Schneier on Security
A blog covering security and security technology.
« Four Irrefutable Security Laws |
| Haystack »
September 20, 2010
Statistical Distribution of Combat Wounds to the Head
This is interesting:
The study, led by physician Yuval Ran, looked at Israeli combat deaths from 2000 to 2004 and tracked where bullet entries appeared on the skull (illustrated above), finding that the lower back (occipital region) and front of the temple areas (anterior-temporal regions) were most likely.
I'm not sure it's useful, but it is interesting.
Posted on September 20, 2010 at 1:58 PM
• 45 Comments
To receive these entries once a month by e-mail, sign up for the Crypto-Gram Newsletter.
What the heck, I'll throw out a theory from the peanut gallery. This theory consumed all of 15 seconds during development.
Soldiers are hit in the back of the head because they are hit when their head is ducked.
They are hit on the temple when they lift their head up to see what is going on.
In both cases, the area struck by the bullet is the most prominent portion of the head at the time of impact.
Headshot, an occupational hazard.
Looking at the picture presented, it looks to me entirely random. Humans are pretty good at picking up patterns, and we have a tendency to find them when they aren't there. I'm pretty sure if you generated a few dozen images with uniform random placement you would see quite a few striking patterns emerge.
One small point. We would need to see data on ALL of the headshots (including non-fatal ones) to make real sense of this. It could be only that FATAL shots are likely to be at the back and sides of the head and non fatal shots are more to the front. The back of the head is where the autonomic nervous system lives. The extrapolation from lethal shot location to the overall "hit" distribution is flawed.
If you were designing helmets, this might be helpful to determine where to reinforce the helmet.
After reading the linked post, I thought exactly the same as statistics...
Fun fact: after soldiers started wearing helmets, the count of head injuries rose sharply. Simply because less people died.
Another thing that might be a contributing factor is that if you hear a gunshot and duck, you might move forwards and downwards. In any case, I agree that it's an interesting tidbit.
Useless. Without at least 1000 samples there is no way to get meaningful results
The modern kevlar laminate helmet does protect the wearer from some bullet strikes, especially in the case of glancing hits, strikes by bullets that have already passed through or bounced off other materials, and because pistols and carbines (7.62x39) strike lighter blows than full-power rifle cartridges used by snipers and machineguns (7.62x54R). The areas not as well covered by this admittedly limited protection are the areas where the study found the hits were concentrated.
Note that Mindhacks seems to have misinterpreted the study he cites http://www.ncbi.nlm.nih.gov/pubmed/15345967 which addresses blunt trauma injury caused by backface deformation when the helmet stops a bullet but flexing helmet materials still strike a dangerous blow to the skull and brain of the wearer.
I also wonder if these hits aren't the result of aiming for the torso of the soldier; a shot aimed at a soldier's chest that lands only a little high will strike the neck, jaw, or lower part of skull more often than the top of the skull.
And the head of a moving target is notoriously hard to hit in a hurry even when aiming at it deliberately. The closer to the center of gravity of the target the more likely you are to make the hit; the top of the head is bobbing and weaving more that the part attached to the neck.
Seems to me the researchers would need to understand how many other hits (not necessarily fatal injuries) occur to the jaw, neck, and unprotected shoulders and upper chest before dismissing the effect of distribution of hits. PubMed points to a related study http://www.ncbi.nlm.nih.gov/pubmed/15995476 that found injuries to the neck and face were the most common cause of fatal injury to IDF soldiers.
I looked at the photo and after thinking about it some, it does make sense. I was trained at Ft. Benning, GA when I got my commission. Setting up defensive positions for small arms and crew served automatic weapons involves overlapping fields of fire. An opposing infantryman taking cover would be most often hit in the head in those areas, helmet or no helmet. This is because the shooter would be at an oblique angle to the target. The helmet provide only marginal protection against small arms fire and virtually none against a 7.62 mm or larger crew-served weapon. The helmet just causes the bullet to expand sooner.
I think this also makes sense in both an open field maneuver or urban combat.
just my tuppence.
It might be interesting to compare those with the statistics from 'Cranio-maxillofacial injuries in Homer's Iliad' (http://linkinghub.elsevier.com/retrieve/pii/S1010518207001254) or 'Cranial trauma in ancient Greece: From Homer to classical authors' (http://linkinghub.elsevier.com/retrieve/pii/S101051821000034X). However, both of those are behind a paywall.
Since this paper was using Israeli soldiers for it’s sample size, we should remember that Israeli soldiers are largely engaged in urban guerilla style warfare. When engaging in such a fight, the shooter, being heavily outmanned and under resourced will be trying to compensate with stealth. Therefore they will choose to try and shoot soldiers from behind, which explains the concentration in the back. As for the front, I can only assume those are situations where the soldiers are engaged in a firefight and therefore they are acing the enemy that they are firing upon.
In response to Bruce's comment about usefulness, this is very useful. If you design helmets or body armor for a living.
I think "statistics" nailed it with their comment earlier. There's a pretty obvious statistical bias here. Even if a pattern exists in this data, it's useless if you don't account for the fact that we're only looking at the fatal wounds.
> ...if you hear a gunshot and duck...
Most bullets travel faster than the speed of sound. Even with subsonic rounds, you'd still need some amazingly fast reflexes...
Sound fades quickly. At long ranges, you won't hear the gunshot. You'll just see sand/dirt puffing up around you where the rounds are landing...
They found clear patterns because humans tend to "see" patterns in random distributions.
Hanging on a chain to remind me is a 7.62 FMJ round that is bent and distorted and caused a small burn on the back of my neck when it went down inside my combat shirt.
Just befor that it had hit the underside edge of the helmet I was wearing and chewed up the webbing strap.
Was I in combat? No I was on a rifle range, how the hell the round bounced back to the 100yard firing line I will never know.
Did it have my name on I guess not but judging by the mess it made of the helmet, I used up a couple of cats worth of lives.
That said there is a UK soldier who was even more lucky just recently. He was hit in the head by an RPG round that bounced before exploding and he got wiplash and minor shrapnel wounds...
I guess you can say these things just happen.
statistics is completely correct, without total number of head injuries we learn nothing.
And following up on Richard Hartmann's comment, I recall that after the introduction of airbags, the number of reported driver foot and lower leg injuries went up. It turned out that these injuries were not reported when the driver was killed. When airbags started saving lives, these injuries suddenly were reported, but they were happening all along.
@Clive Robinson - Thanks you for sharing this story glad you did not get hurt and you had a helmet.
In the commnets of the link there is a great commnet that explains the frontal hits. Soldier usually have a rifle/carbine in frot of them. With the head tilted a bit to the right. The weapon shields the center area.
Maybe there is communications gear on the side of the head?
And the rear impacts are ambush/shots from behind.
Interesting study. It's just like how the bottom and leading edges of the space shuttle have a highly dense material impregnated into the ceramic tiles (making it black) which dissipates heat better, but also seriously adds weight, which is why the entire shuttle is not black: they only put the material where it's most needed. The article says that the same approach could be taken to design a helmet that is not significantly heavier than what they already use.
I do not know why Schneier posted this history here, but I think it is related to perceived risks and statistical risk handling. BTW, a similar study was conducted by the German army in the World War I and caused them to replace the classic Prussian leather spiked helmet (Pickelhelm) by a steel helmet with a larger protection band behind the neck (Stahlhelm).
> Most bullets travel faster than the speed of sound. Even with subsonic rounds, you'd still need some amazingly fast reflexes...
Where there is one gunshot, there well may be second or third one coming soon. Maybe adjusted according to the place where the first bullet struck, if it missed...
a) Not enough data to produce conclusion,
b) Data needs to include wounding shots,
c) Data needs to consider shot entry+exit vs fatality,
d) Data needs to consider type of warfare and firing location - snipers in cities will be 400m directly in front or behind and 10-20deg elevated, rarely the 30m to the side.
Other than that, interesting - but the rule holds "if your number is up, your number is up."
At first glance, this says to me "dont let the enemy be behind you". I would also like to know how many of these troops were actually wearing the helmet at the time of impact.
When my Combat Communications unit deployed, we would be required to take off the Kevlar helmet and wear bright plastic hard hats while unloading the trucks and (cargo) pallets.
While I realize they are designed for different missions (large mass/low velocity/ultra cheap vs. small mass/high velocity/GSA purchasing) - I really cant picture your life being saved by the flimsy plastic hard hat thats not fastened on when you would have died from the same hit in the Kevlar (which has ~20 times the mass, much more exotic material [ie can be designed for multiple missions], better webbing and is fastened on to stay with you after a first hit that would send the hard hat flying). Not to mention if you were in an actual pissed-off-people-with-rifles area wearing a bright yellow hat in lieu of camo would be suboptimal.
I always assumed it was because some Congressman's brother had a company that made hard hats.
A white face looking up at a helicopter from leafy cover is quite visible to those on the helicopter. the common military headlice shaved haircut makes the whole head a white dot to shoot at. helmets add some camoflage and the white shows beneath that which is where the rounds seem to have hit. It would be interesting to know whether the shots that hit were of the spray and pray style of our afhgan allies or were sent by people who use gunsights. Most spray and pray rounds go high over the battle. Aimed rounds from a distance can be thwarted by effective camoflage to prevent target discovery though it seems that much of israeli combat is urban or close range where neither the helmet or camoflage will help there.
If you hear a shot duck is good advice, because the shot missed but it told you that someone is shooting, and that person may have friends anywhere in 360 degrees around you, after ducking crawl and don't stick your head up within six feet of where you ducked if possible because the movement of ducking may have attracted another concealed and unannounced shooter to that spot where he is now sighted on waiting for the head to reappear. Look around the base of any cover, rather than pop up above it.
You will see less but less will see you and you can look at the other areas from another low position.
Wasn't there something similar done with airplanes in one of the world wars? They analyzed the bullet holes in the returning aircraft to figure out how to armor the planes better... and since the planes they were analyzing made it back, they figured the planes that didn't make it back had bullet holes where these planes _didn't_. Seems a similar approach may be useful in this case, despite the clamor about needing to know all the non-fatal wounds, etc.
In WWII, an allied airforce made a study of where bullets typically hit aircraft in combat to help decide where to place armor. An operations researcher then pointed out that you wanted to know where bullet holes *weren't* found, because those were the aircraft that didn't return.
Assuming this is accurate data, it's good for people who
1. design armor (even better if it includes nonfatal shots) and
2. med personnel dealing with head wound patients, as location of damage sometimes correlates with nature of brain injury.
radial engines could continue to operate with jugs shot right off them, the pilots seat is what needs to be armored, and if you need some protection in a car, its good to realize this, back of seat, headrest and door should have some armor belts. its not the total armor that is sold by the big companies, its more like an uparmored humvee idea. most armor is heavy like half inch steel plate, but if you protect the driver from behind and one side you can use less of it.
this isn't perfect, but better than nothing.
They studied fatal wounds.
So the data doesn't show where soldiers get shot.
The data shows which bullet entry points are fatal.
Getting shot in the lower back of the head is specially bad because the bullet tears up the cerebellum, which then swells up and compresses the brain stem, which is fatal.
Many U.S. soldiers in Iraq & Afghanistan have been killed in this way; not by bullets, but by blast and explosion to the back of the head. There is a proposal to extend combat helmets further down the neck to protect against this.
I don't know why getting shot in the temples is specially bad.
I'm reminded of a story - possibly apocryphal - that was told as part of a stats class at a UK university.
During WW2, the Brits applied a lot of data analysis tools in some sometimes-unexpected ways, often as part of larger cost-benefit analyses.
One approach they took was to map the location and severity of damage to bombers as a result of flying missions over Germany. By this means, it was thought, a cost-benefit analysis of weight of armour plate vs reduced bomb load vs number of airplanes lost might be performed, and a calculus developed for what sort of armour should be applied, and how much, and where, for the maximum benefit.
Until some bright spark pointed out that the data represented only those aircraft that made it home - which apparently had no need for further armour at all.
I think this analysis is best described as a definition of which locations of head wounds are the most likely to be fatal. As others observe, it would be both more interesting and more informative to see an analysis of all head wounds, regardless of outcome.
Should have read all the comments, shouldn't I . . . .
That's right: concentrate your armour plating where the bullet holes AREN'T seen.
"a cost-benefit analysis of weight of armour plate vs reduced bomb load vs number of airplanes lost might be performed, and a calculus developed for what sort of armour should be applied, and how much, and where, for the maximum benefit."
IIRC the the Avro Lancaster, a British heavy bomber, had only one pilot and its only armour was an eight disk of plate behind his head. Seems the analysts weren't trying to prevent the pilot from being killed, only from being killed so quickly as to lose control of the aircraft and denying the crew time to bail out.
Perhaps it was a similar cold hearted mathmatician who determined that blowing up German factories by day was not nearly so effective as burning down the factory workers' homes by night.
The study of where bullets tend to hit is nothing new. There has been a lot of research in this area, primarily during WW2. There were some extensive studies of the distribution of flak damage to airplanes, in a try to reduce the loss rates. See for example http://history.amedd.army.mil/booksdocs/wwii/...
Once (15 years ago) I talked to swiss guy on an industrial trade fair who selled kevlar protective equipment.
I asked him why helmets weren't protected with his stuff.
He simply said: we could, but at some bullet velocity and caliber the impact would break your neck...
'IIRC the the Avro Lancaster, a British heavy bomber, had only one pilot and its only armour was an eight disk of plate behind his head. Seems the analysts weren't trying to prevent the pilot from being killed, only from being killed so quickly as to lose control of the aircraft and denying the crew time to bail out.'
True enough, the later marks of Lancaster had only a sheet of armour plate behind the pilot's back and head, which can usually be identified by a bright yellow circle painted on the front. But earlier marks, and other aircraft models (Wellington, Hampden, Manchester, Stirling) had a variety of armour plating in various locations, around the main crew areas, in the turrets, and elsewhere. This was often removed in service as tending to add weight and offering no real protection from essentially-random projectiles. Lancaster pilots were known to fly using a flak vest as a seat cushion, since they had no armour below the waist.
The US B-17 day bombers had a lot more armour plating (which also varied by mark and model) but much of this was removed in the field also. I suspect that this hodgepodge of approaches is what may have led to the diea of doing some actual operational analysis.
The cold-hearted analyst who determined that it was better to bomb the workers out of their homes than out of their factories is probably cousin to the one who calculated that the best way to reduce the combat effectiveness of an army is not to shoot the soldiers but to infect them with venereal diseases.
"statistics" is right on the money. Without the non-fatal distribution, this is less meaningful.
The "medulla-shot" is one selected by short-range snipers to deal with hostage situations. It prevents the hostage-taker from being able to pull the trigger.
These sorts of studies are incredibly useful, that's how we got (most of) our current military helmets. During WWI Germany redesigned its helmets based on head wound statistics and created the characteristic German late WWI/WWII helmet shape, while others stuck with their historic designs like the British tin hat, chosen purely for ease of manufacturability and more of an antipattern for protective headgear design. This actually proved a problem for others because what was a reasonably optimal helmet style came to be associated with the German military, so anyone else who wanted to design a good helmet faced the dual constraints of having to minimise head wound problems *and* not look like the German one. The US, for example, rejected some very good designs because they looked too similar to the German helmet. The PASGT helmet still looks like it was derived from the German WWI design, not because it was but because that's a reasonably optimal shape for a military helmet.
Interesting to compare to (fatal) medieval battle injuries found on skeletons from Battle of Towton, c 1461. These were concentrated on left side of the head and shoulder, which makes sense for soldiers facing right-handed combatants.
My uncle could have told you this for the price of a cup of Earl Grey. Not because he was a sharpshooter for the Black Watch and represented Canada in the Commonwealth Games and as an honour guard for the Queen's visit, but because he spent twenty years working in an abattoir shooting horses.
And if you ask him nicely, he'd probably show you how it's done on a moose or deer.
And give you some even more important advice and that is to shoot them as close to the barn as possible so you don't have to lug the carcass far. If you brought him some pie or smokes, he might even get into telling you the story of working as a guide and the dumb American that wouldn't listen and couldn't wait for the moose to get out of the river before he shot it.
Moose are really heavy when wet. Which, while on the topic, I've heard is why the US went for small bullets on the M-16 as it's been calculated to cost the enemy more to maim rather than kill the enemy. Anyways, the thing floated down stream and he had to chase it. I'll leave it at that not to spoil it and suggest you bring Pecan or Apple pie and he likes that tobacco that comes in the blue bag. But he's also always trying to quit, so don't mention it until he's brought up the topic of having a smoke.
His secret on how to butcher the meat to make it taste like pork, you'll have to go ask him yourself. If he likes you he might tell you. But I tell you, he'd always impress my wife from New York who is quite the chef and has a general distaste for the 'gamey' flavour of most venison she's eaten prepared by 'sportsmen'. But she loves what he calls 'Brushy Beef'.
The biggest benefit from putting on a helmet is to dehumanize the appearance of your troops, which makes it easier psychologically for the enemy to shoot them. Yes, it's counter-intuitive.
Police tend to eschew helmets except in riot work. While the skull is fragile, covering the ears takes away a critical sense and wearing a helmet all shift is tiring.
In an environment full of fragments from explosions (shells or IEDs), it's a different equation -- but the efficacy of a helmet against a rifle bullet is marginal at best.
That WWII bomber survivorship bias story sure gets passed around a lot these days. I remember it came from the introduction to a book, the name of which now escapes me. Wonder if anyone has the first source reference.
Anyway this is a study of soldiers killed so it would be the opposite of the WWII situation, where the study was of the survivors.
Amusing cartoon of the B-17s survivability:
Can't help thinking that those are areas a helmet does not really protect.
1) You want a helmet that allows you to turn your head to look up, so that's the back of your neck and lower back of the head exposed.
2) You want a helmet that allows you a full field of vision, which leaves the temples at least partially unprotected - especially from shots coming from front and little to side. If you're sniping, front works well because there's no protection and a little on the side is even better because they don't see the shots coming. Of course, the edges of the helmet deflects shots and temples are eye-level, so if peeking over anything or around a corner...
WWI British headgear was designed for trench warfare, e.g. artillery and mortars. That's why it was widely referred as a "Shrapnel helmet".
The German helmet was designed to protect against blows in hand to hand combat. Neither would likely stop a bullet.
By the end of WWII both designs had evolved, to better suit a more mobile battle. But the current crop still is quite heavy & awkward.
Read all of the comments, been surprised that no one thought of the simple fact that you don't shoot someone in the face.
The fallen soldiers in this study where essentially Bouncers. On one side of the gate the exclusive Club, on the other the cold and dark Street. Disgruntled men who are not admitted congregate at the corner. Sometimes one of them climbs up and fires. Most likely when they look in the opposite direktion so you don't have to look them in the eye.
Schneier.com is a personal website. Opinions expressed are not necessarily those of Co3 Systems, Inc.