Film of Stuka 87G in action

[John D Salt]

Originally posted by JonS:

[snips] 168 Bde was ditched for NWE, and replaced by 231. I don't know what happened to 168 ... broken up for scrap maybe.

According to Joslen, after leaving 50 Div it became part of 56th (London) Division. It fought in Italy: Monte Comino, the Gargliano crossing, Anzio, the Gothic Line , Coriano and the Rimini Line. Finally reduced to cadres due to lack of reinforcements to the Med theatre on 24th Sept 1944.

All the best,

John.

[JasonC]

JonS - Not that I care too much about supplying Africa, but here are a few minor points about thruput.

No doubt the Germans and Italians managed their overland supply ops in Africa less than optimally. And the totals needed are probably daunting, particularly given the irrelevance of the theater as a whole, compared to the eastern front. But as a general rule you can increase the thruput of even a long chain by dumping more stuff in the back end, if you can choose the composition of the stuff added.

Say the extra tonnage includes 1000 trucks, 5-10 million gallons of gas, more engineer units, etc. They would increase the amount transported through the dump system. At El Alamein, truck supply is going to eat up at least a third of the stuff sent, in the chain itself. Half may be a more realistic estimate actually (you get about a third just for the gas). But the net tonnage reaching the other end might rise 2000 tons a month.

Thing is, a division in active combat might consume 200 tons a day. So a meaningful boost to delivered supply might require 5-10 times the above numbers. Probably not the most efficient use of scarce gasoline. Coastal shipping as far as Tobruk or Bardia would obviously save lots of gas, but it would take a major engineering investment at the ports, as well as the ships.

If you double the supply priority of the entire theater, you might supply a couple divisions reasonably at that distance. But the same resources would probably be much more useful e.g. in southern Russia. Similarly, with extra German first line AC, you can do more to keep British shipping out of the Med.

The whole theater just wasn't worth such investments. In fact, arguably the biggest mistake the Germans made in the MTO was sending a quarter million Axis troops to defend Tunisia, since it bought them only a few months and lost the force sent.

[dieseltaylor]

Would losing in Crete have made the Germans keener to involve themselves in Africa? I suspect it was a totally political decision to save Italy "honour" but then if they had received a good kicking in Crete perhaps a lot of the war would have changed.

Incidentally

"German airborne doctrine was based on parachuting in a small number of forces directly on top of enemy airfields. This force would capture the perimeter and any local anti-aircraft guns, allowing a much larger force to land by glider. Freyberg was aware of this after studying German actions of the past year, and decided to render the airfields unusable for landing. However, he was countermanded by the Middle East Command in Alexandria. They felt the invasion was doomed to fail now that they knew about it, and possibly wanted to keep the airfields intact for the RAF's return once the island was secure. This may have been a fatal error."

oops!

[Soddball]

Originally posted by JonS:

</font><blockquote>quote:</font><hr />Originally posted by Mike:

regardless of your quote, more suplies landed would have allowed for a supply line that was more capable of delivering more supplies!

I note that the tonnage given is quoted as "more than Rommel's current usage" - which of course he thought was not nearly enough........but which the supply line apparently managed to deliver OK.

No, you still aren't getting it. The Italians managed to get to African ports more than Rommel required, despite the best efforts of Malta. The problem was getting it from there, off the ships, and then to the front where there continued to be shortages, despite there being sufficient in theatre.

Simply getting getting more to Africa wouldn't have helped much - it would only have led to larger dumps in Tripoli, some 1,000 to 1,400 miles behind the front. That's 2-to-4 times the distance between Wellington and Whangarei. Travelling at about 10-20 miles an hour.

A.E.B.s point about coastal shipping is valid, but more-or-less the same result could have been acheived from bases in Egypt. Tripoli may have been out of reach, but the destinations of the coastal traffic weren't. Besides, Malta had done the damage to coastal shipping well before any serious consideration was given to capturing the it by the Axis </font>

[pamak1970]

All else is, of course, never equal.An article in the latest BAR (just arrived today) by Rowland, Rooney and Storr based on historical analysis from DOAC in the 1980s shows that correlation of forces is one of the less impirtant influences on battle outcomes. No surprise there, really; Biddle's "Military Power" arrived at the same conclusion, and dear old Trevor Dupuy suggested, only partly in jest, that "the smaller side always wins" would be a good rule of thumb in military history

All else in reality may never be equal but this does not make the theory invalid in pointing some principles.

Most theories make assumsions that do not exist in reality.

"All else" is not identical and static when you throw a dice in real life and each of your efforts takes place in a slightly different enviroment than the previous ones, but this does not discourage us from applying probability.

In fact the concept itself of probability does not fit in our real world,since it is related with infinite trials.

When i say that i have 50% chances , i do not expect each time i repeat a 100 -set of efforts to get 50 times the result expressed through the probability number.

The probability points that if i try to do something infinite times, then i can arrange results in infinite hundreds and each one of these sets will have 50 times the result expressed by the probability.

The other thing is that Lanchester does not talk about ratios only ,it talks about quality also and this is what is difficult to measure in reality.

Dupuy ,was one of the people trying to do it using historical data and here i want to point a common misconception.

Some times an author assuming that Lanchester law holds true,uses historical number of forces and casualties and solves the equations in order to find the value of A/B which indicates the quality difference of both sides .

So he finds a certain number claiming that germans for example had triple quality than Russians.

This number is taken by other authors and at some point comes to us.

So very often someone trying to "prove" that the equations are very good is doing the following.

He uses the "well known number" of superior quality of germans and doing the opposite procedure using Lanchester equations, produces accurate results about historical losses, so he thinks he proves the accuracy of the equations,which is of course not true.

regarding that the smaller side wins frequently,that is because they have acheived such a quality superiority that negates the numerical advantages.

It is not a coincidence that most of the famous generals were commanding inferior in numbers armies.

We do not see examples of small forces of less quality,winning often inspite the odds against superior in numbers enemy of superior or about the same level quality.

There is always a "quality factor" involved ,like training,morale, surprise and so on......

Farthermore, we can recall the fact that we accept that the attacker at the decisive point has to acheive numerical superiority against the defender who posses the quality of better protection ,better observation and so on , we do recall napoleon pointing that the big battallion always wins and we do recall that inferior forces in operational level win cause through superior maneuver acheive numerical superiority also or at least increase their concentration at the critical point.

What "theory"? As far as I can tell, Jason has merely been doing a comprehensive demolition job of both the simple-minded application of Lanchester's square law in places where it's wildly inappropriate, and guying the "romantic" view of combatants spending their entire time slaying each other in vast and Valhalloid numbers. In this, he is simply right.

First if all noone said that combatants spend their entire number slaying each other.

On the contrary, one of my arguments was that they spend a very small portion of the time slaying each other.

The issue for Lanchester and in our conversation in general is about "lethality" WHEN the combatants decide to try slay each other.

Extensions to the Lanchester equations to cover reinforcement have been around for a while.

Yep, as i said there are many modifications and very complicated forms.

I guess when we have such problems in examining the most simple form, there is no point to even attempt talking about others

So what's the expectation of the number of survivors in a square-law fight with an equal number of participants of equal effectiveness on each side?

Mutual destruction.Noone will survive.

Now ,this does not mean that we conclude that in real battlefield ,it is going to be this result.

We know that this almost never happens.

Does this mean that the equation is useless?

No.

The equation points as a principle that in such a situation , the expectation is to inflict and have about the same casualties as the enemy.

Better to say it somewhat differently.

In such a situation, both sides have equal chances to win or lose.

In real life two opposite tanks might have a fight to the end between with all equal.

During many different fights, one tank might win and score a single kill surviving itself and the other half of the battles having the opposite side scoring a kill and surviving itself.

In a way it is like having an equation pointing that if you throw two dices , you will get one even number and one odd.

In reality you can not expect to see this all the time.

In fact if you throw for example a dice 1000 times , it is more probable to see uneven number of even and odd results ,than seeing exactly 500 even and 500 odd numbers,

but the equation still holds value in pointing to certain conclusions.

So, if someone uses this equation to say that you have equal chances to get an odd or an even number when you through a dice, he is right.

Maybe in theory, but who cares? This never happens in real life. The number of tanks engaging other tanks seldom rises above being a succession of duels and truels, even on those rare occasions when tanks fight tanks. The SSKPs attainable by WW2 weapons are so low that over-hitting never becomes a concern.

When we want to examine something in theory, we very often "transform" reality to a theoritical scenario which is convenient for us to examine.

The fact that this theoritical scenario is not identical with what happens in reality does not minimize the importance of theory.

I can not build an identitical scenario but i might be able to build an equivalent one.

if i want to study airflow in aerodynamics, i am not oblidged to take in cosideration the movement of each atom but i can still produce a successful theory and build an aircraft for example.

Nope, I don't believe Fred ever said anything about volleys. His equations plainly indicate a continuous-time treatment.

It is expressed from the very beginning.

The two equations exactly because indicate continuous time treatment , they imply silmulateneous fire.

Both of them combined indicate that at WHATEVER time after the start of the battle,each side produce an atrition to the opposite.

How you produce attrition?

By firing in our case.

it is not only the continous time treatment,it is also a continuous attrittion treatment which means also a continuous fire treatment.

In other words ,no matter which time you choose to examine, both suffer and inflict attrition which means both fire.

If you claim that the fire is not treated as simultaneous,then you say that there is at least one value for the time ellapsed after the start of the battle ,when one equation gives a certain rate of attrition while the other does not.

That is obviously not possible since the "lethality" coeficient is fixed and not time dependent and the X or Y are different from zero .

[ August 19, 2005, 01:17 PM: Message edited by: pamak1970 ]

[dieseltaylor]

Naval history net says

October 1942

"In the build-up to the battle, Royal Navy submarines and RAF aircraft, especially those based in Malta, are sinking more than a third of Axis supplies setting out for North Africa. As the offensive gets underway, the Inshore Squadron continues to support and supply Eighth Army along its right, seaward flank."

This really does not seem to jibe with the huge graph posted above. I am bemused.

[pamak1970]

Since the previous posts became somewhat complicated

i will present a short and simple scenario to demonstrate that the theory that Jason uses to estimate lethality of a weapon is not logical without using any math

Now we all know from personal experience , that if you stand still in front of a soldier ready to shoot at you from a distance of 20 feet,the chances are that you are not going to survive after his first shot.

We understand that a soldier in such a situation is very lethal.

Let's pretend now that we are ignorant of this lethality and we are eager to estimate it.

So we examine a historical example of a firing squad executing a person from a distance of 20 feet and we see that a firing squad of ten soldiers kills a person .

For some people, this is enough to estimate lethality of a soldier shooting at a target from a 20 feet distance.

They take the number of kills in the above historical situation which is 1.

They then take the number of shooters in the above historical situation which is 10 soldiers.

They "spread" the kill among the shooters.

Therefore they divide 1 with 10 and finds as a result 0.1

So they conclude that

"on average each soldier "scores" one tenth of a kill against a target standing still at a distance of 20 feet.

Therefore , if a soldier fires at you from such a distance,he has very small chances to kill you, which is 10%".

or

If a firing squad fires at you from that distance, we will have on average one bullet hitting you and the rest failing to hit you.

And since the above "comclusion" counting on our current experience is obviously illogical,it begs the question of why such a theory which distorts facts about lethality so much in very simple situation like the above, can be used to estimate accurately lethality in much more complex situations?

[John D Salt]

Originally posted by pamak1970:

[snips]

All else in reality may never be equal but this does not make the theory invalid in pointing some principles.

I'm not saying that Lanchester's Theorem is invalid. I'm objecting to "the simple-minded application of Lanchester's square law in places where it's wildly inappropriate".

Originally posted by pamak1970:

[snips]

The other thing is that Lanchester does not talk about ratios only ,it talks about quality also and this is what is difficult to measure in reality.

Dupuy ,was one of the people trying to do it using historical data and here i want to point a common misconception.

Some times an author assuming that Lanchester law holds true,uses historical number of forces and casualties and solves the equations in order to find the value of A/B which indicates the quality difference of both sides .

I think you are misrepresenting Dupuy here. My recollection is that he refers to work by Fain and others pointing out the poor fit of Lanchester-like predictions with historical results. Where does Dupuy say he uses Lanchester's square law in his calculation of score effectiveness?

Originally posted by pamak1970:

[snips]

Farthermore, we can recall the fact that we accept that the attacker at the decisive point has to acheive numerical superiority against the defender who posses the quality of better protection

Jason has already made the point, which you have apparently ignored, that massing more troops on the breakthrouigh sector makes more targets for defensive fire, making the linear law more applicable than the square law. This matches the observation in the Kirke report that achieving tactical success in WW1 depended more on concentrating fire than concentrating troops. It also matches the current British Army principle of massing effects rather than massing troops.

Originally posted by pamak1970:

[snips]

What "theory"? As far as I can tell, Jason has merely been doing a comprehensive demolition job of both the simple-minded application of Lanchester's square law in places where it's wildly inappropriate, and guying the "romantic" view of combatants spending their entire time slaying each other in vast and Valhalloid numbers. In this, he is simply right.

First if all noone said that combatants spend their entire number slaying each other.

Great, So you agree with Jason. Now, I ask again, what supposed "theory" of Jason's is it you are saying is wrong?

Originally posted by pamak1970:

The issue for Lanchester and in our conversation in general is about "lethality" WHEN the combatants decide to try slay each other.

And we know that most casualties in modern high-intensity combat are produced by indirect fire. So why are you apprently obsessed with the direct fire battle? Not because it's more "romantic", perhaps?

Originally posted by pamak1970:

[snips]

I guess when we have such problems in examining the most simple form, there is no point to even attempt talking about others

The only person I can see here who is having the slightest difficulty in understanding Lanchester's equations is you.

Originally posted by pamak1970:

So what's the expectation of the number of survivors in a square-law fight with an equal number of participants of equal effectiveness on each side?

Mutual destruction.Noone will survive.

Right. And we know that zero is not the expected number of survivors in an evenly-matched fight. So, contrary to what you posted earlier, the Lanchester equations do not accurately predict the number of survivors of an engagement (and indeed are rather famous for not doing so).

Originally posted by pamak1970:

Now ,this does not mean that we conclude that in real battlefield ,it is going to be this result.

We know that this almost never happens.

Does this mean that the equation is useless?

No.

What it means is that if you want to predict numbers of survivors, you use a discrete rather than a continuous treatment and fling in some probability. Lanchester's basic equations can be thoght of as time-dependent flow rates. They are not probabilitic, and therefore do not have the concep[t of "expectation". If one replaces them with a model in which a number of discrete entities take shots at each other over time (Bernoulli trials), then you will be able to estimate the number of expected survivors.

In a very similar fashion, elementary modelling courses often show the comparison between predator-prey models written as continuous-flow models, and the same model written in stochastic discrete-event style.

Originally posted by pamak1970:

Better to say it somewhat differently.

In such a situation, both sides have equal chances to win or lose.

Yes, much better.

Originally posted by pamak1970:

When we want to examine something in theory, we very often "transform" reality to a theoritical scenario which is convenient for us to examine.

True; and when that transformation is done with no knowledge or regard of the main characteristics of the system under study that should be preserved in the transformation, the resulting theory is misleading tosh.

Isi Mitrani said that "to model a system is to replace it with something that is simpler, and the same in all important respects". Over-hitting simply is not an important effect in WW2-era terrestrial combat.

Originally posted by pamak1970:

The fact that this theoritical scenario is not identical with what happens in reality does not minimize the importance of theory.

And dignifying something with the gradiose title of a "theory" does not save it from being tosh. In fact we are not dealing with theories, which can be subject to proofs, but with models, which cannot.

Originally posted by pamak1970:

I can not build an identitical scenario but i might be able to build an equivalent one.

if i want to study airflow in aerodynamics, i am not oblidged to take in cosideration the movement of each atom but i can still produce a successful theory and build an aircraft for example.

I rather think that if you'd spent any time working in an organization that employs a large number of aerodynamicists, you would be less keen to tout aerodynamics as a good example of a successful theory.

Originally posted by pamak1970:

Nope, I don't believe Fred ever said anything about volleys. His equations plainly indicate a continuous-time treatment.

It is expressed from the very beginning.

The two equations exactly because indicate continuous time treatment , they imply silmulateneous fire.

[snips]

it is not only the continous time treatment,it is also a continuous attrittion treatment which means also a continuous fire treatment.

So there can be no concept of "volleys", can there?

All the best,

John.

[John D Salt]

Originally posted by pamak1970:

[snips]

i will present a short and simple scenario to demonstrate that the theory that Jason uses to estimate lethality of a weapon is not logical without using any math

I think you first need to say what "theory" of Jason's it is you think you're attacking.

Originally posted by pamak1970:

Now we all know from personal experience , that if you stand still in front of a soldier ready to shoot at you from a distance of 20 feet,the chances are that you are not going to survive after his first shot.

OK, hands up everyone who has "personal experience" of being shot at from a distance of 20 feet?

Originally posted by pamak1970:

We understand that a soldier in such a situation is very lethal.

We also, if we have not gone completely shouty crackers, understand how very far removed this romantic flight of fancy is from what happens on battlefields.

Originally posted by pamak1970:

Let's pretend now that we are ignorant of this lethality and we are eager to estimate it.

So we examine a historical example of a firing squad executing a person from a distance of 20 feet and we see that a firing squad of ten soldiers kills a person .

For some people, this is enough to estimate lethality of a soldier shooting at a target from a 20 feet distance.

Under exercise conditions, against a co-operative target, yes. I challenge you name any of these "some people" who would be so daft as to imagine for an instant that it bore the slightest resemblance to what heppens on battlefields.

Originally posted by pamak1970:

They take the number of kills in the above historical situation which is 1.

They then take the number of shooters in the above historical situation which is 10 soldiers.

They "spread" the kill among the shooters.

Therefore they divide 1 with 10 and finds as a result 0.1

So they conclude that

"on average each soldier "scores" one tenth of a kill against a target standing still at a distance of 20 feet.

Therefore , if a soldier fires at you from such a distance,he has very small chances to kill you, which is 10%".

Either you failed to understand my assertion that over-hitting is not a significant effect on the WW2 battlefield, or you believe that it is. So, let's have it; what reason do you have for believing that SSKPs of WW2 infantry weapons, on operations, were close to unity? I really would be fascinated to know.

Now it is possible for the kind of error you propose to be made. The historical example I'm aware of is the case of Israeli Gabriel missiles fired in the 1973 war. As is usual for missiles, the manufacturers claimed an SSKP in the region of 0.9. Looking at the record of Gabriel firings during the course of the war, the conclusion was reached, by dividing the number of Gabriels fired by the number of Arab missile-boats sunk, that the operational SSKP was about 44%. This didn't look too good until it was pointed out that the usual tactic was to fire them in salvos of two missiles, and often the first would clobber the target and the second be wasted.

However, I do not believe that SAA fired in WW2 had SSKPs even remotely comparable to radar-homing anti-shipping missiles in 1973. I'd be fascinated to know why you apparently do.

Originally posted by pamak1970:

And since the above "comclusion" counting on our current experience is obviously illogical,it begs the question of why such a theory which distorts facts about lethality so much in very simple situation like the above, can be used to estimate accurately lethality in much more complex situations?

No, that's not "begging the question".

A very rough rule of thumb that used to be used at Fort Halstead was that going from firing-range hit probabilities to exercise results reduced effectiveness by up to an order of magnitude, and from exercise to operations up to an order of magnitude again.

I think it was Phil Barker who said that "most engagements are fought just outtside the effective ranges of the weapons involved".

All the best,

John.

[flamingknives]

Now we all know from personal experience , that if you stand still in front of a soldier ready to shoot at you from a distance of 20 feet,the chances are that you are not going to survive after his first shot.

We understand that a soldier in such a situation is very lethal.

Actually, if one were to use S.L.A. Marshall's conclusions about WWII riflemen, you'd have less than a 20% chance of getting shot at, nevermind being hit.

I realise there are some flaws with Marshall's work, but don't know where to find an article explaining what they are.

If some kind soul could point me in the right direction...

[Michael Dorosh]

Didn't Robert Lawrence, MC, get shot at from about 20 feet on Tumbledown Mountain in 1982, take an FN round through the head that destroyed a goodly proportion of his brain - and live to tell the tale?

Ah....he said "chances are".

Back to lurking...

[John D Salt]

Originally posted by flamingknives:

[snips]

I realise there are some flaws with Marshall's work, but don't know where to find an article explaining what they are.

If some kind soul could point me in the right direction...

Pending the arrival of a kind soul, you might take a peek at John Whiteclay Chambers' screed from Parameters at http://carlisle-www.army.mil/usawc/Parameters/03autumn/chambers.pdf

What you really want is a copy of Roger Spiller's piece from the Winter 1988 issue of the RUSI Journal, but AFAICT that isn't available electronically, and the snippets of it that appear on the web involve some thoroughly dreadful selective quoting.

For criticism of SLAM's personal manner, rather than his research methods, there's a bit in David Hackworth's About Face which is entertainingly rude about him.

None of which alters the fact that, depite his self-regarding personality and his fabricated research data, SLAM still appears to have been pretty much right about participation rates, as similar observations were made independently by Wigram in Sicily.

All the best,

John.

[pamak1970]

I'm not saying that Lanchester's Theorem is invalid. I'm objecting to "the simple-minded application of Lanchester's square law in places where it's wildly inappropriate".

The way i see it, you do say that the Lanchester's theorem is invalid in real war cause the model is too simplistic.

Our difference is not that it has good applicability in predicting specific numbers in real situations.

Our difference is that I do accept that the PRINCIPLES described by Lanchester are valid in real war.

For me it is logical to beleive that if an attacker with a certain favorite ratio against the defender, manages to acheive victory after sufferring X casualties, then we can say that if the same attacker under the same conditions could repeat this attack with a more favorite ratio, he would experience most probably less total casualties,(according to Lanchester) .

Same if you play a CM scenario.

If you edit a scenario you win most of the times with an average of x casualties inflicting on you- and decide to add more forces to your side and play the same scenario again, then the more probable thing will be that you will win again the new scenario sufferring less casualties.

It is also more probable that you will win in a shorter time.

All the above hold true to me, inspite the fact that in real life we do not have "same conditions" but "similar" ones.

All this conversation we have is cause of my argument that the "average effectiveness" you use does not actually describe the lethality of a system itself since lethality is independent of numbers while your "average effectiveness" is not.

I think you are misrepresenting Dupuy here. My recollection is that he refers to work by Fain and others pointing out the poor fit of Lanchester-like predictions with historical results. Where does Dupuy say he uses Lanchester's square law in his calculation of score effectiveness?

Dupuy points poor results but this does not mean that he discards Lanchester equations.

He still accepts their validity but since "all else be equal" does not apply in real world,he develops a new system to quantify "this all else".

In general he tries to quantify the A and B factor of the original Lanchester equations which he accepts.

His new equations are similar to Lanchesters .

The difference is that the A and B factor are described through other parameters extracted from historical experience..

You are right that he does not use lanchester square law to calculate effectiveness.

He uses historical data to do so.

As a last point , I do not want to give the impression that my critisism was about Dupuy.

I just took the opportunity to comment about something i have seen occassionaly.

That is people other than Dupuy, trying to prove that his equations are accurate by doing basically the reverse procedure of what he did in order to produce them.

So when these people find results which match approximately historical data ,they think they proved that Dupuy is accurate,while in reality this was logical to happen since Dupuy used this data in the first place to produce its equation.

Jason has already made the point, which you have apparently ignored, that massing more troops on the breakthrouigh sector makes more targets for defensive fire, making the linear law more applicable than the square law. This matches the observation in the Kirke report that achieving tactical success in WW1 depended more on concentrating fire than concentrating troops. It also matches the current British Army principle of massing effects rather than massing troops.

son has already made the point, which you have apparently ignored, that massing more troops on the breakthrouigh sector makes more targets for defensive fire, making the linear law more applicable than the square law. This matches the observation in the Kirke report that achieving tactical success in WW1 depended more on concentrating fire than concentrating troops. It also matches the current British Army principle of massing effects rather than massing troops.

What you say does not make sense

Give a look to linear and square law and notice that concetration of fire is implied in square law nor in linear law.

This is expressed clearly in whatever sourse you want to see.

It is not "my interpretation" .

Consider also that the linear law points that in theory an attacker of a certain size A against a defender of size B will win and suffer x casualties,while in the same situation an attacker of triple size against a defender of the same size will again suffer the same casualties x.

I do not understand why you beleive that this law is more appropiate in describing battlefield effects.

In support of my argument that it is not just me that i beleive in the conclusions of square law,i point you Warden ,the person whose theory was the foundation for the air campaign in the first Gulf war.

From his book ""The air campaign- planning for combat"

"Loss Ratios a Function of Force Ratios

Another phenomenon is important for the air commander to understand: Loss rates vary disproportionately with the ratio of forces involved. Two forces equal in numbers (and reasonably close in equipment and flying capability) will tend to have equal losses when they meet. Keeping the same equipment and personnel, as the force ratios go against one side, that side will have greater loss rates than the changed ratio would suggest. Conversely, for the side for which the force ratios become more favorable, loss rates will fall more than the ratios would indicate. The change in loss rates, either positive or negative, is not linear; it is exponential. Furthermore, no point of diminishing returns for the larger force seems to exist. That is, the larger the force gets, the fewer losses it suffers, and the greater losses it imposes on its opponent.79"

He is clearly speaking about the consequencies of the square law.

This is somehting that it is well accepted inside the military community.

It seems that it is not the same in this forum.

In fact the same studies you use are against your argument.

They point that it is advantageous to mass EFFECTS,fires, shots.

They basically say that you SHOULD want to set a battle with characteristics of a "square law lanchesterian battle".

Massing troops in not sufficient by itself since it does not define the nature of the battle (aimed or unaimed)

Great, So you agree with Jason. Now, I ask again, what supposed "theory" of Jason's is it you are saying is wrong?

I posted many examples to point distortions of lethality based on average effectiveness.

that includes the last post you ignore.

You might have noticed that up until now i do not avoid countering some accusations against square law as they are expressed in your posts.

I beleive i have the right to expect the same from your side when i point specific logical falacies in the other theory.

Read my previous post carefully .

I say that the results of the other theory when they are applied in that simple scenario are illogical.

A simple presentation of what is wrong with the theory .

And we know that most casualties in modern high-intensity combat are produced by indirect fire. So why are you apprently obsessed with the direct fire battle? Not because it's more "romantic", perhaps?

The discussion started with the stuka against a tank effectiveness .

It was logical that it would be focused on "direct" fire .

The only person I can see here who is having the slightest difficulty in understanding Lanchester's equations is you

and Warden also according to his book.

Right. And we know that zero is not the expected number of survivors in an evenly-matched fight. So, contrary to what you posted earlier, the Lanchester equations do not accurately predict the number of survivors of an engagement (and indeed are rather famous for not doing so).

You still do not get it.

I do not talk about numbers ,i talk about principles that still hold true, when we see these numbers.

I do not understand why you insist on that.

If you are shot and receive a total of 9 bullets in 10 seconds, that means that you get 0.9 bullet every second.

Now we know that in reality this does not happen

Same with what we are talking here.

What it means is that if you want to predict numbers of survivors, you use a discrete rather than a continuous treatment and fling in some probability. Lanchester's basic equations can be thoght of as time-dependent flow rates. They are not probabilitic, and therefore do not have the concep[t of "expectation". If one replaces them with a model in which a number of discrete entities take shots at each other over time (Bernoulli trials), then you will be able to estimate the number of expected survivors.

I gave you the example of having the probability definition itself related with infinity which does not prohibit us to restrict it in definite number of trials.

In our case ,the proposition is the following.

If attrittion and fire is considered continous ,we say that it is equivelant with an infinite speed of rate of fire.

The claim is that if two forces have certain effects against each other when both posses a rate of fire of infinite speed, then we expect that the effects will be same when the two forces posses a definite speed of rate of fire.

Let me put it differently

If we were able to say that a certain force destroys half of the enemy before the end of the battle, at a common speed of rate of fire of 1 shell per second, then the same thing wiill happen if the common speed of rate of fire for the two opponents is 2 shells 'per second, 3 shells/per second and so on and so on up to infinity where the fire "becomes" continous .

True; and when that transformation is done with no knowledge or regard of the main characteristics of the system under study that should be preserved in the transformation, the resulting theory is misleading tosh.

Not always.

Up today ,we can not even understand the nature of many things we examine since ancient times.

Gravity is one simple example (do not confuse nature of force with results of this force)

It is true that there is a synergy of assmumsions you make and observations.You can not build a theory in the vacuum trying to describe real world events.

Over-hitting simply is not an important effect in WW2-era terrestrial combat

It is not the overhitting my friend.

It is the accumulation of probabilities the bigger numbers give you.

I give my best effort to give you examples to understand theoritical concepts.

I fail cause you still do not get it that the issue is the accumulation of probabilities.

I rather think that if you'd spent any time working in an organization that employs a large number of aerodynamicists, you would be less keen to tout aerodynamics as a good example of a successful theory.

In fact i was working for 10 years in civil aviation as a mechanical engineer.

I have some years under my back in studying fluid mechanics and aerodynamics in my university.

My example for using aerodynamics was not a coincidence

So there can be no concept of "volleys", can there?

of course they can.

See my comment above about infinite speed of rate of fire.

[ August 20, 2005, 01:13 PM: Message edited by: pamak1970 ]

[JPS]

Hmm, interesting thread. I decided to run CMBB experiment, July 43, midday, good weather, pretty open farmland. All troops regular. Large map (for attacking scenario, 3000 pts).

15 JU87G attacking 15 T-34M43 (also 10 BA-64B and some infantry, but those were never targeted by the planes). In first trial, defender had 5 25mm AA and 5 37mm AA (290 pts value). In second trial, only the 5 25mm AA.

Results:

Overall impression - the Stukas hit their targets often! However, ...

In first trial, 11 aircraft destroyed (2 for 25mm, 9 for 37mm), 1 T-35 abandoned, one immobile, 2 men lost.

In second trial, 1 aircraft destroyed, 1 T-35 abandoned, 2 immobile, no personel casualties.

Conclusion gamewise - Ju87G is waste of points if opponent is expected to have any T-35s.

However, what would be a historically reasonable level of AA guns? One per tank platoon? More? Less? How were these distributed/deployed, e.g., in the initial Soviet defensive stages of Kursk?

[pamak1970]

I think you first need to say what "theory" of Jason's it is you think you're attacking.

I said it many times.

That the average effectiveness he uses is not the same with leathality

OK, hands up everyone who has "personal experience" of being shot at from a distance of 20 feet?

Let me change it them to "personal knowledge" .

Unless you imply that you will raise you hand up when someone asks about personal experience in knocking out a T-34 with a sTuka.

We also, if we have not gone completely shouty crackers, understand how very far removed this romantic flight of fancy is from what happens on battlefields.

The difference of the nature of the battlefiled and a firing squad is totally irrelevant to your theory cause it does not examine this nature at all,meaning the mechanism through casualties are inflicted.

It sees only "how many",not " how and under what conditions it happened"

Your theory uses only the number of shooters and casualties which are present in any case to define a term like lethality which has a standard meaning , definition..

The definitions of lethality, shooter,kills or "average kills per shooter'remain the same.

The math that estimates "average" is the same.

This type of math is not related to conditions,it is related to the same definition of "average".

Therefore always you divide kills to shooters under any conditions mo matter if it is during war or hunting..

Now if the definitions and mathematical logic that your theory is based on , is not affected at all by the different nature of a battlefield and a fighting squadron ,then if i see illogical results in my conditions, i claim the problem is with the foundation of your theory itself.

It is not a bi product of "different conditions".

I think "my data" were reasonable.

If theory does not calculate reasonable values for a standard definition of "lethality" in my case then there is a logic gapso the theory will fail in your scenario too

An additional question i would like to ask.

Why it predicts lethality in wwii?

Can it do the same for the gulf war?

I guess your answer is yes

Still the two situations are different ,right?

Why you assume here that the procedures of applying lethality are similar between two totally different wars but when someone points you the problems with the firing squadron example, you point that a war is different from the scenario of a firing squad?

Prove and explain under what conditions your theory applies .

Lanchester and his differential equations did not convince you.

Now it is your turn to try to prove that your theory estimates lethality during wwii.

If you can not do it, then why i should accept your claim?

I can easily say without a need to prove that your theory does not predict lethality in wwii.

or that it predicts lethality during napoleonic wars only.

Under exercise conditions, against a co-operative target, yes. I challenge you name any of these "some people" who would be so daft as to imagine for an instant that it bore the slightest resemblance to what heppens on battlefields.

I am not sure what you point.

if you beleive that your theory gives illogical results in my scenario but argue that " war is different",then you do not point anything worth for discussion.

War might be different and your theory may apply or war might be different and your theory may not apply either.

WE have to see your reasoning and it is always your resposibility to provide proof to support your claim.

Now if you do beleive that your theory does apply in my example and estimate that a soldier of a firing squad has an average lethality of 10% under exersize conditions against a cooperative target at 20 feet , i have no farther comments.

I think it is much more beleivable to say that a sTuka can knock out tanks very easily

[Bigduke6]

Kursk is a little of a special case, as the Soviets were in a prepared defense, in depth, and Red Army proportion-wise had concentrated in the sector even more equipment like tanks and AAA than men.

A standard infantry division as I remember had .50 cals to keep the airplanes away. AA units using "tactical-level" automatic cannon were in separate battalions usually controlled by Corps and Army. I don't recall what the standard distributions were, but that's easy enough to look up.

If you are talking the standard Soviet armor manuever unit - a brigade - then I would expect it to have about a platoon to a company of some kind of towed AAA under its direct control, and maybe more if it was in an assembly area, crossing a bridge, or otherwise particularly vulnerable to air atttack.

So I would say the AAA you used in your first test scenario was a good deal more concentrated than the average case in RL, and in your second scenario a bit more concentrated than the average case in RL, but certainly within the realm of possibility.

Originally posted by JPS:

Hmm, interesting thread. I decided to run CMBB experiment, July 43, midday, good weather, pretty open farmland. All troops regular. Large map (for attacking scenario, 3000 pts).

15 JU87G attacking 15 T-34M43 (also 10 BA-64B and some infantry, but those were never targeted by the planes). In first trial, defender had 5 25mm AA and 5 37mm AA (290 pts value). In second trial, only the 5 25mm AA.

Results:

Overall impression - the Stukas hit their targets often! However, ...

In first trial, 11 aircraft destroyed (2 for 25mm, 9 for 37mm), 1 T-35 abandoned, one immobile, 2 men lost.

In second trial, 1 aircraft destroyed, 1 T-35 abandoned, 2 immobile, no personel casualties.

Conclusion gamewise - Ju87G is waste of points if opponent is expected to have any T-35s.

However, what would be a historically reasonable level of AA guns? One per tank platoon? More? Less? How were these distributed/deployed, e.g., in the initial Soviet defensive stages of Kursk?

[JPS]

Thanks BigDuke. I informed myself a bit more based on Red Army Handbook 1939-1945 (Zaloga and Ness, 1998). Below my summary for those interested:

For integrated Soviet (light) AA assets, in early war one would see quad-maxims (in trucks or fixed positions) and 37mm. Later on, quad-maxims are replaced by 12.7mm heavy MMGs. The 25mm AA is not typically used by front line troops, but rather by PVO air defence force regiments. In CM setting its rarity should be high. This is unlike the German equivalent.

Some production numbers:

25mm AA 2k from 1941 to 1943

37mm AA 12.1k from 1941 to 1943

12.7mm heavy MG 23.2k from 1941 to 1943

Organizational deployment:

- Motorized rifle brigade (from April 1942) would have 12 37mm AA and 3 12.7mm HMG

- Tank corps (end of 1942) would have only the above AA assets for the whole corps (63 light tanks, 99 medium tanks)

- Tank brigade (December 1941) would have 4 37mm AA and 3 HMG (for 16 light, 20 medium, and 10 heavy tanks)

- Tank brigade (August 1941) would have 8 37mm AA and 6 12.7mm HMG (for 22 medium tanks and 32 light tanks)

- Tank brigade (from November 1943) would have 9 12.7mm HMG (for 65 medium tanks)

- Mechanized brigade (from September 1942) would have 8 37mm AA and 12 12.7mm HMG (for 23 medium and 16 light tanks)

- Mechanized brigade (from Feb/Sep 1943) would have 9 12.7mm HMG (for 32 medium and 7 light tanks)

- At mechanized corps level, the 37mm AA varied from 40 (Sep 42) to 26 (Jan 43) to 18 (Jan 44) to 16 (May 45); these are for tank strenghts varying from 175 (Sep 42) to 246 (May 45). Heavy AA MMGs were not reported.

- A dedicated AA division would have equipment as follows

1942: 48 quad-maxim, 32 12.7mm HMG, 48 37mm AA

1943: 52 12.7mm HMG, 48 37mm AA, 16 85mm AA

1944: 52 12.7mm HMG, 72 37mm AA, 16 85mm AA

As a rough baseline rule-of-thumb, I'd expect to see around one 37mm AA per 8 tanks, and somewhat more 12.7mm HMGs. So 2 37mm AA and 3-4 12.7mm HMGs could be seen "typical" for my test scenario (althought it was underrepresenting infantry).

I do not have sources that would be useful in describing how the integrated light AA assets were deployed tactically (i.e. at level CM is addressing, both in defence and in attack). If someone could provide insights on this at least I would be interested.

[Andreas]

Originally posted by Steiner14:

</font><blockquote>quote:</font><hr />Originally posted by Andreas:

</font><blockquote>quote:</font><hr />Originally posted by Steiner14:

...propaganda hero...rudel's war diary a propaganda book...blahblahblah.

Rudel was that ineffective and such a braggart, that he was forbidden to fly. Hm, strange.

His kills are all confirmed? All wrong! Propaganda numbers.

Rudel was involved in the development of the 87G: bah, this plane doesn't hit tanks and therefore it was judged as highly effective just for - guess right - propaganda issues.

Why don't you just answer the question? </font>

[JasonC]

Gee, does every person shot have 11 holes in them? Do 3/4 of them manage to survive their wounds anyway? Does ammo consumed exactly track average unit sizes engaged tactically at a single time?

Average effectiveness is an integral. The item being integrated is tactical effectiveness. The limits of integration track ammo use, and vastly exceed unity. There is no way to combine a hypothesis of high per engagement effectiveness with low average effectiveness, without contradicting known facts at dozens of points.

That is how cross checking works, why accounting works, etc. Errors falsify not just one measure, but scores of implications. The average lethality of most weapons is dramatically lower than advertized, per tactical occasion even more than overall.

You can't kill 50% of the time 500 times and average only half a kill.

Since rationality left three exits back, I see no point in continuing this conversation.

[Andreas]

Coming back to Rudel - from the TDI thread it appears that he claimed 12 Soviet tanks on his first outing with the 87G somewhere around July 7th 1943. Are these part of his 500 claims?

Jason - rationality left a long time ago. Unfortunately however, Rudel was a neo-nazi icon (highest decorated soldier, vanquisher of 500+ tanks and a battleship, loyalty to his Fuehrer to the death, yada yada), and as such questioning his kill claims is about as heretical to Nazi nutcases as walking up to the head of the Spanish inquisition somewhat around 1550 and suggesting that all this God-stuff is really a load of nonsense, and asking him what he thought of that nice Mr. Luther in Germany with all his interesting and reasonable ideas would be, and if done in the 'right' place, the response would be similar.

[pamak1970]

You can't kill 50% of the time 500 times and average only half a kill.

Your numbers have nothing to do with reality.

Noone claims that the kill probability was 50% and noone claims that each tank engaged enemy tanks during 500 battles on average.

So i guess you are talking theoritically,in which case it is possible to have the above result Since the average kill as you define it is always related only to number of shooters and number of destroyed targets,it is possible to have all the above numbers you gave true

For example if 1000 tanks engaged in 500 hundred different battles and during each one of the battles they were engaging a single tank with a 50% kill probability ,then at the end of these battles the average kill per tank would be half a kill.

Same if 1000 tanks with a 50% kill probability destroyed 500 tanks in a single battle.

If you want to talk about reality,then you have to examine real data from real battles and get rid off the average .

Coalition forces during the first gulf war had according to your average 1 kill per tank after 100 hours of operations.

If you think that during a tank battle between M1 Abraams and T-72 of about one hour duration, you are going to see M1 tanks have an "average" kill of 1% , you are dead wrong.

In a similar way you are wrong if you "conclude" that a force of 100 M1 tanks will kill about 10 T-72 during a tank battle of one hour duration.

Battles and the opportunity to shoot against an enemy target are very rare but when they do happen attrition is much more than the one you estimate through "average kills".

Both sides will continue fight and inflict or suffer attrition until someone reaches the "break point" and either the attacker will abandon the attack or the defender will retreat.

[pamak1970]

There is no way to combine a hypothesis of high per engagement effectiveness with low average effectiveness, without contradicting known facts at dozens of points.

From what i see you refuse to read what i am saying cause i gave many clues why you can not match average effectiveness and battlefield effects,like rare opportunity to shoot against an enemy tank, multiple actual kills in tactical terms -much more than the actual number you use -since tanks could be repaired and return again on the battlefield-favorite ratios and so on and so on,

The only known facts are tank losses during intense fights.

You contradict data when we have a lot of tank losses in a matter of days while average perfomance was half a kill for a tank for the duration of its life which according to you was one year and if your argument is that these losses were cause of other weapons also like AT tanks, all you have to do is calculate the average perfomance of an AT tank during the duration of its life ,which will be extremely low also.

I can give you the problem of assumming that all aircraft ,all artillery , all tanks and all antitank mines were focusing on killing enemy tanks.

Calculate the "average kills" for each weapon system and try to add their average kills per day.

When you will see that your results do not match historical data during heavy battles,you might think again about your method of calculating battlefield effects.

[JasonC]

500 is a realistic figure for AP ammo supplied per tank in WW II.

They didn't have laser range finders and on board ballistic computers.

For that matter, what is the average kill chance per engagement of a T-72 fighting M-1s? For the T-72s that is. You want to look only at the above average cases, but that is exactly what averages will not let you do.

Lopsided battles won't tell you anything about WW II because WW II was not lopsided. Losses were very high on the winning side. Loss rates of tanks deployed were 100% on the losing side of course, but they were around 75% on the winning side as well. The war took years, not hours. Most tanks were lost, and most tanks did not take out one enemy tank. It is not like they wiped out the enemy in a few hours without getting their hair mussed.

The per round lethality of tanks against tanks in WW II was very low. That *all* weapon systems fail to take out their own value on average is a mathematical law. That tanks were no exception in WW II is just an irrefutable, established, empirical fact. You don't have to like it for it to be true.

And I am done offering the truth to you. The horse has been lead to the water. If you want to believe nonsense and lies, it is your funeral.

[pamak1970]

Originally posted by JasonC:

500 is a realistic figure for AP ammo supplied per tank in WW II.

They didn't have laser range finders and on board ballistic computers.

For that matter, what is the average kill chance per engagement of a T-72 fighting M-1s? For the T-72s that is. You want to look only at the above average cases, but that is exactly what averages will not let you do.

Lopsided battles won't tell you anything about WW II because WW II was not lopsided. Losses were very high on the winning side. Loss rates of tanks deployed were 100% on the losing side of course, but they were around 75% on the winning side as well. The war took years, not hours. Most tanks were lost, and most tanks did not take out one enemy tank. It is not like they wiped out the enemy in a few hours without getting their hair mussed.

The per round lethality of tanks against tanks in WW II was very low. That *all* weapon systems fail to take out their own value on average is a mathematical law. That tanks were no exception in WW II is just an irrefutable, established, empirical fact. You don't have to like it for it to be true.

And I am done offering the truth to you. The horse has been lead to the water. If you want to believe nonsense and lies, it is your funeral.

I told you before why number of supplies per tank does not show number of ammunition actually fired per tank,plus i pointed that during an engagement each tank has a load of dozens of AP round to use trying to kill enemy tanks,it does not use one Ap round per battle

Regarding the Gulf war,

Your theory was used in exactly the same way you use it to find the average kills per American tank.

Your theory uses number of shooters and number of kills,spreading kills among shooters.

If you say that the equations to estimate M1 effectivenss are not right, give the proper ones.

If your theory has any logic, then it should apply in all cases, regardless if the battle is lopsided or not.

Your theory does not give any "conditions" regarding when you could apply it and when not.

That is something that you say in the same way i can say that your theory can not be used to calculate effectivenss in wwii and that it should apply only in Napoleonic wars for example.....

Last but not least,even during wwii large tank battles that produced a lot of tank losses on both sides lasted for a few days, not months or years.

The other misunderstanding is that there is no dispute on the fact that on average we have half or one third of kills per Soviet tank for example.

The calculation is right.

The dispute is the interpretation of what this really means.

It is similar to the example i gave in another post when someone sees a soldier firing a shot against an easy target from a close distance and "expects" to find one third of a bullet inside the dead body because he misinterpretends "average"

numbers.

Regarding of your idea that you offer truth, my opinion is that you offer gross misinterpretations.

Anyway, i think the best advice i can give you is trying to use your theory to estimate average enemy AFV kills per tank/artillery pieces and planes.

When in doubt what numbers to use, select the one that favors attrition and kills per asset

Then try to combine these results to see if expected enemy tank losses are realistic and fit with the historical data you have during big tank battles when both sides used thousands of the above assets.

Your theoritical number will be much lower than the real ones.

I know that in this forum right now you are not going to accept anything i say.

When you are alone trying to test your hypothesis the way i propose, things will be different

[ August 24, 2005, 08:48 AM: Message edited by: pamak1970 ]

[JasonC]

You haven't the slighest idea what "my theory" is, you are arguing against a straw man between your own ears. Nuance is wasted on you because you pay no attention to the exact logical meaning of careful statements, you just jump to imaginary generalizations you can quibble with.

Of course I have partitioned all losses among all causes of loss, necessarily crudely given the uncertainties, but using estimates and rankings of relative effectiveness. That is where statements like the average German vanilla AFV got no more than 1 kill come from, with figures no higher than 3 for Panthers and 5 for Tigers overwhelmingly likely. Everything from fausts to 88s is included in those estimates.

Enjoy your last words. You've never taught me a thing or mentioned a single point I hadn't long since considered, and won't take instruction. So you are useless to me.