Tank gun test, extreme ranges. (Warning: large pics inside)

[JasonC]

M Hofbauer noted that most of the German AT weapons at Kursk were 75mm, so it is not surprising that they got most of the kills. Quite, and my point rather exactly. He just didn't extrapolate the uber-Tiger tale to see it is incompatible with this, if we are meant to draw any overall lesson from it. If we are not, if it is an anecdote and known to be exceptional (at the high end), then its is fine. It is perfectly believable that 2 Tigers KOed 16 tanks on one occasion, of course. (But actually they claim 34, which is distinctly less believable). What makes the use of it an uber-Tiger tale is the insinuation that this is perfectly typical, everyday performance - 17 kills per vehicle per outing, without loss (curiously, in -both- examples - 1 vs. 17, and 2 vs. 16+18).

What the Kursk figures do is falsify that general picture, if your own native common sense can't do it for you. If every 88 AFV at Kursk (there were ~300) killed 17 enemy AFV per outing without loss, all Russian tank losses at Kursk would be accounted for in one outing per 88 vehicle, with some to spare, and with none left to other arms present. The uber-Tiger picture requires enourmous numbers of dead AFVs specifically killed by 88 vehicles, and there aren't enough dead AFVs to go around to fit that picture literally, as a supposedly average result.

If we take the Russian loss portion of 1/5, or make some allowance for error and up it to 1/4 of all kills, and if we ignore the contribution of towed Flak, then a average German 88 vehicle present KOed 3-4 Russian tanks in the entire Kursk campaign, over two months of fighting both offensive and defensive (they lost up to 5000 AFV). With towed 88s included, 2-3 might be a more reasonable figure. This may well be consistent with 1 Tiger KOing 17 enemy tanks in one fight, without loss, on one particular occasion, especially if the range of that fight were below average, the targets particularly exposed, the crew particularly crack, etc.

As the best outlier out of a population it is a believable "run". In fact, the top 1/300th of a negative binomial distribution with 2/3rds success at each trial is 1, so you'd expect a 1 random run 17 long out of 300 attempters, if the chance at each trial were only 2:1 in favor. If the odds were 4:1, you'd expect 20 out of 300 vehicles to see runs that long. But it is not consistent with that being an average occurance. Obviously, the mean result of 2:1 odds is 2 kills per loss, and of 4:1 odds is 4 kills per loss.

How is this related to the more general subject of the thread? The initial poster, calling for increased accuracy of ranged fire, does so because he expects his German ubertanks to reliably kill again and again at the longest ranges, without expending much ammo to do it. An average, not an outlier. And for a few tanks to therefore be able, over one ammo load, to KO many times their own number of enemy AFVs, even at long range. Which, except for the range, is what the uber-Tiger anecdote is meant to suggest was normal or average, rather than exceptional.

The expectation the initial poster had that led to his call for vastly increased long range accuracy, is precisely that such occasions should be typical and not exceptional, even if the range is extreme. But if what he wanted were the reality, long ranged fire by 88 vehicles alone would have "swept the board" at Kursk. Which they didn't. c3k admits as much when he says "this was an exceptional attack on an exceptional defense". But exceptions do not tell us -average- figures, by the definition of average.

On c3k's other questions, I gave the AP ammo figure for 88 longs precisely to avoid any potential confusion with 88 Flak. I also excluded the HE ammo. 88 longs - only 5000 weapons, Nashorn, Elephant, Jagdpanther, King Tiger, and mostly towed PAK - had 2 million AP rounds made for them alone. That's a lot of seriously dead bushes. AP, not HE, per fielded weapon, consistently has 3 digits. I don't think 90% of it was sitting in warehouses at the end of the war.

Numerical analysis can indeed be a useful tool, but to use it properly you have to put in numbers or ranges for unknowns, that you actually believe possible, which is just what tss did not do. He himself says he thinks the shots per hit might be in the 6-12 range or might well be in the 10-20 range, particularly for ranged fire. He thus admits he knows his other lines are off by a factor of 2 to 8 at least, in cumulative effect, but he does not bother his head about that. Because he is not honestly applying the method, he is just stuffing in things he doesn't believe and noting he doesn't believe what pops out the other end. Duh. The outcomes are only as good as the estimates of intermediate stages, and when the outcomes are obviously wrong - as tss himself insisted - that reflects not on the general method, but only the purposefully unrealistic assumptions about 95% targeting of bushes and only 10% deliveries to the front.

As Wreck noted, the people proposing 95% targeting of bushes do not believe it sufficiently themselves to actually want to see it modeled. This shows they too are arguing for an ulterior purpose, or rhetorically, instead of saying what they actually think. Recon by fire with MGs, especially the high ammo coaxial, is not only believable but documented, and I can even see recon by fire with HE, especially when supporting an infantry assault, much like artillery barrage fire at suspected map coordinates. But I have never heard of recon by AP. The reason is obvious - you have to hit a point target to do anything with AP.

More generally, the whole point about different facets checking each other requires that one stick to estimates one actually believes, and is willing to see modeled any significant effect one regards as real. Peddling out recon by AP to suit an accuracy boost, then refusing the resulting phantom targets, is transparently a desire to have more fire effect per round expended than one believes the actual participants had. Which again is simply not honest.

In the same category are silly claims about enourmous supply chain losses to boost average accuracy figures, combined with no willingness to see infantry without loads of automatic weapons, or availability of one's favorite rare tank or gun, where the supply chain is supposed to work perfectly to get everything into the hands of the ubertroops. What all such rhetorical manipulation of assumptions betray is a desire to increase the game effectiveness of one's favorite forces, instead of modeling them accurately. And a lack of the honesty about one's own estimates needed to engage in any sort of modeling of real historical processes.

As for my main question, why not fire at range, tss gives two answers. Exposing oneself and thus drawing artillery fire, and maybe you can get better. He admits the former is only a real risk to towed guns, which makes it somewhat beside the point. A tank can fire off all the ammo it is likely to want to expend at long range within about 5 minutes. Certainly there is precious little risk of drawing down artillery fire that close upon opening up, (especially vs. Russians), and the tank can move at any time, anyway, and the imaginary immediate response artillery can't hurt it very much to begin with. Moreover it is not obvious that the shooter will be spotted, when the range is so long. Instances of long ranged fire from hidden PAK, never spotted despite minutes of fire, are quite common in the histories.

The other answer is in the category of "maybe". Maybe you can do better when they get close. Better than what? You can't knock out all the enemy tanks 4 times. If you have superior odds, you are likely to KO them all even with a 5% hit probability per round, even in just a couple of minutes of firing. Say a company fires at a platoon, and say because of range the ROF is only 5 rounds per tank per minute. The platoon will be dead within 90 seconds. Against even numbers a 5% hit probability could KO the enemy force within 4 minutes, expending only about half the typical on-board AP loads to do it. I can see holding fire until the hit probability rises to around 10% if one is outnumbered, sure.

But what happens when you wait? The target opportunity can pass. The enemy may disappear in a fold of ground. The enemy might get close enough that his replies become dangerous to you, as well. As the range drops, enemies within range achieve wider and wider side angles, limiting the usefulness of superior front armor. Did tankers still hold fire in ambushes and such? Certainly, but the most common reason is waiting for the range to be short enough for a round to actually penetrate - e.g. with 75 long vs. T-34s or something. It can also make sense to hold fire in an ambush until the kill chance with each round is high, to avoid more of the enemy reply fire, if that is seriously dangerous, and if enemy detection of the ambushers is a very unlikely.

But if the gun can kill the target, and the target has presented itself, and ranged fire can be delivered with perfect safety against the replies - and the accuracy has risen even to a few percent, then opening up makes perfect sense. As is typically the case with ranged fire by Tigers (and Panthers), which is the subject that sparked the whole thread. Because the replies aren't going to hurt you, and you aren't going to run out of AP before the enemy runs out of luck, and you are going to get hits if you just keep firing with all guns, for minutes rather than seconds. We know they did deliver such ranged fire. The point is, doing so would be an effective tactic even with modest hit percentages per round. The ammo was there, it makes even low probability per round fire tactically effective, and there is precious little reason to believe it wasn't there for just such a reason.

[Andreas]

<BLOCKQUOTE>quote:</font><HR>Originally posted by Vanir Ausf B:

IIRC it was 11% of the pilots got 90% of the kills.

<HR></BLOCKQUOTE>

I see you have a new sig

[M Hofbauer]

<BLOCKQUOTE>quote:</font><HR> JasonC: M Hofbauer noted that most of the German AT weapons at Kursk were 75mm, so it is not surprising that they got most of the kills. Quite, and my point rather exactly. He just didn't extrapolate the uber-Tiger tale to see it is incompatible with this, if we are meant to draw any overall lesson from it. If we are not, if it is an anecdote and known to be exceptional (at the high end), then its is fine. It is perfectly believable that 2 Tigers KOed 16 tanks on one occasion, of course. (But actually they claim 34, which is distinctly less believable). What makes the use of it an uber-Tiger tale is the insinuation that this is perfectly typical, everyday performance - 17 kills per vehicle per outing, without loss (curiously, in -both- examples - 1 vs. 17, and 2 vs. 16+18).

<HR></BLOCKQUOTE>

sorry but I think you are still wrong, even though you try to throw smoke and shift your position.

back to your original previous statment which still lacks logic: if there were hundred Pz IV killing 80 T-34, and one Tiger killing 20, then the Tiger still would validate the uber-theory, even though the example perfectly fits your statement.

again, your above new statement is *also* wrong, namely because such feats (single Tigers killing over a dozen T-34s on some occasions) are hardly ever heard of from Pz. IVs, or from Shermans (Korea), or from Shermans vs. Pz IV, or from T-34 vs. Pz. IV et cetera perge perge. Tigers did it not do that regularly but on a frequency so as not to call it absolutely single incidents, and with a frequency that would create the uber-status which you are disputing here: there's a reason why enemy tank crews had specific fear just for the Tiger. They (allied tank crews) were neither stupid nor cowardly, and I hope not that you are trying to make them out to be for being Tiger-phobic.

in other words, tio use above example again, if there were hundred Pz IV killing 80 T-34, averaging between 1 and 5 kills per Pz IV, and there was also one Tiger killing 20, and five braindead deafblind Tiger crews which didn't hit anything at all, then this example would still validate the Tiger-88-uber-theory, even though the example perfectly fits your new enhanced statement.

Or perhaps you can relate a similar vehicle which created a similar amount of incidents with such kill ratios from the same time period such as the Tiger I to support your argument above (but please not other german cats as they are exactly at issue here)?

But I guess you'll probably just throw another Nebelkerze and withdraw to appear again from another location with another new statement.

btw sorry if I sound harsh, it was a long day....

[tss]

*Thump* *Thump* *Thump* That damned horse doesn't seem to want to move. Maybe if I give it one more gently prod. *Sput* No effect. Maybe I have to admit that it's dead and it is time to bury it and let it rest in peace.

This will be my last post on this thread. I don't have really anything new to say and just add this response to point out few interesting features of Jason's answer to my post. There's nothing left here but a flamewar and I'm not interested in participating in those.

Jason, you don't have to bother to reply to this. I won't be even reading it. And after I have time to migrate my UBB-killfile code to Mozilla, I won't be reading any of your future posts, which is in a way a pity since you do often post highly informative posts. However, it is better for my blood pressure to stop reading them altogether.

JasonC wrote:

Numerical analysis can indeed be a useful tool, but to use it properly you have to put in numbers or ranges for unknowns, that you actually believe possible, which is just what tss did not do.

Translation: "I can take numbers out of thin air so that they will to get conclusions I like. If you do that, you are a liar".

At least that is the message I get here.

How do you know that your reasonable sounding assumptions lead to a reasonable conclusion? That first one of my datasets led to an unreasonable conclusion even though the individual variables were, at least to me, quite reasonable.

He thus admits he knows his other lines are off by a factor of 2 to 8 at least, in cumulative effect, but he does not bother his head about that.

You see, I'm not a psychic. If I don't have enough data to get a reliable average for the ammo ratio, I don't pretend I can do it.

No, once more, for the final time: the whole point of the formula and two sets of values was to show that without knowing the real values of parameters the result is meaningless.

Because he is not honestly applying the method, he is just stuffing in things he doesn't believe and noting he doesn't believe what pops out the other end.

So, apparently, if I honestly believe that my estimates are correct, they are correct even if I have no way of knowing the actual values.

that reflects not on the general method, but only the purposefully unrealistic assumptions about 95% targeting of bushes and only 10% deliveries to the front.

And I assume you have some sort of magical oracle that tells to you that your assumptions are realistic.

This shows they too are arguing for an ulterior purpose, or rhetorically, instead of saying what they actually think.

By my count this is the third time you accuse me of being a liar. But then again, how could I count accurately when I'm just a stupid engineer who couldn't even win a debate against a piece of a railway track.

Peddling out recon by AP to suit an accuracy boost, then refusing the resulting phantom targets, is transparently a desire to have more fire effect per round expended than one believes the actual participants had.

Thank you for telling me that. It is always enlightening to find out what I really think.

Which again is simply not honest.

Third time in the same post.

... combined with no willingness to see infantry without loads of automatic weapons, or availability of one's favorite rare tank or gun, where the supply chain is supposed to work perfectly to get everything into the hands of the ubertroops.

Well, after that paragraph I'm pretty certain there's no danger of any birds invading this thread; that number of strawmen could frighten away a flock of angry ostriches.

I think your argument would be a little more convincing if you would rebut what I wrote and not what you think I wrote.

Hint: I haven't called for preferential treatment for "übertroops" during this argument. In fact, the only thing that could be interpreted that way (and then only if one really wanted to do so) is that I've mentioned that Finnish artillery doctrine called for batallion-strength one-minute long strikes fired with intesive fire (120 rounds in a minute for 75 or 76 mm guns). And I hadn't mentioned that in this thread before this.

And a lack of the honesty about one's own estimates needed to engage in any sort of modeling of real historical processes.

So, it is not honest to admit that one doesn't know real figures?

... and the imaginary immediate response artillery...

followed by:

Instances of long ranged fire from hidden PAK, never spotted despite minutes of fire, are quite common in the histories.

And instances of artillery fire called on noticed AT assets are not?

By the way, I noticed that you didn't answer that one direct question I asked from you.

OK, that's enough. I'll go have some sleep.

- Tommi

[c3k]

JasonC,

As always, an interesting exchange.

Kursk is an anomaly if you're basing any kind of argument of Tiger lack-of-accuracy on kill ratios on that battle. A well prepared (in fact, text-book example of prepared defense) Soviet force with concealed, tightly controlled anti-tank defenses, including tanks. The Germans moved forward into what was essentially a 4 day continuous ambush. And still they advanced. Pretty sure a Sherman couldn't have done that.

So, let's look at some of the anecdotal evidence of 8.8cm long range accuracy. (By the way, I give a lot of credence to these reports. You seem to dismiss them as tales spun near the camp-fire. Unit awards, valorous citations, etc., are based on these. They are backed up by post battle analysis. How did all those Soviet tanks get destroyed?) Anyway, to my point: most of these encounters (the ones highlighting the accuracy of these weapons) were with long fields of vision, Soviet advance or stationary and unaware, Tiger or other "long-gun" on a hill, well camouflaged, tactically on the defensive (important point). Time and again they destroy their targets. Anecdotally. Since, if they didn't tell anyone about it, we'd never know about it.

Let's use your 3-5% accuracy at 2000 meter figure. My references are in another room and I'm on a typing roll, so let's assume a Tiger (or any long 75 or 88) has an ammo load of 40 rounds. Using 5%, that means one hit per 20 rounds. So, my total load out, leaving nothing, yields 2 kills (assume a hit is a kill). Tigers often operated in pairs (a zug). 2 Tigers on a hill taking out just 4 tanks at 2000 meters??? C'mon. If that happened the crews would be reassigned as infantry.

Now, how often were the above conditions met? In the steppes, often enough to establish a reputation. In North Africa, the same.

What does this have to do with CMBO? Easy. My 8.8cm flak missed a stationary target at 982 meters 4 times in a row. The other one missed an oncoming tank, ambush, at 400 meters 3 times.

[Jeff Duquette]

Tss

Great posts as usual. I especially enjoyed the function you came up with…although I would have added a few more variables to it. One equation....fourteen unknowns...now solve for x,y,z,a,b,c,d,e,f,g,h,i,j,and K.

<BLOCKQUOTE>quote:</font><HR>Marcus Said: and there was also one Tiger killing 20, and five braindead deafblind Tiger crews which didn't hit anything at all<HR></BLOCKQUOTE>

This is the crux of what I have been talking about, and the biggest reason why I haven't bothered responding to Jason's - albeit interesting - speculation that accuracy should be established as a function of manufacturing output. The original concept was that a few do most of the killing. So if the "many" are firing 100 or even 1000 rounds before they register a kill it has no bearing on the "few" that understand their weapon systems and remain reasonably calm in combat and kill "many" with only a few rounds. Obviously everyone isn't Radley Walters, Lafayette Pool or Kurt Kniepsel (Someone insert a British, French, Finish, Soviet, Australian, South African, New Zealander, Indian, Italian, Rumanian, Hungarian, Bulgarian, and Polish Tank ace here please…or any other nationality I accidentally left out).

Personally I think an "elite" or even "crack" rating in CM for a tank crew should push these crews into the realm of "uber mench". Whether these crews are German, Canadian, British or Martians. IMHO Their ability to hit targets at ranges indicated in some of the AAR's already posted on this thread should correspond with ammunition expenditures also indicated in these same AAR's. IMHO point value as well as rarity should also be tweeked accordingly. Am I going to stop playing CM if this don't happen…no. Am I going to continue to bitch about this on occasion…you bet.

[ 10-04-2001: Message edited by: Jeff Duquette ]

[c3k]

JasonC,

As always, an interesting exchange.

Kursk is an anomaly if you're basing any kind of argument of Tiger lack-of-accuracy on kill ratios on that battle. A well prepared (in fact, text-book example of prepared defense) Soviet force with concealed, tightly controlled anti-tank defenses, including tanks. The Germans moved forward into what was essentially a 4 day continuous ambush. And still they advanced. Pretty sure a Sherman couldn't have done that.

So, let's look at some of the anecdotal evidence of 8.8cm long range accuracy. (By the way, I give a lot of credence to these reports. You seem to dismiss them as tales spun near the camp-fire. Unit awards, valorous citations, etc., are based on these. They are backed up by post battle analysis. How did all those Soviet tanks get destroyed?) Anyway, to my point: most of these encounters (the ones highlighting the accuracy of these weapons) were with long fields of vision, Soviet advance or stationary and unaware, Tiger or other "long-gun" on a hill, well camouflaged, tactically on the defensive (important point). Time and again they destroy their targets. Anecdotally. Since, if they didn't tell anyone about it, we'd never know about it.

Let's use your 3-5% accuracy at 2000 meter figure. My references are in another room and I'm on a typing roll, so let's assume a Tiger (or any long 75 or 88) has an ammo load of 40 rounds. Using 5%, that means one hit per 20 rounds. So, my total load out, leaving nothing, yields 2 kills (assume a hit is a kill). Tigers often operated in pairs (a zug). 2 Tigers on a hill taking out just 4 tanks at 2000 meters??? C'mon. If that happened the crews would be reassigned as infantry.

Now, how often were the above conditions met? In the steppes, often enough to establish a reputation. In North Africa, the same.

What does this have to do with CMBO? Easy. My 8.8cm flak missed a stationary target at 982 meters 4 times in a row. The other one missed an oncoming tank, ambush, at 400 meters 3 times.

[Wreck]

Tommi writes: the whole point of the formula and two sets of values was to show that without knowing the real values of parameters the result is meaningless.

They are by no means meaningless. Only by disclaiming all knowledge of the subject to you get meaninglessness. (But then why care enough to even propose the formula?) In fact you, and everyone else reading this, knows enough about WWII to be able to use the formula to draw some conclusions.

For instance, we know that the total shells delivered to each tank were at least one full load. That's why they built them able to carry a full load of shells. Many of us also know that resupply in the field was not a rare thing, but rather, common. So each tank would get at least a couple full loads over its lifetime.

For instance, the total shells fired to kill an enemy AFV cannot be below 1.0. I used this in my previous attempt to point out to you that in fact you do know enough to use the calculations productively.

But these are relatively trivial. Even without knowing any specific facts, anyone can make guesses as to the other values in the formula. Is 10% shell delivery to the front believable? No. Is 95% shooting at bushes believable? No. Between these, your "worst case" scenario is rendered unbelievable.

Jason chooses (at least rhetorically), to regard this as evidence of dishonesty. I regard it as evidence of incomplete contemplation. (I think you need to take Jason a bit less seriously.)

50% delivery to the front, and 50% shot at bushes (or cleared from the barrel when wanting HE, etc.) -- those I might believe. Though I suspect both figures are still too pessimistic by far. But those give a very different result, don't they?

What do you believe? Or perhaps I should say, what do you believe?

[Brian Rock]

It's a pity this has become so heated, as there has been quite a bit of provocative discussion over the last few pages.

My perspective is that CM isn't broken, but that it could do with some adjustment at the edges.

One point that I think needs to be made up front: I don't think anybody is denying that the 88mm gun was a better gun than the 75mm, or that the Tiger was a better killer than the Pz.Kpfw.IV, Sherman, etc. The debate is how much better, and how that should be represented in the game.

What usually prompts these debates is someone has one or two experiences where they don't get the results they expected and this is generalised to "88mms are undermodelled". I'm not saying this means that 88mm isn't undermodelled, simply that a small non-random sample is very poor evidence for it.

Part of the problem is that expectations are built on largely on anecdotal evidence, it seems to me that the uber-Tiger anecdotes are highly selective. Tiger-philes recount countless stories of a Tiger singlehandly blunting attacks by Russian divisions, but they never talk about the Polish tank commander mentioned in "Victory at Falaise" who took twelve Shermans up to 100 metres of sixteen German ‘Tiger-type’ tanks, killed eight and made the rest bugger off. The former is presented as valid evidence, the latter is dismissed as some strange anomaly.

I once had a debate with some guy on the Talonsoft board who had a bad case of "Wehrmacht penis envy" (to quote Bolger), and he simply refused to even accept the stories of Tigers failing. His answer to these was to provide another case of Tigers achieving tactical successes.

(He also refused to concede that the fact the Germans lost every one of these campaigns meant that the Tiger may not have been the operational uber-tank he wanted it to be, but that's another thread, isn't it?)

The point is not that the Tigers weren't capable of exceptional performances, but how typical was it, and what are the variable that made these possible?

I think JasonC's point about a large number of cases will eventually throw up some outliers that are exceptional at either end. If so any weapon system will eventually rack up a series of surprising kills, and a better weapon system will probably produce more extraordinary results at its end of the bell curve (eg a 37mm gun is going to struggle to kill 17 tanks at 1200-1500m unless you are prepared to run a very large number of trials).

Now it may well be that we are getting too many at the "underperformance" side of the bell curve due to the gun being undermodelled, but I haven't seen any conclusive evidence that it's far off. The reasons I think the current model is close are:

1) JasonC's suggestion that units will take low PK shots when they can. This seems reasonable to me, and supported by data I've seen about ranges tankers will open fire. There are cases of units opening fire when they had PKs of 0% (!). Given this, if I was a Tiger CO I'd be quite happy to take shots at 2000m+ if I only had a 5% PK. Still, I concede that this in no way proves this is how Tiger COs did react, merely that it is a logical and tactically sensible response.

This is probably the norm for most Tiger crews. In this case the current modelling is close enough for all practical purposes.

2) I also think it's reasonable that superior crews will get superior results (the "ace" factor). It is common for a relative handful or people to make a disproportionate number of kills, be it aerial combat, infantry fighting or armored vehicles. Some of this will be due to random variation, but it's also true that some people are just better at certain skills than others, and when we look at the truly exceptional experiences we are probably looking at less than the top .5% or so of the population. These units are undermodelled in CM. I'd have no problem with future versions of CM adding a new class of unit that represents the "best of the best" that is significantly better (we can argue the numbers later). I would also make them *hideously* expensive.

On a separate note, I actually like Wreck's idea of "ghost" units appearing. I'd think this could be related to crew quality (green troops see more boogeymen) and environment (increased likelihood at night).

[PJungnitsch]

That's a pretty good summary, and this quote is probably what this thread boils down to.

<BLOCKQUOTE>quote:</font><HR>The point is not that the Tigers weren't capable of exceptional performances, but how typical was it, and what are the variable that made these possible?<HR></BLOCKQUOTE>

BTW, the engagement I posted was not, as has been pointed out, meant to glorify the Tiger as an all around wonder machine (there was actually an engagement in the same book in which the Tigers did extremely poorly -close range fighting in soft terrain ).

But it was meant to show that for a time at least mid war Tigers could wreak havoc especially if they could take on the T-34 opposition at the extended ranges this thread is about.

As far as factors that would enable them to do this (aside from the high velocity of the gun) could be listed precision manufacture of gun and aiming components, heavy stable gun platform, high resolution optics, and well trained crews. Use of rangefinders by at least some gave an added advantage.

Plus as well I imagine the policy of using the heavy battalions as 'firefighters' gave the crews lots of practice while the heavy armor kept them alive, which would help moving the talented crews up the 'ace' ladder.

[Jeff Duquette]

Some additional fodder for thought. The following is derived from several Operational Studies conducted by both the British and American Armies in ETO 44-45. The studies are:

"Data On World War II Tank Engagements Involving the US Third and Fourth Armored Division" By David Hardison. This study includes a database of 129 different Tank on Tank engagements fought by the 3rd and 4th AD in ETO Sept 1944 through Feb 1945.

"A Survey of Tank Warfare in Europe From D-Day to 12th August 1944" AORG Report No. 6. This study includes a database of 112 separate tank engagements fought by Commonwealth armored units in Normandy.

Here are some basic conclusions drawn from these little gems of information.

<UL TYPE=SQUARE>

<LI>From AORG 6: The average range of engagement within some of the more open terrain of Normandy around for example Caen, or Verrieres Ridge was approximately 1200m with a standard deviation of 735m.

<LI>From AORG 6: The average range of engagement within closed terrain of Normandy ala Villiers Boccage was approximately 405m with a standard deviation of 380m.

<LI>From AORG 6: The side that fired first was typically the winner of tank vs tank engagements. Apparently 77% of the actions assessed indicated the side that fired first prevailed in the engagement. This statistic held true even when the side firing first was outnumbered by as much as 3:1.

<LI>From 3rd and 4th AD report: Average range of engagements was approximately 1000meters. This includes both closed Terrain and Open Terrain engagements. This avg. is heavily weighted by two operations\campaign fought in relatively closed terrain one of which is the Ardennes.

<LI>From 3rd and 4th AD report: 50% of the engagements assessed involved one side rapidly disengaging after suffering only a few casualties. Total annihilation of a force consisting of more than 3 tanks was "rare".

<LI>From 3rd and 4th AD report: A great advantage was had by a force which initiated the aimed-at-target direct fire action. The ability to initiate the action was associated with the tactical employment of the weapons, the "defenders" possessing a marked advantage. (i.e. the fellow that fires first is at a big advantage…the fellow in defense is also at a big advantage)

<LI>From 3rd and 4th AD report: Of eleven engagements in which they fired first, the attackers lost twelve of eight-eight weapons while destroying thirty of sixty-four defending weapons. Of fifty-seven engagements in which they fired first, the defenders lost twenty-two of 238 weapons while destroying 154 of 397 weapons. Thus in sixty-eight engagements thirty-four of 326 weapons were lost by the side which fired first while destroying 184 of 461 opposing weapons.

<LI>From 3rd and 4th AD report: The records repeatedly indicated that the tank engagements generally developed suddenly, often unexpectedly, and usually ended quickly.

It should strike one as odd that the tank firing first wins on an average of 77% of the engagements assessed. Odd that is if each side is expending 100 or 200 rounds apiece to hit an enemy tank as has been suggested on this thread. If we take the route of establishing accuracy as a function of total ammunition produced divided by guns produced to arrive at battlefield accuracy it would seem inconsistent that he who fires first wins. One hundred rounds expended per kill implies accuracy in gunnery plays absolutely no role in tank combat. It suggests that scoring a hit on a target is a function of pure luck. Therefore there would be no particular advantage derived from firing first. If 100 rounds per kill is indeed a benchmark (A theory which I personally feel lacks any real substance…or hard data) the side that would most often win an engagement would be the side with the largest ready racks in their tanks…he with the most ammo stuffed in his turret always wins.

The skills typically emphasized in tank crewman training of the period are quick and accurate range assessments, rapid target acquisition, and rapid accurate fire. He who fires first and accurately wins. Speed and accuracy.

The US Army's watch words for tankers used to be (back in the 1980's) If you can see it you can hit it…if you can hit it you can kill it. Tank crew drills today emphasize basically the same critical factors of speed and accuracy. Again with the CAT score here, but crack M1A1 crews could complete an engagement at range of 1200 to 2000m in about 7 to 16 seconds. These are surprise targets (pop ups) depicting hull-down tank sized aiming areas. Average engagements involved 1 round one hit. Accuracy being 93% for CAT teams in the M1A1. Leopard II crews were in the same realm being 92 to 96%. The more modern training standards stressing urgency and speed would seemingly imply training standards back in the oldin' days weren't to far off in their emphasis on what was important. This seems to make sense if we assume that he who acquires and fires first typically wins. But rapidity of crew drill would really have little effect if 100 plus rounds fired is resulting in only one hit.

To switch gears, the "uber" Tiger anecdotes are not so far fetched when one considers that a Tiger can calmly plug away at 1500m to 3000m at T34/76's or Sherman (75) without having to continuously displace to avoid being aquired. Tends to have an effect on a gunners aim when the gunners tank has to move between each round fired. Each time you do this the gunner must reaquire his target. Braketing advantages for multiple shots is limited if not eliminated in this case.

So the Sherman and T34 in the attack is conducting: quick halt fire…moveout…halt…reaquire…fire…move out.

The Tiger in a defensive posture is conducting:…fire….fire…split the bracket fire …hit…cease fire…traverse and engage next target.

The armor on a Tiger gave its crews confidence that the T34/76 or Sherman 75 wasn't going to punch a hole in it at moderate to long ranges.

The Tiger as an uber vehicle continued. The German tank ace that most wargamers could name off the top of their head would have been a fellow that earned his rep in a Tiger. The most infamous tank engagement in Normandy would have been a Tiger Crew that just about single handedly stopped the advance of the British 7th Armored Division. To the GI every other Panzer they saw was thought to be the dreaded Tiger. I am willing to bet that more literature has been produced on the Tiger I than any other tank in history. The mystique of the Tiger was such that it still infiltrates American Culture in the form of WWII movies. Moriarty to Kelly in Kelly's Heros "WHEN WE WAS IN THE BOCCAGE COUNTRY WE WAS ASSUALTED BY TIGERS!!!…YOU KNOW WHAT I MEAN BY ASSAULTED!!!…I MEAN ASSUALTED!!! I reckon there is probably a reason for all the "bad-mo-fo" mystique surrounding the Tiger although some nuevo historians would smugly tell us the Tiger really wasn't so hot.

These "Tiger tales" are also not so fanciful when one considers operational studies detailing Tiger losses as a function of Allied tank or anti-tank gunfire. The percentages of Tigers KO'd by gunfire is relatively small when matched against Tigers having to be abandoned from lack of spare parts or lack of petrol. Example:

AORG report No. 17 detailing German Tank Losses in Normandy 6 Jun to 31 Aug.

35 Tiger I's in the sample (recovered by Commonwealth Units over the time period considered)…22% KO'd by gun fire…74% abandoned.

This same argument can also be somewhat made for the Panther. Also from AORG No. 17:

154 Panthers in the sample…30% KO'd by gunfire…5% by PIAT…5% by Artillery…7% from air attacks…48% were abandoned.

Of additional interest from AORG No. 17…

Avg number of hits to knock out a Tiger 4.2

Avg number of hits to KO a Panther…2.55

Avg number of hits to KO a MkIV…1.2

Avg number of hits to brew up a MkVI…5.25

Avg number of hits to brew up a MkV…4

Avg number of hits to brew up a MkIV…1.5

One additional bit here. If we assume 100 rounds of ammunition are expended per hit as is being suggested on this thread, the implication is that the Tigers assessed in this study were requiring on average 420 rounds of ammunition being fired at them. 525 rounds if the Allied tank crews involved were of a mind to get the thing to brew up. And these 525 rounds were all being expended in an engagement that "developed suddenly, unexpectedly, and ended quickly".

[Jeff Duquette]

ohpps...double post

[ 10-05-2001: Message edited by: Jeff Duquette ]

[Berkut]

This has been an informative debate, even if it got a little heated. Much like one I was involved in regarding the penetrative abilities of AP shell and AP shot. (I'm sure M.Hofbauer will remember that one, we agreed to disagree and moved on)

I think we need to accept that we all have extensive libaries of military history books to back up our opinions. But we shouldn't allow ourselves to be dragged into a shouting match over these things...it's not worth the stress

We should remember that this is a computer wargame, and although BTS have done their best, it cannot model reality 100%

At the end of the day, all we have is a collection of books and personal anecdotes to draw on. None of us fought in Tigers during Kursk nor sat in a sherman looking through a drivers slit as an 88 picked off your mates during operation goodwood.

The Tiger was a damn good tank, and in the right hands was capable of astounding things. But there were more nasty ways to die out there, than at receiving end of a Tiger's main gun & they didn't all have an 88 as their main armament.

I think the genuine fear of the Tiger tank, lent it an almost mythical status. Something that all allied tankers dreaded coming across. And it was this Mystique that has come across to us here in the present day, and coloured our thinking.

Like the anecdotal story of the soldier, who is saved from death by a bullet hitting the steel shaving mirror he had left in his uniforms breast pocket...We have stories of Tiger tanks, picking off targets from great ranges, and stories of Tiger tanks killing off large numbers of vehicles during close range engagements.

Sure some of these things did happen, but they were the exception rather than the rule.

[Wreck]

<BLOCKQUOTE>quote:</font><HR>

It should strike one as odd that the tank firing first wins on an average of 77% of the

engagements assessed. Odd that is if each side is expending 100 or 200 rounds apiece to hit an enemy tank as has been suggested on this thread.

<HR></BLOCKQUOTE>

Suggested by who? If you read the thread you will find no such suggestion. Rather you will find the idea that 400 rounds produced per gun and only a handful of kills per gun means something. What you make of it depends on what you think happened to the shells, but smart folks come to the conclusion that the shells fired per kill are in the range of 10 or more, and perhaps higher.

<BLOCKQUOTE>quote:</font><HR>

If we take the route of establishing accuracy as a function of total ammunition produced divided by guns produced to arrive at battlefield accuracy it would seem inconsistent that he who fires first wins.

<HR></BLOCKQUOTE>

It is only inconsistent if you assume that all engagements have the same accuracy. If you allow differences in accurate (i.e., for range, say), then firing first can be a statistically high value thing at the same time that average rounds fired are high.

For example, consider this extreme made-up case. It may be that one particular Tiger fired over a million shells and never got a hit, while 1000 other Tigers each fired just one shell, but hit and killed with each one. In this case, the average shots/kill is 1000, but every single kill obtained was a first shot kill.

A plausible example has a large number of engagements happening at short ranges, leading to one or two shot kills, while there are still other engagement, fewer in number, which take place with very low accuracy and therefore expend a lot of shells with little effect.

<BLOCKQUOTE>quote:</font><HR>

One hundred rounds expended per kill implies accuracy in gunnery plays absolutely no role in tank combat.

<HR></BLOCKQUOTE>

Wrong, as in my example.

[JasonC]

Wreck suggests a more charitable reading of incomplete connections between thoughts, and I am quite open to that suggestion. Charity is a useful principle in debate, and I take the point.

Jeff raised some excellent points but also misunderstood my own claims rather seriously. I have never suggested the average shots fired per kill are 100. I defending the idea of 3-5% hit probabilities for fire well beyond 2500 meters, stated that the roughly 1/5 - 1/6 hit probability noticed in a test at 2000 meters seemed about right, and insisted that the higher accuracy at closer ranges must be balanced somewhere by a form of shooting in which ammo expenditure per hit is high. I said I think the average accuracy per shot has to be around 10%, and since accuracy is certainly higher than that for closer range fighting, there has to be some firing going on at less than that accuracy.

The distinction between average shot and average kill must be understood plainly. This would seem to be the basic issue between me and Jeff. I hope I can explain it clearly. I preface it by saying I see nothing in the data he presented that contradicts my own picture of what happened.

If you look at all kills, you will find some portion of them made at close range by little ammo, some additional portion at medium range using somewhat more ammo, and some presumably small portion made at long or extreme range using significantly larger amounts of ammo. The average kill probably takes place at medium to close range. But the average round expended may be expended at long range. The two things are not the same, because ammo use is weighted toward the engagement types that use lots of it.

I will illustrate numerically in case that is difficult to follow.

5 kills with 1 round each - 5 rounds

10 kills with 2 rounds each - 20 more rounds

15 kills with 3 rounds each - 45 more

10 kills with 5 rounds each - 50 more

10 kills with 8 rounds each - 80 more

10 kills with 10 rounds each - 100 more

5 kills with 15 rounds each - 75 more

5 kills with 20 rounds each - 100 more

3 kills with 25 rounds each - 75 more

2 kills with 30 rounds each - 60 more

75 total kills, 610 total rounds fired, average shots per kill 8.13.

But now look at the median kill, the 38th one - it falls in the category of 5 rounds each. Look at the median shot fired, the 310th one - it fails at the 15 rounds per kill line.

That is, the average -kill- is made at 20% accuracy (and thus within one minute). But the average -round- is fired with only 6.7% accuracy (and thus in engagements up to 3 minutes long).

Now, notice what happens it we extend the table to include more extreme-range fire, with say 2 kills each 35, 1 kill with 40, and 1 kill with 50 rounds.

We just added only 4 kills, bringing the total to 79. The median, 39th kill still falls around 20% accuracy, maybe a little less. But we have added 160 rounds fired, bringing the total to 770. The average shots per hit rises to 9.75. And the median 385th shot occurs at the 5% hit probability line (5 with 20 each, above).

In either case, over half the kills are scored with 5 rounds or less. But that does not mean the average round is fired at a range close enough to have a 20% hit probability. Over half the rounds are being fired at hit probabilities below 7% or below 5%, depending on how far out the extreme range fire extends.

(Of course these numbers are illustrative only, but I tried to keep them reasonable close to what I think happened).

It takes more ammo to get even a modest number of kills at extreme range. The number of kills at such ranges being modest, this leaves the average -kill- largely untouched - most of them still occur at close ranges with relatively high hit probability. But it increases the total -rounds fired- significantly, and so moves outward the hit probability of the average round fired.

Also, I never suggested even that the mean rounds fired per kill are 100, I actually suggested around 10 (which fits with the extended version of the tables above, incidentally). I said explicitly that factors like supply chain, rounds lost in dead tanks, overrun depots, hits that don't kill, and overkill, etc, might well account for -one- order of magnitude of difference between rounds produced and enemy tanks destroyed. I denied they could account for -two- orders of magnitude. Thus hit probabilities on the order of 10% for the average -round-, I consider plausible and consistent with the known production figures, and suggested quite a while ago.

The point has been all along, that with such an average hit probability (around 10%), and with it being well known that a significant portion of the kills are scored at closer ranges where the hit probability per round is high, there has to be a significant expenditure of ammunition somewhere else to account for the remainder. The average -kill- can still be at close range with high hit probability, while the average -round- is expended with poor hit probability.

So far in response to Jeff's arguments, which other than misunderstanding my own with the 100 per hit notion (when I said around 10 quite a while ago) and the difference between the average kill and the average round fired, I basically agree with.

Next there is the confusion over uber-Tigers. One fellow still doesn't see anything unbelieveable in the idea that an average Tiger is killing 17 tanks per outing without loss, and thinks if they killed 2 each without loss in one engagement they'd be busted out of the panzer service. I think the latter performance was about average, as I will try to show.

I recur to the Kursk case, and I now realize that I cannot assume anybody nows the raw data.

There were 133 Tigers at Kursk, not 2. There were another 90 Elephants, with even better range. Also 90 Nashorn and at least 120 towed 88mm Flak (just in corps level units). That is 433 88mm weapons at least.

The Russian field reports say 20% of their dead tanks had 88mm holes in them. Over the course of 2 months, they lost 5000 tanks. Suppose they may be lowballing the percentage and it is really 1/4, to be generous. There are then 1000-1250 dead AFV killed by 88mm.

Say the Nashorns are nothing special and the towed Flak relatively ineffective, because neither has the armor. They can still account for 150, giving the Flak only .5 each and the Nashorns an average of 1 - which are extreme low-ball estimates in both cases.

Then 850-1100 AFVs are left to be killed by 90 Elephants and 133 Tiger Is. Right there you can see an upper bound for the average kill per AFV is 4-5. The Elephants claimed 500, which might be right or might be high by a factor of 2, conceivably, because battlefield kill claims often are. That means a range of 350-850 for the Tigers themselves, or 3-6 each. Not in one afternoon, in -2 months-.

If one is more generous to the Nashorns and gives them 2 each, and the Flak and gives them 1 each, and takes the 20% figure at face value, then 3 per Elephant or Tiger is what comes out. Depending on how much you believe the Elephant claims, the average for the Tiger Is could be as low as 2 each. But the center of all of these accountings, which give a range of 2-6, is more like 4 each, and that is a plausible figure in itself. I think it is right.

Now, notice how completely implausible 17 each per outing without loss is, by comparison. Alone the Tigers would get 2261 kills in one outing each. Surely they can get more than one outing each, even with a somewhat declining total operating, in the course of 2 full months. Perhaps they can get such a fight in once a week, and perhaps half of them are left by the end. Then the Tigers alone ought to account for a measly 13,566 AFVs in the course of Kursk.

Can we still account for the anecdotes, knowing the realistic average number of kills per Tiger in the course of a 2 month battle is only around 4? Yes, we can.

Suppose in each engagement the Tiger wins or loses, and wins 4 times as often as it loses. Then the average over the full 133 tanks will be the same as the mean created by the ratio of those probabilities, thus 4 each. But some of those will be Tigers that were put out of action (by whatever, you can include breakdowns if you like) before they got any kills at all. Others will be tanks that got impressive runs, which were then lapped up by anecdote writers for the armed forces propaganda magazine.

How many would you expect to get long runs, with 133 trying and a prior probability of a win of 4 to 1 in favor? You'd expect 20 of them to get runs 17 long before being knocked out. By pure chance, without accounting for any superior skill effects at all. Set 133 casino players rolling dice, and see what the longest run before throwing a "1" is. Same sort of effect.

Mining outliers for uber tales consists in taking the outliers - even random ones will do - and pretending they are the average. The same technique is used to sell lottery tickets. The luckiest fellow in a large set of players is paraded around as proof that you too can win. As a mere morale booster, that is fine. It encourages emulation, and builds confidence about the risks being run. But as supposed proof of mean behavior, it is ruinous statistical nonsense.

Does this mean the Tigers weren't superior to the Panzer IVs? No, of course not. But the kills from each Tiger at Kursk were probably around 4, and those from each Panzer IV long or StuG long were probably around 1.5. Not 0.8 vs. 17 times the number of outings. That is an uber tale, created by comparing an average to an outlier, and it overstates the superiority of a Tiger over a Pz IV long by about an order of magnitude.

Incidentally, the same 4:1 probability of a win, over the whole war's population of 1350 Tigers, you'd expect one of them to get a streak 42 kills long without losing his own tank once. Some German aces got higher total scores, but by going through more than one tank, and including scores carried over from previous vehicles. Presumably some skill factor was also present to make for even longer runs. Because one run 42 long - which after the fact you can be sure everyone would ascribe to uber skill - would actually be expected by chance alone.

Superior vehicles create longer runs because they increase the probability of success at each trial. A larger vehicle pool can also help create an larger outlier, though much less powerfully, unless combined with a high success probability per trial.

I hope this remains interesting.

[Andreas]

This is an excerpt from an AAR by Hauptmann Friedrich Stück, Sturmgeschütz Brigade 276, from a fight in a village named Babin(?) on the Rowno-Nowgorod-Wolynsk road.

<BLOCKQUOTE>quote:</font><HR>I could not say how long the matter lasted. We had used all our aproximately 80 rounds; maybe that's a sign how long it lasted?<HR></BLOCKQUOTE>

Hauptmann Stück's Stug knocked out three T-34s in a knife edge fight, firing at point-blank ranges (one T-34 attempted to ram his Stug after the cannon was knocked out). Another was knocked out by an AT gun, the last by another Stug. The one that was knocked out by the AT gun was driven back by Stück using HE, which made some of the T-34 crew bail.

I know, it is not a Tiger, and it does not prove anything, but it is a good illustration as to what can happen in a close-range firefight. They just fired, any round that came to hand of the loader. ~80 rounds gone for three T-34s (three-and-a-half if you feel generous).

The account is in 'The Combat History of Sturmgeschütz Brigade 276'.

[WineCape]

<BLOCKQUOTE>quote:</font><HR>Originally posted by M Hofbauer:

this gem almost got lost amongst the firefight.[... referring to combat modifiers for TC's as mentioned by Bad Monkey!...]

I really really like that idea. It would make tanks less generic, yet still variation would be limited to a certain envelope, just like you can expect a certain minimum performance from a HQ unit even without modifiers.

I think you really hiot onto a very good idea here. And it seems to me it should be possible to include this with a fair, feasible amount of effort (it worked for the HQ units).<HR></BLOCKQUOTE>

The more I read this excellent thread, the brighter this idea! Seems also little trouble of coding this in instead of a complete engine re-write regarding the formulas used for hit%, no?

Regards,

Charl Theron

-----------------

"Sparkling Muscatel. One of the finest wines of Idaho!"

-- waiter in The Muppet Movie (1979)

[Amedeo]

Very interesting thread. About the 'you can't miss up to 1200m' and the Tigerfibel I'd like to point out (if it wasn't already done) that the Tigerfibel correctly says that the 88mm APBCB has a very flat trajectory that is less than 2m 'high'. The problem is that those 2m are to be counted from the height of the gun muzzle. So if the Tiger is firing on a perfectly flat area with its gun boresighted at 1200m, the round will reach at least an height of 4m above ground. On the other hand if the Tiger fires from a pit with the gun almost at ground level nothing taller than 2m will be missed...

Regards,

Amedeo

[M Hofbauer]

<BLOCKQUOTE>quote:</font><HR>Originally posted by JasonC:

Next there is the confusion over uber-Tigers. One fellow still doesn't see anything unbelieveable in the idea that an average Tiger is killing 17 tanks per outing without loss, and thinks if they killed 2 each without loss in one engagement they'd be busted out of the panzer service.<HR></BLOCKQUOTE>

I would be interested to know exactly *who* said those words, and *where*.

[Roksovkiy]

Like some fellow said early 'Your taking Jason C, too seriously'.

He is voicing his opinions based on speculation using a ‘play on numbers’, not hard evidence.

All primary evidence indicates the long-range accuracy of the 88L56 and 88L79 are under-modelled in the game compared to reality.

[Jeff Duquette]

<BLOCKQUOTE>quote:</font><HR>Jason Said: I recur to the Kursk case, and I now realize that I cannot assume anybody nows the raw data.<HR></BLOCKQUOTE>

With respect to the raw data perhaps fixating on Soviet tank casualties for the operation in question would be more relevant. First Kursk is being discussed, than an estimation of Soviet tank casualties suffered over a two-month period enters the discussion. If we are talking about Kursk maybe we should stick to Kursk and not insert a front wide amalgamation of tank casualties for a time period that far exceeds the actual Kursk Offensive.

Recomended references:

“Kursk, 1943” by Niklas Zetterling

or

“Soviet Casulties and Combat Losses” by Krivosheev.

Krivosheev indicates 1,614 tanks and assault guns were lost (irrecoverable losses) by the Red Army during the defensive operations around the Kursk salient.

Zetterlings research – which is founded upon review of Red Army staff reports of the period in question – indicates irrecoverable tank losses for the Voronezh Front were 1,254 tanks and assault guns. Central Front tank\assault gun losses were 651. Total between the two fronts was 1,905.

I am more inclined toward the Zetterling figures as Krivosheev breaks out his losses by Defensive operations around Kursk and the Soviet Counter Offensive toward Orel. The Soviet Orel counterattack was occurring concurrently with the last several days of the German offensive and arguably was one of the main causes for the cessation of the German attack. Although some will argue the Allied invasion of Sicily was the main cause…be that as it may.

Lets split the difference between the two estimates and assume a figure of 1750 Soviet irrecoverable tank\assault gun losses.

<BLOCKQUOTE>quote:</font><HR>Jason Said; Then 850-1100 AFVs are left to be killed by 90 Elephants and 133 Tiger Is. Right there you can see an upper bound for the average kill per AFV is 4-5. The Elephants claimed 500, which might be right or might be high by a factor of 2, conceivably, because battlefield kill claims often are. That means a range of 350-850 for the Tigers themselves, or 3-6 each. Not in one afternoon, in -2 months-. <HR></BLOCKQUOTE>

Using your same ratios for the potential spread of Tiger I inflicted tank losses and plugging in the above figure of 1750 irrecoverable Soviet tank\assault gun losses – we have a 7% low end (i.e the 350 figure) and 17% high end (i.e. the 850 figure) – we are left with a spread of 123 to 298 Soviet tanks\assault guns KO’d by Tigers.

Now than Zetterling also provides us with an exhaustive bit of research on German staff reports and maintenance\recovery units for the German Army at Kursk. For the period in question German irrecoverable Tiger 1 losses included:

Army Group Center, July 5 – 14: 4 Tiger I’s

Army Group South, July 5 – 17: 6 Tiger I’s

Ratio of irrecoverable losses: 10 Tiger 1’s to 123 to 298 Soviet tank and assault gun losses.

So for every Tiger tank lost, Tiger tanks inflicted 12.3 to 29.8 Soviet tank casualties.

[ 10-06-2001: Message edited by: Jeff Duquette ]

[JasonC]

If Jeff would have it that in the offensive phase, the average Tiger got 1-2 kills while their losses were small, and in the defensive phase they got another 2-3 kills to total my 4 for the whole period through August - with additional presumably minor losses incurred, and with breakdowns probably running ahead of battlefield losses, and many more in short term repair than total write offs - I have no objection to any of that, and consider it perfectly realistic.

Still isn't quite the picture of the 17 each anecdotes being ordinary, is it? In fact, it would fit rather well with the 2 each without reply that another fellow thought would get the crews busted out of the Panzer arm.

The reason I used the longer period for loss accounting, incidentally, is that is the period for which the 20% by 88mm survey figure is available. We don't know for certain, one way or another, whether the losses in the offensive phase had a similar ratio of 88mm kills to kills by other calibers, or not. The Russians couldn't go over the wrecks measuring holes until after the whole battle.

Some may wonder how the losses to each vehicle can be so low, when the shooters are still intact (as most are over short periods of a week or so). I think the primary answer can be found in the reports on the western front fighting, that typically the losing side breaks LOS quickly. That probably prevents every tank in the winning formation from getting kills. The underlying cause is the way the opposing armor fleets creep into each other's lethal and visible zones tenatively, with only portions overlapping.

[M Hofbauer]

forget it.

[ 10-07-2001: Message edited by: M Hofbauer ]

[RMC]

LOL.

[c3k]

JasonC,

What do you mean by way of reply? I'm the 2 kills would get them busted out of the panzer troops poster (my paraphrasing of your paraphrasing of my actual posting).

I saw nothing posted requesting a reply.

If you maintain that for the duration of the war, on average, it took 10 shots per kill, fine. If you maintain a Tiger, set up, with long site-lines, stationary, yadda, yadda, only had a 3-5% chance of a hit at 2000 meters, again, you're wrong.