Only bullets that hit count

Wednesday, April 10th, 2019

The Hitchman report paved the way, Emeric Daniau argues, for numerous studies that tried to evaluate the rifleman’s effectiveness under what was considered “realistic” stress conditions and tried to provide a technical answer to the perceived lack of effectiveness at ranges longer than a hundred meters:

The founding idea of those studies (like ORO-T-160, but also SALVO I & II, SAWS and several others) was that “only bullets that hit count”, and that the only military effect of small-arms fire (and Individual Weapon fire in particular) was hitting and disabling a target. Suppression effects that are known to greatly reduce enemy fire effectiveness and enemy movements (two very interesting military effects) were simply not taken into account and no effort was made to try to evaluate them, or incorporate results of other studies devoted to such topics.

This idea of limiting the effectiveness of small-arms fire to hitting and disabling enemy soldiers immediately calls upon past visions of glorious battlefields where dense masses of soldiers were shooting at each other, or where a handful of brave souls stand against a “human wave” assault of mechanized infantry (“high density” battlefields), but seems totally remote from the “low density battlefields” so frequently encountered during “decolonization” wars or peacekeeping / stabilization engagements.

From a methodology point of view, this choice (deliberate or not) to reduce military effectiveness of Individual Weapon fire to “bullets that hit” had major implications. First, the complexity of evaluating small arms effectiveness was greatly reduced, “scientific” evaluations could be performed and focused on hit probability (pH) and terminal effectiveness (pI/H) against unprotected targets, or after defeating personal protection (but not intermediate barriers).

Second, since the maximum effective range considered (300 m) is relatively short, almost any bullet pushed fast enough could do the assigned job (hitting and delivering “sufficient” terminal effectiveness).

Of course, the capability to hit something is very valuable, but what is the hit probability of a soldier in a real combat, as opposed to simulated combat?

The “shots to casualty” ratio of small-arms fire is a highly debatable issue, and numbers as high as 100,000 have been quoted, but without a strong database to sustain that claim.

More reliable values could be found in the experience of the First Australian Task Force (1ATF) during the Vietnam war, with (mean) values of 187 shots per casualty for the 7.62 mm SLR and 232 shots per casualty for the M16 in the context of day patrol.

Nearly 80% of those engagements took place at ranges shorter than 30 m, not really long range, and still the average hit probability was around 0.5%, compared to a hit probability of ~100% found in ORO-T-160.

Of course, “mean” values are only average and in particular events close to “ideal” shooting scenario, shots-to-casualty ratio around 30 to 1 were achieved. While this number (pH ~3%) is definitively higher than 0.4% or 0.5% (nearly one order of magnitude), it’s still a very substantial difference from results commonly found during simulated combat.

The French operation in Mogadiscio in June 1992 could be seen as a very good example of effective firing, but even in this scenario ~3500 small arms (5.56 mm and 7.62 mm) rounds and ~500 12.7 mm rounds were expanded to produce a maximum of 50 casualties (pH of 1.25 %).

Police shootings that take place at very short range (generally less than 7 feet) exhibit the same symptoms of very low hit probability, one or two orders of magnitude less than expected. For example, during the famous 1997 North Hollywood shootout, the two heavily armed bank robbers fired approximately 1100 rounds during a 44 minutes battle and wounded 11 police officers (pH ~1%) and 7 (probably untargeted) civilians.

In return, police officers fired an estimated 650 rounds and killed both perpetrators (it is possible that one committed suicide after being wounded). Both bank robbers wore homemade bulletproof garments and one was hit several times in rapid succession in his legs until he surrendered (he died later from blood loss), so it’s difficult to evaluate the hit probability of the law officers, but even at a few feet, with good visibility and superior training (the final part of the shootout was conducted by SWAT members at a distance around 3-4 meters), one should expect results probably not much higher than 5% to 10%, again a substantial difference between “real life” results and results recorded during simulated combat.

This difference could be easily explained because of course, during simulated combat, no matter the amount of “realism” of the shooting scenario (sounds, fumes, explosions, fatigue or even electric shocks on the shooters), the targets are not returning fire so soldiers could focus on “clearing the range” (and freely expose themselves during the process), while during real combat trying to minimize exposure time to avoid being hit is mandatory.

So, if we look back at Figure 30, we have an idea of the hit probability of a soldier firing his M1 rifle at a human-size target with an exposure time of 3 seconds.

In order to be able to hit this target, the soldier needs also to expose himself to incoming fire (from his target, or from other people waiting for a shot of opportunity, the battlefield is not a place for a duel) during roughly the same amount of time.


Evaluating the dispersion of hand-held weapons and trying to improve the hit probability was at the heart of both ORO-T-160 and ORO-T-397 (Salvo II study).

Most results found in ORO-T-160 used a target exposure time of 3 seconds and shooters were discriminated between “experts” (highest skill) and “marksmen” (lowest skill).

During those tests, “experts” scored significantly higher than “marksmen” (another “argument” against long-range firing in the hands of the masses). For example, during the second test, “experts” scored 8 hits (25% hit probability) on a man-size target at 310 yards (Figure 33), when “marksmen” scored only 2 hits (6% hit probability, Figure 34).

During those test, a large cloth was used to record as many shots as possible (hits and near-misses), and it’s interesting to notice that in both cases:

  • the number of shots recorded was nearly the same (25 for “experts” and 26 for “marksmen”),
  • the dispersion of shots was nearly the same (~7.25 mils for “experts” and ~7.5 mils for “marksmen”),
  • the number of rounds impacting in a 1 m circle around the head of the target was the same (23 shots, or 72%, in both cases), the only significant difference is the mean point of all impacts, “on target” for “experts” and slightly off target by 35 cm for “marksmen”, this shift of MPOI alone explains the difference between a hit probability of 25% and 6%. A deviation of 35 cm (~14 inches) at 310 yards is what you can expect from a 15 mph lateral wind acting on the .30-06 M2 bullet,
  • a third test was performed, using a 1 second target exposure time and a random schedule between two targets, one located at 110 yards, and one located at 265 yards (table A3 in ORO-T-160, p.100), and under those firing conditions the “marksmen” greatly outperformed the “experts”. Unfortunately, results obtained during this third test were not reviewed as deeply as results obtained during test n°1 and n°2, and no conclusion was drawn from it.

That last point makes it clear that we weren’t training riflemen to shoot quickly.

During the SALVO II study (ORO-T-397), a similar test was performed. The dispersion of each individual soldier was measured (“Experts”, mean dispersion between 1.97 mils and 2.93 mils; “Sharpshooters”, mean dispersion between 2.12 mils and 3.67 mils; “Marksmen”, mean dispersion between 2.30 mils and 3.70 mils) and from the conclusions of this report, it is clear that the level of marksmanship (for a given training program) plays only a small role in hand-held weapons dispersion compared to target exposure time and visibility (Figure 35).

The 7.5 mils and 7.25 mils obtained for “marksmen” and “experts” respectively, for a target exposure time of 3 seconds found in ORO-T-160 (published in 1952) are close to the “upper bound” found in ORO-T-397 (published in 1961) and probably reflect the change of marksmanship training (TRAINFIRE I was introduced in 1954, and TRAINFIRE II in 1957), from “bull’s-eyes” targets to “popup” targets.

Mean dispersion found in ORO-T-397 was around 3 mils for a target exposure time higher than 5 seconds, and this dispersion was increasing as the exposure time was decreasing, up to 7–7.5 mils (24 MoA to 26 MoA).


So, there is a wide difference between the way we evaluate small-arms hit probability and “real-life” results which seem to range from 0.3% to 3%, even at fairly close distances. From an individual fire perspective (one soldier, one target and one bullet), this value could be considered low, but from a tactical point of view, with tens to hundreds of soldiers, each carrying more than a hundred rounds, it’s high enough to produce decisive military results.

For example:

  • during the battle of Magersfontain in December 1899, fire from the 8,500 Boers’ individual weapons (Mauser bolt action rifles) at a range of around 400 yards (366 m) was sufficient to kill and wound 665 British soldiers (24.5% of the total) in the first 10 minutes of the battle,
  • a few days later, during the battle of Colenso, 2 British batteries (12 guns) of field artillery were engaged by rifle fire at a distance of 700 m. Suffering heavy casualties, the British were forced to fall back to their camp, losing 10 guns in the process.


The adoption by the US, followed by NATO, of the .223 Remington cartridge as the 5.56 x 45 mm, was not the result of concluding that the battlefield depth (measured in kilometres before WWI) was now reduced to 300 m, but an acknowledgement that effective HE support could be provided now at very short range in most conditions, and that the fire delivered by the infantry individual weapon should be used only for defeating adversaries in the 0 to 300 m bracket, longer ranges being devoted to collective weapons firing heavier ammunition.


  1. Bob Sykes says:

    Based on after action interviews during WW II (and Korea?), S.L.A. Marshall concluded that most infantrymen did not fire their weapons during a combat engagement. The ones that did were usually close to a fellow soldier firing an automatic weapon. These conclusions have been disputed. I don’t know of a similar study since then.

  2. Isegoria says:

    S.L.A. Marshall may have been a liar, but his point that riflemen won’t shoot has been corroborated by other sources — if not to the same degree.

  3. Kirk says:

    Marshall and Grossman didn’t do any real “studies”. Marshall made up crap that sounded good, and reported made-up anecdotal evidence instead of doing any kind of solid research to back up his ideas. Calling what he did a “study” is an insult to the gods of knowledge.

    Grossman is more of the same, at a secondary remove: He merely regurgitates the bullshit that Marshall and others crapped all over the historical record. Precisely no primary research of any kind has been done by Grossman–It’s all scholarship, and lousy scholarship, at that. He can’t even bring himself to question the basic premise of Men Against Fire, despite the fact that it’s been in disrepute since before his time at West Point. Grossman doesn’t even know his own field, that well. It’s like he did a survey of the library at West Point, collected up everything that supported his thesis, and then reported it as gospel truth to a credulous public.

    The outright fraud the two of them perpetuate with the exact numbers they quote for “non-participation” is right there, in front of anyone doing the reading. Marshall’s work is backed up by no documentation, whatsoever: You go looking for the primary records of the oral histories he supposedly based all of it on, and there’s literally no “there” there. All the numbers he quotes in Men Against Fire were, as near as I can tell, pulled right out of his ass. What’s mind-boggling is that nobody ever questioned his bullshit–Ever. It just got reported as fact, down the years, because that’s what everyone wants to believe, that people don’t want to kill other humans.

    C’mon down to the killing fields, y’all… I have some evidence I’d like to present to you that humans can positively thrive off of killing other humans. Cambodia’s Killing Fields, Rwanda, the Balkans, the Armenian genocide… All places and times where “normal humans” raped and killed like champions, with no inhibitions whatsoever on display for the majority. In fact, the scary thing is just how few had issues with doing the killing, when you look at it all. “Reluctance to kill”, my ass. You want me to believe that the guys out on the line in combat during WWII were somehow nobler, better men? Bullshit; absolute, utter fatuous bullshit. I don’t know why Marshall told the lies he did, or why Grossman chose to ignore the already-identified issues others have had with Marshall’s “work”, but neither one of them have any bearing on reality whatsoever.

    If someone can ever present actual evidence that Marshall’s numbers were based on anything other than his wild imagination, I’ll be shocked. So far as I’ve been able to research, there’s nothing there, under the publicity–He never produced the original records of his interviews, ever, and never published even the basic data on which his calculations were supposedly based. When he was called on that, he produced excuse after excuse for why he couldn’t. And, because of his fame, everyone shut up.

    The other thing that’s telling about Marshall is to speak to people he actually interviewed during WWII. Those guys all deny him ever bringing this subject up during the group bull sessions he did with them, and years later professed that they had no idea where he’d have come up with any of it.

    Combat non-participation is something that sounds likely, because we want to believe that it’s difficult for people to kill other people. It’s something we want to believe, but it isn’t true. At. All. Given the right circumstances, human beings are the most uninhibited killers on this planet, and there’s no distinction made between other humans and the rest of the animal kingdom. Doubt me? Look at the world around you, sweetheart, and do it with open eyes.

    What Marshall may have observed is that humans behave in rational, survival-oriented ways, ways that can be taken as “reluctance to kill” if you squint reallllly hard, and look at them sideways. The trick is, you have to differentiate what you observe from what you want to be true, and what is true is that most people are cowardly shits, when you get right down to it. There are probably only about 20% or so of any given unit that are what could be termed “combat enthusiasts”, or “men indifferent to their own survival”, and those guys are the ones that carry things forward. Who those men are will shift from moment to moment, as the adrenaline and motivation ebb and flow through the human psyche, but to say that 80% don’t participate because they don’t want to kill? LOL… Nope; not at all accurate. You take most of that supposed percentage of “non-participants” into a situation where their minds can do the risk/success calculation in favor of killing, and they’ll kill until they haven’t got any more victims left to kill. This is one reason the MG teams are so much more successful–You’ve got this great big noisy and effective weapon, you’re usually in the prone or dug in, and you’re laying waste to the enemy. Killing is easy, and you’re doing it with a buddy. Individual rifleman, though? He’s out there on his own, and more worried about drawing attention to himself, so he’s not gonna fire until and unless he’s got something to shoot at. Put him in a position where he’s safe, and the enemy is laid out before him like so many little duckies in a row…? Most of those guys are going to kill with an enthusiasm that would make your toes curl.

    I’d say that it’s true that probably about 20% of the unit carries the battle, but that’s not down to them being the only guys who don’t have a problem with killing; it’s because they’re the guys whose adrenaline and desire to kill is somewhat higher than the ones who make up the mass. And, who that specifically is, across a unit? That will vary. Guy who cowers in a hole today may be the guy who stabs some poor bastard to death with a messkit spoon tonight, and that guy who charged the machinegun position this afternoon may be the guy who hides under the truck in the dark, tonight, while the enemy raid force sweeps through the perimeter. It’s all variable and highly subjective; today, Carl may not give a f**k, because he’s got nothing to live for, but tomorrow, after tonight’s mail call and the letter telling him his girl is pregnant, all that may change 180 degrees. Likewise, Fred may get a letter telling him that his wife has left him, and all of a sudden, he’s out there in front with a bayonet, going after the stupid twats that dared fire at him and his buddies…

    It’s all down to the “people factor”, and people are just about entirely unpredictable and inconsistent. Even the ones you think you know…

  4. Alistair says:


    We agree there is a massive drop-off in infantry lethality between Instrumented Exercises and Actual Combat, right?

    That effectiveness degradation is coming from somewhere. Reduced-participation (and exposure) from fear (possibly suppression induced) seems by far the best candidate, right?

    Back in the day…looked at a lot of outcome data from Falklands, Italian, and Western Desert infantry defences at Platoon / Coy level. Actual killing effectiveness, in kills-per-defender-rifle-equivalent was consistently short by a factor of >10 over what the firepower models predicted. Even with controls for force ratio and terrain and troop quality. The crew served weapon density also tested out as a strong correlate of the killing that did happen.

    So I’m not saying that Marshall wasn’t full of it, but all the data I’ve seen didn’t disprove his hypothesis when we tested it.

  5. Alistair says:


    RE: Unpredictable and Inconsistent people and combat effectiveness.

    Maybe on an individual level, sure. But on a group level, well….that’s another matter entirely, and statistics come to bear.

    Actually, even on an individual level… there’s this thing called genetics…You’re Scots Irish (iirc), for God’s sake. Do you think it’s just your culture that makes you all so delightfully pacifistic?

  6. Adar says:

    “Nearly 80% of those engagements took place at ranges shorter than 30 m, not really long range”

    In Vietnam. And 50 % of ALL engagement took place at a range of 20 meters or less. An ambush or a meeting engagement in thick jungle or undergrowth. A lot of shots quick desirable. In this regard the M16 better than the M-14 or the Aussie SLR.

  7. Bruce says:

    Paul Fussell mentioned US troops being scared to fire in the cold air of northern France because the smoke hung in the air in front of them and provided a target. He thought Germans had better smokeless powder.

  8. Kirk says:


    I don’t mean to question or attack you personally, or the work you did for (I presume…) the Canadian Forces, but I have to ask: Did you guys take the glib pronunciations by Marshall, et al at surface value, the way I did back when I was young, or did you do what I later (under the pressure of embarrassment), and actually go looking at what the hell he based his “work” on?

    I shamefacedly have to admit that I could find nothing, anywhere, to support or serve as a primary source for what Marshall was saying. And, I looked–He claimed that all that stuff in Men Against Fire was based on actual post-battle interviews he conducted, but never, ever provided the precise interviews for corroboration. There’s nothing out there that says “On this date, Marshall conducted post-combat interviews with this unit, including these men, and this is the precis of what was said…”. It’s all crap “…straight out of his memory…”, and pulled up the same kind of bullshit excuses that Stephen Ambrose and Michael A. Bellesiles came up with when people started asking questions–”My notes and files were destroyed” “This is a reconstruction of lost notes”.

    What pisses me off about Marshall is that he and his reputation convinced me of a bunch of things, early on, and I had the signal experience of expounding (extremely unknowingly, believe me…) on his thesis points in front of a small group that happened to include actual WWII combat veterans and a guy who’d been interviewed by Marshall. I’m still pissed at that one, because there’s nothing so embarrassing as discovering that your “primary source” for a bunch of stuff was full of shit.

    I’m also angry at Marshall because of his incessant insistence that the whole Trainfire thing “grew out of his work…”, which is another whopper entirely. I’ve been all over the genesis literature of that program, and I can’t find a single damn mention of Marshall, his work, or much of anything else indicating that any of the people who worked on it were ever even aware of Marshall’s existence, or his “combat non-participation” theories. Marshall claimed credit for helping create that program, and there’s actually no sign in the literature extant that I’ve seen proof of that. Might still be out there, but I’ll be damned if I could find it. It is, sad to say, an arena of scholarship which is not well-studied.

    Grossman? Don’t get me started on that intellectual fraud. I’ve sat in his little pre-deployment “seminars”, which were paid for by well-meaning commissioned types who bought into his BS, and I think they actually did more harm than good. The line of thinking he puts out is that “…men don’t want to kill…”, and talks about how there’s this “natural reluctance” which has to be overcome. Care to guess how that effects those who discover they possess no such “natural reluctance”? Yeah; they start thinking there’s something wrong with them, which I’m afraid isn’t quite accurate. I’m of the opinion that at least a couple of cases of PTSD I know of personally got started right there in that theater, listening to that idiot Grossman expound on his fantasies about the intrinsic good of human nature. Not only is he a fraud, unintentional or otherwise, but he’s inimical to gaining an understanding of what is really going on, in precisely the same way that Marshall put everyone on the wrong track for literal decades.

    I don’t think you’re mistaken about the drop-off between simulation and training, at all. I just think you are misattributing the cause of it. “Reluctance to kill” is not it, at all–It’s mostly down to two things, in my opinion: One, there’s a hell of a difference between being shot at for realsies, with actual potential death on the line if hit, and training simulations. No matter what you do with MILES gear, until you’re somehow duplicating the effect on someone’s nervous system of taking a mortal wound, you’re naturally going to have one hell of a significant difference between what they do in non-threatening training simulation and real life. You may be able to condition somewhat better response with realistic training, but there is always going to be that missing “Reality” component to it–And, the training subjects damn well know it.

    Two, there is another factor that people leave out from the idealization of “What Should Be”, and that’s the fact that most soldiers, even the least well-trained ones, understand the impact of what they’re doing with those bang-sticks they’re firing. You don’t fire unless you’re absolutely certain that you’re shooting at the enemy, and that the fire is going to be effective–Yeah, it’s comforting to make noise at the big, bad, scary thing, but in reality, you don’t shoot at what you’re not certain of, unless you’ve gone past the stop on your fear-gauges. That factor is another thing to consider–The subject’s understanding of the effect of what they’re doing, and reluctance to ‘eff things up by shooting something or someone they shouldn’t.

    Together, I think those two factors cover a lot of the missing bits you find statistically. More so than this fantastic delusion that “people don’ wanna’ kill people, mmmkay…” thing that Marshall and Grossman came up with. Trust me–Spend some time watching kids on a playground, or who have access to a means to kill and some ready victims. There is no “inborn reluctance to kill”, in any regard: The little bastards will kill anything, even their own peers, without so much as a second thought.

    And, this is really what I find the most disturbing thing about these twin assholes: Marshall and Grossman both make it seem as though civilization and decent behavior are easy, inborn and inbred things, and not what they actually are, which is the result of hard work and constant reinforcement of behavioral and societal norms. Every generation has to be tamed and civilized anew, and if we were to listen to the feckless morons like Marshall and Grossman, we’d ignore that reality.

    Which is the genesis of a lot of the problems we see around us in the world today: People have actually bought into this bullshit of the inherent goodness of all mankind, and that’s why they’ve quit doing the hard work of diligently teaching and enforcing the norms and values. In their own way, Marshall and Grossman have probably done more to undermine Western Civilization than a lot of termites manage, gnawing away at the joists of your house.

    In short, Marshall and Grossman have taken us down an intellectual blind alley, when it comes to trying to understand human behavior in combat, and how to win the fight when we take those lessons to war. The fact that they may have gotten some of the numbers right, probably by sheer accident, does little to affect the fact that their misapprehensions became so widespread. You start to try to “fix” the non-participation problem by addressing this supposed “reluctance to kill” issue that they’ve ginned up, and you’re not going to fix anything–Because the actual causes are simply too damn different from what those solutions to an imaginary problem would address.

    The way you address this crap is to do two things that the Germans did very well in WWII: Ensure strong primary group formation, and ensure excellent junior leadership is there at all times. Anomie and isolation on the battlefield are the things that lead to men not doing their part in the fighting, and if you address those two factors, rather than trying to overcome some fantastic human trait that doesn’t actually exist, you’ll stand a chance of actually getting better combat performance from them.

  9. Kirk says:

    Alistair, second post:

    The statistical aggregate is great, but as you learn the hard way as a junior NCO, statistics don’t mean squat down on the individual level. “Hey, guys… DuPuy says we shouldn’t break and run, here, ‘cos not enough of us have died yet…” “Screw that, sergeant, that idiot DuPuy wasn’t working for that moron LTC Smith… That stupid fat f**k is gonna get us all killed…”.

    So, yeah… The aggregate is all well and good, but we out on the pointy end have to work with the actual raw material of those statistics, and it ain’t at all easy. Nobody tells you about how having that one guy on your team who seems like a rock, normally, can start an avalanche of fear and flight by seeming to lose his shit and run from the enemy. The fact that he lay down in a pile of fire ants, by accident? Yeah; that’s one of those little “incidents and accidents” that vanish into the maw of the statistical norm. And, those are the things that a junior leader has to deal with, not the higher math up at battalion which says “Hey, 3rd Platoon of Alpha company has taken a bunch of casualties… Better put them in reserve before they break…”.

    Statistics can be a valuable guide, when you don’t know the people, and you’re dealing with them in terms of impersonal distance. Being the poor bastard on the scene, you’d better know your people, and be able to work with what you’ve got, not what the stats say is “normal”.

    As to the Scots-Irish predilection for sheer bloody-mindedness, I’m not sure whether that’s a set of inborn traits, or just a mass of predispositions that get activated by exigent circumstance. Just like with anyone, or anything–There are reasons for stereotypes existing, and many of those are solidly based in real experience. But, the trouble with stereotypes is that you’re almost inevitably going to run into situations where they don’t pan out, at all.

  10. Kirk says:


    Fussell told the truth, here: US powders were not anywhere as “smokeless” or “flashless” as the German powders were, which is something that goes to reinforce my contention that an awful lot of the brass had no understanding of combat, or what was really important in combat.

    There are copious cites to show German concern with both those issues; on the US side, the biggest concern was cost and availability, not the combat-critical characteristics of those propellants. This was something left unlearned after WWII, despite many people pointing it out, and not really addressed until recently.

    I got to fire a bunch of WWII German MG ammo at night, once upon a time. You compare that experience to even what we issued in the 1990s, and you’re left going “Uh… I think we may have missed something, here…”. The signature from the German stuff was nowhere near as distinct as our ammo, even after fifty-odd years. Noticeably so, I might add.

    And, again–I have to point out that there are a lot of people who don’t “do” combat and who are making procurement decisions who dwell in a general fog of ignorance and idiocy. They don’t know what they don’t know, and since the guys on the line don’t know that there are changes that could be made to make their ammunition perform better, they don’t ask for them.

  11. Alistair says:


    S’ok; no offense taken, mate; I value your perspectives and experience.

    I came in cold to the programme as a wet-nosed graduate. The whole thing was run by an old-school engineer who had come across Marshall whilst trying to make sense of historical data which was definitely NOT behaving; it was supposed to supplement the trials data but looked nothing like it. Faced with a legion of Desk Officers defending their very expensive trials and models, he took the principled decision that it was all crap and went off to run his own stats based on the original war diaries, maps, casualty reports and survivor accounts.

    Any individual action was messy; but once you had a few dozen statistical patterns emerged. Once you had a few hundred they became irrefutable. The MG were doing a disproportionate share of the defence killing. Not all. But at lot. There was also a big difference between the SAW class weapons and the crew-served MGs; more than raw firepower would suggest.

    By now we had found Marshall. We looked at team and supervision effects. We had a team of historians crawling through the archives. We found guns that had Sgts or Officers in proximity fired more and seemed to kill more. Some people where doing a lot more than the others.

    We looked at heroism awards; we found people who won heroism awards and their sections/crew went on to be much more effective killers. We then checked before the award; and found they were already good killers before the brass noticed. These people just seemed to be less phased by danger. A disproportionate number of them ended up dead, of course. But that indicated a strong link between participation / exposure and effectiveness.

  12. Alistair says:


    Don’t get me started on Dupuy. He meant well, but oh gosh, what a mess from a data handling and hypothesis formation POV.

  13. Paul from Canada says:

    There is a lot of data that suggests “actual” performance is less than theoretical performance.

    I recall a study where they tried to replicate Napoleonic era weapons’ effects. They knew from studies done at the time (the Prussians did several, and so did the British), that there was a reasonably scientific hit ratio using a typical smoothbore musket.

    They took these studies, and using period muskets and reproductions, re-produced them, and compared the results. They validated the theoretical hit probability (i.e. “X’% hits at “Y” distance against a 5×10 foot target by a platoon and so on). Then they tried to wargame an engagement using laser guns programed with the same level of accuracy (dispersion), using volunteers. The result was that the re-creators achieved a far higher “kill rate” than that reported historically.

    Basically, when you are actually in mortal danger on the proverbial “two way range”, your actual performance is way worse. Understandably so.

    I recall an incident in the biography of a British soldier turned mercenary, of his experience in Rhodesia. An older rifle shooter, with trophies under his belt from Bisley etc. was criticizing the performance of modern soldiers compared to the long range marksmanship of the rifle association shooters. In response, he set up a scenario on the range, getting the shooter to fire on a figure target (rather than a bulls-eye), running progressively from firing point to firing point, and finally, when the older man started shooting, he yelled, screamed and shot up the ground around him with his own rifle. Needless to say, the results were not pretty, and his point was taken in by the older shooter.

    Myself, I think that the Pareto distribution is at work here. To follow up what Alistair said, there is a small minority actually capable of thriving in the unnatural environment of combat. I commented before about the psychology of survival situations, and the distribution of reaction to same.

    A minority panic, go catatonic and are completely useless. Another minority, are capable of staying calm, thinking and reasoning, and functioning more-or less normally. The middle of the curve responds to training and follows instructions/orders, to varying degrees of efficiency, but do not perform optimally. Hence the need for training and drills. This is the majority of people and the majority of soldiers.

    There are some who can stay calm, control the adrenaline, and shoot calmly and accurately, BUT THEY ARE A MINORITY.

    Interestingly, there is some literature out there to suggest that most people under stress get tunnel vision and time dilation, the frontal cortex being swamped by adrenaline/epinephrine. Those that perform best do not, they keep being able to use the frontal cortex. Interestingly, they also suffer less PTSD. A lot of Special Ops types fall into this group, and the discussion is currently a debate as to whether Special Forces selection selects people with this ability, of if the training inculcates this ability, especially since some of the brain chemicals concerned, increase with physical fitness and training.

    There are lots of very competent and professional soldiers, who excel in peacetime, who will likely fail in wartime, simply because there are very few with the natural talent to do so. There are plenty of fit, driven, dedicated cyclists, but very few Lance Armstrongs.

    I have a suspicion that proximity and danger have a lot to do with the relative degradation of performance over time. In the Boer war, if you were firing controlled volleys at long range, in a group, under orders and control of your superiors, you were in little actual danger. Or rather, the few enemy rounds cracking overhead did not SEEM to be really dangerous. On the other hand, in the jungles of Vietnam, or Malaysia, or the dense bush of Rhodesia, the enemy blasting rounds into your position form 20 yards away, and you by yourself, the nearest comrade 10 yards or more away, is a very different thing.

    The other point to make, is do infantry smallarms really matter in the grand scheme of things? It seems to me that the infantry is either pinning the enemy in place so that support fire can take them out, or allowing a maneuver element to move to such a place as to make their position untenable. Alternatively, driving the enemy out, or occupying their former position, AFTER support fire has already killed, demoralized, driven them out, or at least, “suppressed” them.

    Infantry tactics are far simpler than they used to be. We used to teach various tactical bounding and flanking movements, but back then, the infantry was fighting essentially other infantry. Today, we expect support fire of all kinds, including direct fire support from the 25mm on the vehicle that we deploy from, and direct frontal assault is more common.

    I think Kirk has it right when he describes the German way of war, MG and crew served weapons centric, rather than the traditional rifleman centric doctrine that we still to a certain extent hold onto. If you watch video of engagements in Afghanistan, what you often see is infantry taking cover, and engaging the enemy, but not really attacking/assaulting, but rather locating and keeping them busy while the CAS arrives. In that context, does it really matter how “effective” and “accurate” our smallarms fire is, when most of the killing is done by aerial bombs?

  14. Bruce says:

    Kirk, thanks.

  15. Kirk says:


    Every time you post, you just keep teasing that infovore that lives between my ears. I would so love to go digging through the results of that work you describe, but I’ll bet money it’s buried deep in the system and unavailable to mortal man.

    I don’t profess to knowing what the hell really goes on in combat, any more than the next idjit. I just know BS when I see it, and I also recognize how little real work has been done in this regard–Reading your description of what your work was does give me hope, though.

    The fundamental problem here is the same one that bedevils so much of modern life–The theorists have taken over everything, and they have so thoroughly polluted the mental commons that people are unable to see past their failed ideation of reality. You come to an issue with preconceived notions, you’re going to template it along those lines, and if you’re starting from flawed theory, it isn’t very likely that you’re going to be doing much more than add to the confusion.

    When it comes to combat effectiveness, I think there are a number of things going on: Just like with most fields of human endeavor, there are always going to be the “10 percent” that just about everything in human endeavor relies on. There are always certain individuals that provide the yeast, the movement, the shaking, the bit that gets things done. And, who carry the other 80% along to success.

    This isn’t a hypocritical rehash that agrees with Marshall or Grossman, either–What I’m saying is that the statistical phenomenon that they observed exists, it’s just that they got the causes completely wrong. It’s not that the 80% is reluctant to kill; it’s that they’re that component of society that just goes along to get along; give them a chance, and they’ll happily kill if that’s what everybody else is doing–But, they’re not going to motivate themselves or anyone else to go do it, because that’s what they do. It’s like that across society–You have key influencers that somehow manage to shift entire social movements, just ‘cos they make it look like the thing to do. If you look at human behavior from a marketing standpoint, and then do a cross-correlation to behavior in combat, there you go: The 10% is that bit of the equation that you have to take into account, because they’re the guys who’re going out in front, taking the initiative, and doing the actual “thing” that needs doing. They’re like those first few rocks in an avalanche; inertia keeps the hillside solid, and then one or two rocks fall, then a few more, and finally the whole damn thing falls on your ass.

    What is crucial to military leadership is identifying these guys, and ensuring that you co-opt them for the good of the mission and unit. Not all of them can serve as formal leaders, either, but they’re the crucial guys who every junior soldier looks to for cues as to how to behave. “Oh, if Smith is doing it, then I’d better…”. Smith charges the MG nest, you’re golden. Smith turns tail and runs from the tanks…? You’re screwed.

    The really amazing thing is how one day’s coward can be the hero on another occasion, or how just a slight change in circumstance can completely undo someone who was a pillar of strength. An acquaintance of mine was an Emergency Services tech, someone who’d dealt with it all for years with no apparent psychic damage. He’d pulled people out of car wrecks, searched for dead bodies (finding them…), and all of that. Thing that broke him, though? One day, he’s working a car wreck, literally picking up the pieces, and he looks down at his feet to see a stuffed animal just like the one that belonged to his daughter. Bang. That was the last shift he could handle doing EMS work, and the way he completely lost his shit that afternoon freaked out everyone he worked around, mostly because he’d always been this vast bastion of strength and calm.

    People are the wild variable in every one of these equations, and I think it is an error of massive proportions to try to do what DuPuy did (along with a bunch of Soviets…), and try to reduce everything to a nice, neat equation. There are “best practices” that can be identified, as you allude to, like having leadership nodes with the major weapons, but… In the final analysis, you need to know your people and use them effectively. A really good leader knows who to put into place on his gun team, and who to have playing scout out in front, as well as where to place his flawed subordinates in order to get the most out of them.

    One of the major issues I see with regards to a lot of this is that this all comes down to “tribal knowledge” and it isn’t formally studied or taught; you have to pick it up by serving a long tuition as a junior leader, and graduate to a knowing seniority that quite often isn’t attained because the majority of the people who aspire to these things are not the introspective sort who bother to think deeply about what they are doing, or the systems around them. In my experience, there’s a severe lack of deep introspective thought that goes on down where it would need to, in order to gain a true understanding of these things. You need a John English, and there just aren’t that many of him running around. The latest aspirant is Anthony King, and I’m afraid he suffers from the same set of flaws that most academics take to these things–He’s never lived the life, and he doesn’t know what he doesn’t know. You want at the truth of the matter, you have to be that guy who is deathly tired after 72 hours on his feet, with thirty-five miles of cumulative marching under a ruck that’s two-thirds of his body weight, and then trying to organize a live-fire defense that actually does anything past expending some ammunition. You can’t extrapolate from dry academic “studies”, you have to have been that guy, before you really begin to even get an inkling.

    Make me “King for a Day”, and I’d go out and find some bright young men in the forces, identify them after a successful tour in the ranks, sponsor their way through a truly objective course of study to qualify them as field sociologists, and then send them back into the forces to try to winkle the secrets of these realities out. I’d also start wiring units for sound, send them off to war, and then gather the data from actual engagements they were in, in order to try to get at the actual ground truth of what the hell was happening. Precisely which fires impacted the enemy the most–Not just supposition and inference, but fact. What works, what doesn’t, and which systems do the most damage to the enemy, as well as which are a complete waste of time.

    I’m morally certain that while a good deal of what we think we know would be proven out, there’s also a potentially massive set of “esoteric knowledge” that we just don’t know, and will be shocked by.

  16. Alistair says:


    Actually, the public domain stuff got published, or I wouldn’t be discussing it on here. The author is the senior engineer I mentioned. Great guy; retired now but I have immense respect for the work he did driving historical analysis almost single handed. He never got the recognition for it; the management squawked about rewarding technical excellence whilst actually rewarding smooth-arsed charlatans with meaningless influence diagrams.

    Don’t buy it: I have several spare copies lying around if you would like one (leave an email and I’ll contact you; or get Isegoria to give you mine – this post constitutes permission.). It is about as readable as telephone directory in parts, but stuffed full of interesting charts. I think you’d find it interesting.

  17. Kirk says:

    Alistair, I would cheerfully throttle the clown or mime of your choice to get my grubby mitts on a copy of that work, which looks to be seminal.

    What I can’t figure out is how the hell I’m only now hearing of this, when it was published in 2006. Yeah, esoterica erotica for the military mind, but… Nobody I’ve read has cited it, and I’ve never run into a review. Just goes to show you.

    Couple of works I wonder if you’ve run across, now: John English’s On Infantry, the original version that grew out of his doctoral thesis, not the tarted-up and butchered version that later got published by Praeger and “co-authored” (butchered, really…) by Gudmundsson. Then, there’s The Combat Soldier, by Anthony King, where I note that this work by Rowland isn’t even cited in the bibliography, which you think it would be, given the muchly-touted status of King’s work as being definitive.

    Ah, well… Scholarship in this area sucks. Really, really sucks. In the US, it’s a flock of dilettantes like Grossman who’re stone-ignorant and pretty much about as dumb as posts, when it comes to historical antecedents, and all the really switched-on types do one or two papers while they’re at Leavenworth, and then abandon the field to the lightweights. I presume that’s because the scholarly tradition here in the US is inimical to real military history and inquiry on the matters pertaining to it, but who knows? All I can say is that when I find stuff that’s really well-done, pertinent and well-thought out, with good research and no faddish false “conclusions” that they started out with, it’s usually from a Canadian, an Australian, or an Englishman.

    Thing I find really annoying, down the years: Nobody, but nobody, has gone back and done the research to find, translate, and then publish in English the work I know the Germans did during the interwar years. I ran into a gentleman, years ago, at the Lake County Gun Show (Illinois Nazis all over the place, there, and strangely ignored–Bastards once had a booth selling concentration camp memorabilia ferchrissakes’…), and he’d picked up a fascination with German MG tactics, techniques, and history. While he’d been in Germany from the late ’50s to the early ’80s, he’d spent a lot of his spare time in the archives, out collecting stuff at used-book stores, and doing solid research on the subject. When he’d finally retired from being a Department of the Army civilian, he’d gone home to Lake County, Illinois to put it all together and publish it. When I ran into him, he’d been selling off some of his excess “stuff”, like rangefinders and tripods, at the gun show. Holy crap, but did he have a treasure-trove of stuff, including 8mm movies and a ton of still un-translated material that had come out of the archives and various other places he’d ferreted this stuff out of. His intent was to write the definitive work about German machinegunnery from the tactical/operational standpoint, not the typical Wehraboo gun nut one that focused strictly on the guns themselves. I cribbed a bunch of notes from him, and made a note to buy the book when he published, because it was gonna be amazing.

    I should have done more, looking back on it, because this guy died not much after that, and I had moved on to another assignment when that happened. One of those casual meetings you have, ignore as interesting-but-unimportant, and then later recognize as a massive wasted opportunity.

    I presume his collection got broken up by his heirs, or went to a dump somewhere when they cleaned up his house. The man had literal tons of material he’d collected, and then had shipped back to Illinois from Germany.

    There’s a huge void in the scholarship available in English. I doubt that its available in German, either, except in the original sources, because nobody is looking at this stuff. It should have been, but wasn’t.

    Ah, well… If I ever win the lottery, I’ll go looking for it. If not, maybe someone reading this will be inspired.

    So, yes… I’d absolutely love to exchange contact info with you, if Isegoria would be so kind as to expedite the matter. This book by Rowland sounds amazing.

  18. CVLR says:

    Kirk: “(Illinois Nazis all over the place, there, and strangely ignored–Bastards once had a booth selling concentration camp memorabilia ferchrissakes’…)”

    …Say what now?

  19. Kirk says:

    Lake County Gun Show, circa 1990-ish. Literally had a booth selling concentration camp memorabilia and books discussing the Holocaust, and not from a “…this shouldn’t happen ever again…” standpoint, either–The general gist of what the asshole running the booth was saying and putting out was “…ain’t this cool…”.

    Swear to God, he had some of the crap the guards made out of tattooed human skin on display.

    I still don’t know what to make of that whole thing. Jewish friend of mine at the time was a retired Chicago PD cop, owned a gun store, and I delicately asked him about it, and the response I got from him was seriously weird. I was left wondering whether or not the whole deal was a false-flag operation by the ADL, some sort of law enforcement scam, or genuinely a sick bastard who had a concentration camp fetish.

    That whole region of the country is just shot through with the strange and unusual, in a lot of ways.

  20. Kirk says:

    Brian Dunn, over at The Dignified Rant, has put up a post referencing some of these issues, and he’s linked three of the better articles outlining the essential fraud perpetrated by Marshall.

    Links are down near the bottom; I’d recommend reading them to all and sundry, particularly the one where they interview another historian who’d been an escort officer for Marshall. Note especially where the light comes on over his head, and he realizes just how badly he’d been had by Marshall, and how thoroughly Marshall exploited the trust of everyone around him, as well as his sloth and essential incompetence around the battlefield. This is a guy who needed keepers to provide him with the basics like a helmet…

    Marshall was a fraud, from start to finish. Some of his work does possess some value, but the majority of it…? I would stay the hell away from citing anything that crooked bastard had to say. Where he was right could be easily explained via the stopped clock paradigm.

    There’s also a lot more out there–These cites by Dunn are just the tip of the iceberg. I’ve got a file folder somewhere full of a bunch of other stuff I’ve run across, over the years. One of these days, I’ll have to digitize it–The resources outlining the actual genesis of the Trainfire program are quite telling, when juxtaposed with Marshall’s self-adulatory prose taking credit for that program.

Leave a Reply