The effects of missiles - how important were armour and shields?

[Patrick Waterson]

Patrick what is the Latin word used for bristling in your Ammianus quotation??
Roy

horrentes, meaning ruffled, rough, bristling

Ammianus writes: horrentes indutibus rigidis, literally 'apparel bristling with stiff things', which I suppose is one way of describing scale mail.  One MS adds 'aere' (of bronze) before 'rigidis'.

[Nick Harbud]

Question for Holly - what were the max ranges you were capable of shooting?  And what did you consider to be your max effective range?  I bow hunt here in the US and I will never take a shot longer than 40 yards (well, okay, I once shot at a mountain lion at about 60 but just because he was stalking me and I was getting nervous - missed by the way but got close enough that I scared the cat off).  I don't know anyone who would take a shot at an animal beyond 50 yards and expect to hit - a good hunter wouldn't take that shot because even if you hit the animal chances are you won't kill it.  I shoot a bow with a 30" draw, 55 pound pull and about 280 feet per second speed at release.  My son (who is a bit taller than I) shoots at 70 pounds and about 310 FPS.  We killed two elk this year - both at about 30 yards.  So I'm curious what y'all learned in the reenactment business compared to what we do in the hunting business...

You make a valid point, certainly with respect to individual effective range.  FWIW, in modern target archery one shoots at targets of between 80cm and 122cm diameter at various ranges between 40yds and 100yds.  Field archery, which aims to provide a competition with greater resemblance to a hunting trip, sets its targets at variable distances up to 80yds.  Often in field archery the range is not marked.  At longer ranges the effect on accuracy is more pronounced.  In clout archery competitors shoot at a post (the clout) up to 200yds away and typically any arrow landing within 7 meters scores.  Indeed, the above is one of my supporting arguments in favor of the hit probability curve that featured in the earlier postings.

With regard to bow strength and range, most people researching reconstructed longbows seem comfortable with 140lb draw weight, although bows recovered from the Mary Rose are estimated to be up to 180lb.  A replica of a 150lb longbow found on the Mary Rose was able to shoot up to 360yds.  (See The Great Warbow.)  Also see Secrets of the English War Bow page 44 for ranges achieved with different arrow types and bow strengths.  Stretton reckons that each additional 10lb draw weight adds 10yds to effective range.

Most of the replica arrows have a weight in the range 53-95g (2oz to 3.3oz) with a drag coefficient between 1.8 and 2.1.  (Although Mark Stretton developed a quarrel-type bodkin that was very good at penetrating plate armor, which weighed in a 125g.)  A good initial velocity is around 60m/s.  I guess you are using a compound bow to achieve 280-310fps?

[Patrick Waterson]

... Often in field archery the range is not marked.  At longer ranges the effect on accuracy is more pronounced.  In clout archery competitors shoot at a post (the clout) up to 200yds away and typically any arrow landing within 7 meters scores.  Indeed, the above is one of my supporting arguments in favor of the hit probability curve that featured in the earlier postings.

Individual accuracy at range is of less importance when a unit of 200+ archers are all shooting indirectly at the same approximate point of aim.  The idea behind clout shooting is to saturate an area similar to the space occupied by the typical enemy target - hence although at the ranges typically involved only Robin Hood could hit an individual target, the arrival of so many shafts with a fairly narrow 'circular error probability' is going to be bad news for quite a few people in the 'beaten zone'.

Hence the accuracy curve and the hit probability curve will be two different and essentially unrelated things.   In WW2 the USAAF went for accuracy (yes, that is what the Norden bombsight was for) and the RAF went for hit probability - hitting a city was easier.

[Nick Harbud]

Hence the accuracy curve and the hit probability curve will be two different and essentially unrelated things.   In WW2 the USAAF went for accuracy (yes, that is what the Norden bombsight was for) and the RAF went for hit probability - hitting a city was easier.

Have I understood your argument correctly?  30 men firing one shot will score more hits in specific area than 1 man firing 30 shots?

Incidentally, Bomber Command is probably not such a good analogy to support your argument.  I mean, when the RAF wanted to hit a specific target rather than any large urban area they went for accuracy.  For example, Peenemunde or the dambusters raid.

[Patrick Waterson]

Hence the accuracy curve and the hit probability curve will be two different and essentially unrelated things.   In WW2 the USAAF went for accuracy (yes, that is what the Norden bombsight was for) and the RAF went for hit probability - hitting a city was easier.

Have I understood your argument correctly?  30 men firing one shot will score more hits in specific area than 1 man firing 30 shots?

No, sorry, I explained it poorly.  The point is that 300 men trained to shoot together on command at a given range will drop 300 shafts into a crowd of enemies and cause losses at ranges where 300 men shooting individually would be unable to hit individual targets they were aiming at.

Incidentally, Bomber Command is probably not such a good analogy to support your argument.  I mean, when the RAF wanted to hit a specific target rather than any large urban area they went for accuracy.  For example, Peenemunde or the dambusters raid.

Or the 44 (Rhodesia) Squadron and 97 Squadron raid on the Augsburg MAN diesel works in 1942, or the Amiens jail breakout in 1944, or the Shell House Copenhagen raid in 1945, but these were exceptions.  99% of the time it was a case of the Pathfinders mark the city (usually the right one) and everyone else unloads on the markers.  Interestingly enough, the US strategic bombing offensive against Japan went the same way: after six months of repeatedly failing to hit factories through traditional attempts at high-level accuracy, Curtis LeMay took over, adopted RAF bombing doctrine wholesale and went for hit probability instead - massed night attacks against cities with incendiaries at low altitude - and although accuracy per se was near zero, the percentage of hits was around 100%.  When a lot of rounds are dropped together on a large target it is hard to go wrong.

[Mark G]

The pacific bombing campaign is not a safe comparison with Europe - the aircraft available had a much lower altitude, and le may took the decision to go in fast and dirty.

that many houses were wood and paper helped in the decision.

[Patrick Waterson]

The approach was still a shift from high-altitude precision attacks by day to area incendiary bombing by night at medium altitudes, discarding attempts at accurate bombing of specific targets in favour of massed bombing of targets that offered a much greater hit probability.

The point (hopefully) being made is that an accuracy curve is not necessarily the same as a hit probability curve, given a suitably-sized target.  If preferred, I shall go back to battleships and salvoes as a means of illustrating the point.

[Mark G]

Actually, I may have mis-remembered the details - it may have been the reverse - given bombers which could fly higher than the Japanese air defences, and thus, either fly low and accurate and be under threat, or go high, inaccurate, and have minimal casualties - hence, higher and with no thought to accuracy (paper houses no doubt added to that calculation).

either way, the point is, in Europe there was a thought to accurate bombing, in the East their was not.

the source was Robert MacNamarra's documentary film, BTW - well worth taking a couple of hours out to watch if you ever get the chance - its not often you get to hear directly (and AFAIK without any political calculations behind the answers) from someone who ran strategy for two wars.

[Patrick Waterson]

US strategic bombing in the Far East used B-29s, which could fly at and bomb from 30,000 feet with accuracy - over the Nevada ranges.  Over Japan with its turbulent jetstreams bomb strings were deflected on the way down so only about 5% landed close to where they were intended.  Because the jetstreams were unpredictable (read: not really known about yet) there was no effective way of compensating for them.  Ironically, on the very last mission on the very last day of the high-altitude precision bombing offensive the winds were for once in abeyance and the target was hit with almost perfect accuracy.

Because most of the secondary component manufacturing firms were located in Japanese cities, Curtis LeMay decided to use the RAF approach and area-bomb the cities, eliminating component manufacturing firms and the skilled workforce at one stroke.  No longer would the bi-ni-ju-ku (B-29s) delight Japanese watchers with their glittering high-altitude formations: henceforth, with armament largely stripped out to reduce weight, they would drone in by night, packed with incendiaries, and 'Hamburg' Japan's main industrial cities from 10,000 feet.

All of which brings us back to shields and shooting. (How? Don't ask - I just have to get back on topic. )  It would appear that in classical (and possibly Biblical) times the equivalent of controlled volley shooting gave, or was believed to give, effective results to the limit of bowshot.  Diodorus Siculus' account of Cunaxa (XIV.23) is interesting in this respect.

"When the armies were about three stades apart, the Greeks struck up the paean and at first advanced at a slow pace, but as soon as they were within range of missiles they began to run at great speed. Clearchus the Lacedaemonian had given orders for them to do this, for by not running from a great distance he had in mind to keep the fighters fresh in body for the fray, while if they advanced on the run when at close quarters, this, it was thought, would cause the missiles shot by bows and other means to fly over their heads. [2] When the troops with Cyrus approached the King's army, such a multitude of missiles was hurled upon them as one could expect to be discharged from a host of four hundred thousand. Nevertheless, they fought but an altogether short time with javelins [paltois, light Persian throwing-spears] and then for the remainder of the battle closed hand to hand."

Clearchus followed Nick's (and my) preferred technique of picking up the pace as the first volley was shot, causing it to land behind one's troops, and increasing the pace again when the second volley came over.  This works if the opposition are shooting indirectly en masse ('clout shooting') and, for that matter, if shooting with individual aiming at long ranges, but in each case relies on the opposition being reasonably good shots!  I suspect that at Cunaxa the average Achaemenid archer could not be expected to hit targets at 200 paces or so and that massed shooting techniques would therefore be employed - which might also go some way to explaining the Persian panic as the Greeks closed, the mass shooting techniques having been rendered ineffective, whereas archers shooting individually and directly would welcome a closer target and put shafts into it.  Xenophon's own account refers to plenty of dropped and discarded arrows lying around after the battle but none in Greek shields (or in Greeks).  Hence my belief that Achaemenid archers shot indirectly, en masse.

Now for a piece from the same author indicating that with missiles of sufficient weight/impact/penetration, shields and for that matter armour become almost an irrelevance (Diodorus XIV.27.4-6):

" ... the Greeks made their way for seven days through the mountains of the Carduchi, suffering greatly at the hands of the natives, who were a warlike people and well acquainted with the region. [5] They were enemies of the King and a free people who practised the arts of war, and they especially trained themselves in hurling the largest stones they could with slings and in the use of enormous arrows, with which missiles they inflicted wounds on the Greeks from advantageous positions, slaying many and seriously injuring not a few. [6] For the arrows were more than two cubits long and pierced both the shields and breastplates, so that no armour could withstand their force; and these arrows they used were so large, we are told, that the Greeks wound thongs about those that had been shot and used them as javelins to hurl back."

This suggests to me that in assessing the effects of armour and shields we need to consider not only the degree and quality of protection but also the missiles in use: some were more easily stoppable than others.  Carduchian arrows might laugh at Greek shields: Persian arrows seem to have been stopped by them, not so much at Cunaxa as at Thermopylae, Plataea and perhaps also Mycale.

[Nick Harbud]

No, sorry, I explained it poorly.  The point is that 300 men trained to shoot together on command at a given range will drop 300 shafts into a crowd of enemies and cause losses at ranges where 300 men shooting individually would be unable to hit individual targets they were aiming at.

Well the technique you describe is the same as the one that generated my probability curve.  The original researchers lined up a group of musketeers and had them shoot at suitably sized canvas screens set at various ranges.  In other words, the curve you see already takes account of such effects.

Now I accept that the bows are different from muskets, but both will have a hit probability curve.  So what do you think the curve for bows should look like?

[Patrick Waterson]

I would suggest the technique is different: for assessing indirect archery one should not have an upright screen but one laid flat to represent the depth of the target, which for indirect shooting will be more important than the height.

[Imperial Dave]

" ... the Greeks made their way for seven days through the mountains of the Carduchi, suffering greatly at the hands of the natives, who were a warlike people and well acquainted with the region. [5] They were enemies of the King and a free people who practised the arts of war, and they especially trained themselves in hurling the largest stones they could with slings and in the use of enormous arrows, with which missiles they inflicted wounds on the Greeks from advantageous positions, slaying many and seriously injuring not a few. [6] For the arrows were more than two cubits long and pierced both the shields and breastplates, so that no armour could withstand their force; and these arrows they used were so large, we are told, that the Greeks wound thongs about those that had been shot and used them as javelins to hurl back."

This suggests to me that in assessing the effects of armour and shields we need to consider not only the degree and quality of protection but also the missiles in use: some were more easily stoppable than others.  Carduchian arrows might laugh at Greek shields: Persian arrows seem to have been stopped by them, not so much at Cunaxa as at Thermopylae, Plataea and perhaps also Mycale.

The thing I guess to note here is that the Carduchians would presumably have to be firing at close range because the longer/heavier arrows would have a much reduced range compared to "standard" arrows?

[Nick Harbud]

I would suggest the technique is different: for assessing indirect archery one should not have an upright screen but one laid flat to represent the depth of the target, which for indirect shooting will be more important than the height.

A normal flat shot loosed at 60m/s requires an elevation of 22 degrees to hit a target at 200m.  At 45 degree elevation, it is necessary to reduce the velocity to 51m/s.  Incidentally, the latter shot soars to a height of around 60m compared with 24m of the flat shot.  The extra length of the trajectory means it takes roughly twice as long to travel to its target requiring to be loosed 3-4 seconds ahead of shooters in the front rank.

Notwithstanding all of this, what do you think the hit probability curve should look like?

[Patrick Waterson]

A normal flat shot loosed at 60m/s requires an elevation of 22 degrees to hit a target at 200m.  At 45 degree elevation, it is necessary to reduce the velocity to 51m/s.  Incidentally, the latter shot soars to a height of around 60m compared with 24m of the flat shot.  The extra length of the trajectory means it takes roughly twice as long to travel to its target requiring to be loosed 3-4 seconds ahead of shooters in the front rank.

Notwithstanding all of this, what do you think the hit probability curve should look like?

On the assumption that volley-shooting archers can land arrows within, say, 20 feet of the 'call' at 200 yards and keep this grouping down to c.100 yards, the number of shafts landing on the opposition will depend upon

a) the depth of the target
b) the ability of the caller to judge timing and distances (if he is no good he would not be doing this).

In practice, given a caller who knows his stuff, archers who have had adequate training and an opponent advancing at least 8 deep (8 yards/24 feet depth) one would expect 24/40 of the shafts to land on the target formation, and of those perhaps half would hit empty space, the remainder distributing themselves between shields, helmets, shoulders, throats and various other bits of temporarily exposed anatomy, which may or may not be armoured.  Hence expect about 30% of shots to be hitting targets, most of which will be sticking in upheld shields or glancing off armoured troops but potentially or actually incommoding or maiming anyone without any sort of protection.  Up the percentage by, say, 4% for every extra rank the target possesses up to 12 ranks (c.46%), after which the beaten zone is of less depth than the target so no increase is likely.  Conversely, drop it by about 4% for every rank less than 8, down to about 2% for a single rank.

Things become trickier as the range shortens below, say, 100 yards, because like a mortar an indirect shooter has to shoot higher to bring the volley down closer, and timing starts to become a significant consideration, particularly as the target will begin a charge at, say, 40 yards or so.  Archers will have to, or at least want to, switch to direct shooting, which will have a lower angle but will still spread the shot over the height and depth of the target, until the target is at point-blank (flat trajectory) at which point only the front rank or two of archers can realistically participate (and the majority may be wishing to betake themselves elsewhere unless provided with stakes, good close-combat weapons, etc.).

Hence I would see massed archery as being effective from c.200 yards down to c.80 yards, at which point it starts to become less effective and direct shooting more individually effective (better penetration at short ranges) but less effective overall (fewer shooters able to participate and trickier calling, especially if the target speeds up for a charge).  Unfortunately I have not been able to locate any studies on this subject as yet, so the above figures are 'guesstimates'.

It may be worth noting that at Cunaxa, the Greeks were faced by (assumed) Egyptians with large shields and yet were still shot at by archers.  This suggests that archers were deployed en masse behind the spearmen and shooting indirectly over them, i.e. would have to be depending upon called shooting (from a mounted officer who could see the target) and would be relying on indirect shooting entirely.

[Erpingham]

Unfortunately I have not been able to locate any studies on this subject as yet, so the above figures are 'guesstimates'.

I do have a copy of an article on the effect of pavises in absorbing arrows (in the sense of physically blocking them).  It has lots of diagrams but I don't know if it has serious maths.

I'll also, when I get the time, cut and paste some of Saxton Pope's experiments.