, and then aim for that. Shrug. It is FAR easier to be on target when you know the range and have a calibrated sight with which you can adjust the reticle to compensate for the drop. (Ladder type rear sights are far different. Better than a notch, but far worse than adjustable optics.)

 

I like what the graphs show and how they show it.

[JasonC]

On sights and zeroing, yes yes, we all know the combat sights were typically zeroed for 100 meters or yards (though for some SMGs it might be more like 50), and that they had notched sights for longer ranges. Out to 500 or 1000 yards on the bolt rifkes, even. But anyone who has actually tried to use those frankly primitive sights to actually get hits at 500 yards will tell you, having a mechanical graduated doohickie stapled to the top of the gun and actually hitting a target at 500 yards with iron sights are two very different things. Yes it is easier at 200 - but still significantly harder than shooting at or near the zeroed range of the riffle. I will walk through a few of the reasons why, for the benefit of any here who haven't done it personally.

First you need to know which range notch on the sight to use. WWII infantrymen were not issued a moden hunter or sniper's laser rangefinder. If you use the 200 meter notch at 300, you will be shooting low, the reverse and you will be shooting high. At near ranges, because the total bullet flight time is low and the furst 50 meters or so it is still rising (that's what a 100 meter zero means and does, roughly), the difference in falling time from the apex of the bullet trajectory is small and is in the early oart of the 16 feet times time in seconds squared bullet drop from that apex. The longer the range past the 50 meter half zero distance, the longer that flight time, and the greater the total drop, and the vertical inches of difference for getting the range estimate wrong by a notch. Beyond 300, you are also more likely to get the estimate wrong by more than a notch etc.

Then there is the actual sight alignment. On the bolt rifles, this is a small rear notch cut like a grove into the top of the rear leaf sight, and raised byna ramp thing sliding under that notch to increase the angle to the front sight. They aren't very good sights by modern standards (a few like the US M-1 were better ghost rings, but thats the exception not the rule). The longer the range, the harder it is to have the sight alignment correct. The target apparent size in minutes of arc is falling, and the front post isn't getting any narrower. Soon it is the entire width of a man sized target, and you can't tell the difference between aligning on the center of mass of the target and just barely being on it.

Then you have to hold the weapon as steady as you can as you pull the (generally, pretty stiff and heavy 'Weight' by civilian shhoting standards) trigger. In a supported prone position or resting the stock (not the barrel) on a wall or rest, this isn't too hard, but in any other firing position it is significantly harder. The longer the range, the more the same uncertainty or sway in the hold moves the target around in the sight picture. That is growing purely linearly with the range in the horizontal, but uncertainty sway in the vertical is moving the impact point more at the longer ranges. Higher arcs the bullet with more of its flight time rising and less falling time, etc. you can do the ballistics if you doubt it, but the mathis simple enough that the same uncertainty in the vertical angle makes a more than linear difference in the height of the shot, the farther that distance is.

All of those more than linear effects are worse with a pistol caliber round that with a full rifle round, because it is moving only half as fast, and is thus in the air twice as long for a given range. Automatic fire also adds muzzle climb spray to this, and more of it with a lighter SMG than a heavier full LMG, with a standing hand hold that prone bipod supported, etc.

The difference in accuracy between prone or rest supported rifle fire and standing off hand is profound. It is easier to hit consistently at twice the range in the former stance than to hit from the latter. Even 100 meter shots are relatively challenging, standing unsupported with bad irons, right on the weapon zero, because sway is just so much more of an issue. Whereas the same shot with the bolt rifles prone or supported is a piece of cake.

The main moral is that the period sights having notches supposed to solve the soldier's holdover problem for him, in fact just don't. A modern scope with BDC reticle does such things much better, but still relies on accurate ranging, and longer shots are still significantly harder, using them, than 100 meter shots right on the weapon zero, or 200 meter ones with a full rifle round and thus small total drop even without adjustment.

FWIW...

[c3k]

Umm...so you agree that the curves in the graphs are about right?

[JasonC]

c3k - no I am saying the curves on the graph are about wrong.  They don't rise consistently enough, and they are way too kind to low muzzle velocity SMGs at middling distances, specifically.  They should not look like S curves but more like moderately accelerating exponentials.  In shape, the K98 curve is the only one that looks about right, but in level, pretty much all the others, except for the sniper lines, should be above it. 

[Bulletpoint]

Hm, the SMG curve might be off for another reason too.

 

The chart shows how many shots it takes to make a casualty, not how many pulls of the trigger.

 

So, let's imagine a rifle and an SMG firing at each their own target at 30m distance:

 

The rifle fires at its target, hits, and kills the guy.

 

The SMG fires, lets out 10 bullets, the first bullet hits and kills the guy. The rest of the bullets are wasted.

 

In your chart, this result would place the SMG as 10 times more inaccurate than the rifle, even though both shot and hit with the first bullet.

 

No matter if the SMG is really as accurate as the rifle, it will always score less accuracy in your chart, because it will always let out more bullets.

 

Or did I read your test wrong?

Edited September 9, 2015 by Bulletpoint

[Migo441]

@Bulletpoint, yes, that's correct.  I am treating "average rounds per casualty" as a surrogate for accuracy which, I think, is defensible in the sense of something being precise or exact.  But, as you say, if you were to define accuracy as "average trigger pulls per casualty" it would be a whole different picture.

 

I do want to be clear that Rate of Fire is a whole separate issue.  In a firefight, you certainly wouldn't want to reflexively pick the weapon with the lowest  "average rounds per casualty" at a given range.  At that point, you'd always pick the sniper rifle and then get shot from 60m by a guy who picked the submachine gun.    

@Sgt Joch, I didn't test the PPS-43 and PPSh-41 separately.  I just used the '44 SMG Co. where every squad had some of each and that's why I had to label it 'SOV SMG' rather than the individual weapon.  They fire the same round and the PPS-43 has, what, a barrel a smidge over 90% of the length of the PPSh's barrel?

[bobo]

I love this comment,

"(2) I know it's a more challenging shot when there is an elevation difference between shooter and target. That's not from experience, that's from reading Stephen Hunter novels."

Bobo

[Bulletpoint]

On 9/9/2015 at 10:56 PM, Migo441 said:

I do want to be clear that Rate of Fire is a whole separate issue.  In a firefight, you certainly wouldn't want to reflexively pick the weapon with the lowest  "average rounds per casualty" at a given range.  At that point, you'd always pick the sniper rifle and then get shot from 60m by a guy who picked the submachine gun.    

In the game, the sniper rifle would probably win though.  Because the way it works in the game is that the SMG chooses one (1) aim point and then the whole salvo either hits or misses that location. There's very little spraying or scattering effect that would increase the chances of hitting.

In real life, you might know there's an enemy in a group of three bushes, but you don't know which one. You can then spray all three very quickly with your SMG, and have a good chance of hitting something. With the rifle, you'd have to slowly fire shot after shot. In the game, it seems there's little benefit from the additional bullets in each burst. The whole burst effectively counts as a single shot.

Same goes for LMGs. I'm currently in a game where my LMG team has been firing for two full minutes at an enemy soldier kneeling in the open about 180m away, but they just can't hit him. First burst goes 4 metres to the right, next burst goes 4 metres to the left, third goes 4 metres to the right again, etc. etc.

(sorry for digging this thread up again, I keep thinking about accuracy and how it's modelled in the game)

Edited April 1, 2016 by Bulletpoint