I am quite late to this discussion, but looked at the references provided by Wiking with interest; especially the Hitchman Report.
I would first like to note that the Hitchman Report was prepared to argue that the Army should develop a 5-shot salvo weapon instead of a new rifle or automatic weapon. So there is a definite and intentional direction to its conclusions.
With respect to "9 years", here is what the footnote in the report actually says:
"One expert rifleman at Fort Benning, Georgia, estimated that it required nine years of continuous training on fire arms to develop marksmanship to the proficient level which he now enjoys. Sgt. Justice's performance in demonstrating the use of infantry hand weapons is most dramatic. His skill in marksmanship actually approaches the accuracy of the weapon; he has attained a level or performance roughly commensurate with the design precision of the weapon. However, it is estimated that less than 10 percent of the men in the normal recruitment stream could possibly reach this level of small arms proficiency, even if time allowed for training were long."
Now, just to be complete, the authors state that the accuracy of the "standard M-1" is 2 MOA (indirectly based on their quoted "expected dispersion"). The fact is that Sgt. Justice is beyond good if he can shoot 2 MOA with a military rifle using iron sights. Hell, I qualified expert (M-1, M-14, M-16, M1911), but the good sergeant is light years beyond that. One doesn't need to be that good to become a sniper.
I have personally known one man who was that good. He was an ex paratrooper rifleman who combat-jumped with the 82nd at Sicily, Italy, Normandy and Holland (and highly decorated). He was my mentor and coach for a couple years in civilian life; nevertheless, people like Ron and Sgt. Justice have a 5/6 sigma level of skill in my experience. A level of skill that is completely irrelevant to the discussion and not achievable with training alone (IMHO). I suspect Sgt. Justice knows exactly how good he is and that what he really said was something along the lines that he has been training and practicing for nine years. The footnote does not quote the sergeant, not does it attribute the 10% conclusion to him.
Next "average distance... 75-100 yards":
This comes from two retrospective reviews of wound ballistic data. While the original reports are not available (nor is their data or methodology included in the report), the report and citations in the appendices imply that it is based on a medical review of wounds. While one could extrapolate distance from wound depth (knowing the weapon, muzzle velocity, etc.), it also makes sense to conclude that the wounds subject to study is limited to some subset where the depth of the wound was measured (or could be measured). Not "generalizable" on its face (because it is far from complete, i.e., depth determination would require surgery, radiology or autopsy for bullets which did not hit bone, through-and-through wounds don't offer data... - I seriously doubt it represents a random sample of "hits" by any approximation). Population estimates (extrapolations) require random samples of sufficient size to make any meaningful claim. [No, I am not a statistician - scientist and researcher]
Nevertheless, the 75 yard figure comes from Bougainville (can you spell jungle) and the 100 yard figure comes from a sample of 109 wounds suffered by soldiers in the Turkish Brigade during 1950-51 (can you spell small sample). It is rare, if not impossible, for retrospective studies to have a random sample; meaning that broad-based conclusions are frequently just plain wrong. This has been a serious problem in research for ever and has led to some amazingly false conclusions.
Furthermore, ponder how one can possibly determine the range at which hits are actually made in combat (beyond sniper shots of course). I can not conceive that data will ever be available without some type of computer technology that does not currently exist. It certainly didn't exist at the time these "conclusions" were made.
It has been pointed out that "you can't hit what you can't see" and that Afghanistan is a different situation. Quite true.
80% of effective rifle... at ranges less than 200 yards
The cited "studies" were based on interviews, "effective" was not defined, and the actual question was based on the veteran's estimate of engagement range. While the author did not include actual data from the reports, the reports themselves don't contain data either. The actual question really addressed "engagement range" which is clearly a function of conditions (day, night, etc.) and terrain. What Hitchman doesn't mention is the infrequency with which the soldiers actually aimed their shots (virtually never at night and not often in offensive operations - most aimed shots were during daytime defense). Nevertheless, they conclude on several occasions that the riflemen are nowhere near as accurate as their weapons (preparing an argument that too much emphasis is placed on rifle accuracy). More on that later.
In effect, they have data on engagement range from the interviews (nothing new there relative to the 300 yard number) but inappropriately render a conclusion related to distance and accuracy.
Marksmanship:
The "original research" included in the Hitchman Report comes from a very small study conducted at Fort Belvoir, Virginia. In this study, 32 soldiers (16 who qualified Expert and 16 who qualified Marksman) participated in a marksmanship test. They fired at silhouette targets at 205, 310, 265 and 110 yards (left to right). They fired in groups of four (2 expert and 2 marksman) and with a 3 second exposure every 3 seconds. From the diagram, the angle between the 205-yard (B) and the 110-yard (A) is close to 90 degrees (which makes a B-A or A-B sequence a fairly difficult shot).
So, some problems
Individual Marksmanship - one person shooting at a time. Eight passes through the 5-target sequence (40 rounds per shooter). But each individual was assigned a different sequence (e.g., A-B-C-D-A, C-D-A-B-C ...), meaning that each individual had a different test. Two had a B-A switch, two had an A-B switch; one had two shots at 110 yards, the rest had one; one had two at 310 yards, the rest had one. Therefore, a real problem comparing different tests between individuals or groups. Must also assume there is no effect of sequence (and as a shooter myself, I believe there would be); or perhaps period (i.e., time of day since that was not stated).
Group Marksmanship - all shooters fire on the same target in the same sequence (B-A-D-C-B). Four passes through the sequence (20 rounds per shooter). In this test, everyone fired the same sequence (which is good), but a real problem comparing this to the individual result.
In both cases, it appears that targets were not switched between individuals. Meaning that all the experts shot the same target and all the marksmen shot the same target. Therefore, it is not possible to get error or accuracy measures for individual shooters. This is a very poor design and the statistician attempts to "estimate" some voodoo to account for this problem (which isn't possible).
As a result, it is:
- problematic to compare individual marksmanship between groups (expert versus marksman)
- not appropriate to compare individual marksmanship to group marksmanship
- not appropriate to extrapolate the results to the rest of the Army
Piss poor study design before they ever fired the first shot and the sample size is way too small to make a determination concerning marksmanship skill for the Army. Also, the report did not indicate whether there was uniformity between the shooters in either the amount of recent practice or time since last qualification (extremely important factors).
Next, when they presented the results they did not present the results for the entire sample. Why? This is highly suspicious in the presentation of research data and usually cause for instant rejection of the conclusions by peer reviewers.
For example, with the experts shooting the individual test results are presented for 12, 10, 9 and 9 shooters at 110, 205, 265 and 310 yards, respectively.
Within the "expert group" 84% hit rate over 12 shooters at 110 yards in the "individual test" (different sequences) versus 100% hit rate for 8 shooters at 110 yards in the "group test". Are these the same shooters? What were the results for the others? Did all 16 experts shoot both tests (or marksmen)?
When examining the sequences, I suspect that what is reported as "number of shooters" is actually "number of shots"; so if they are correct in reporting that there were 16 expert shooters, it looks like there were 2 different groups of 8 (which further complicates the comparison - especially since there is no ability to compute a standard deviation for the individuals).
OK, way too technical. But this is a crap study and WAY TOO SMALL to reach conclusions related to marksmanship across the Army.
Does accuracy decrease with range in both groups? No doubt in my mind.
Do people who qualify as expert shoot better than those who only qualify as marksmen? Duh.
Can you hit something you can't see? Nope.
Are their computed hit probabilities accurate? Only by providence, if at all.
Is it rational to compare the "inaccuracy" of an group to the "inaccuracy" (probability of a miss) of the weapon? YGBSM.
They actually stated that the accuracy of the shooters was inferior to the accuracy of the rifle (duh). Citing a probability of missing at 300 yards being 0.04 for the rifle in a rest/machine to 0.76 for a "marksman" firing the rifle ("individual" fire I believe). Part of their conclusion that the weapon need not be as accurate or as long-ranged. But think about that, it also means that a rifleman of the lowest qualification can hit a silhouette target at 300 yards 1 time in 4. That's really not too bad for a 3 second exposure.
I'm not trying to be a troll. I'm a retired researcher, pulling the layers off the onion is my thing. Just making a point about military "research" over the years. Lots of poorly designed/executed "studies" of very small sample sizes leading to broad and unjustified conclusions.
Is it really rational at all to even attempt to judge Army marksmanship based on 32 men from the Army Engineer school? What about 100 shooters from 8-10 infantry commands around the country (all taking the same test)? Think there might be differences between commands? Perhaps the level of training and currency might have a small impact?
Best regards,