quote:

---

Official accounts give varying numbers of German vehicles destroyed at Abbaye-Blanche during the five days of the Battle of Mortain. A reasonable estimate would be at least 24 vehicles, including 3 Mk V tanks, 8 Mk IV tanks, 9 half-track armored personnel carriers or assault guns, 3 cargo trucks or cargo vehicles, 3 full-tracked personnel carriers, 2 motorcycles, I staff car, I armored car, one command car (Kuebelwagen), and at least 10 other vehicles, as well as 150-200 German infantry killed and 17 captured, and one machinegun nest destroyed. Lt. Andrew, in an interview soon after the battle, guessed that the roadblock knocked out at least 60 vehicles, many caused by the 3-inch TD guns under Lt. Springfield. He also said that 12 German armored vehicles had been destroyed on Highway 3, apparently by bazooka teams or 57mm guns, but that they had been hauled away at night by the Germans. These were in addition to the TD gun's kills. The 120th Inf Regt history mentions that one of the roads near the roadblock, probably Highway 3, had at least 24 destroyed German vehicles on it alone, all lined up bumper-to-bumper as if destroyed while parked that way. Official losses of the roadblock garrison were 3 killed and 20 wounded.

---

I have a question regarding tank strength returns indicated in Normandy 1944 for the 12th SS Panzer. Strength returns for 1 June indicate 91 Panzer IV’s Combat Ready, and 7 in Short Term Repair. The next return for 16 June indicates 52 MkIV’s combat ready and a question mark in short term repair. Can I assume that information is just not available for vehicles in short or long term repair for this period?

The 26th June strength return indicates 60 Combat Ready MKIV’s and 12 in short and long term repair.

In your section on replacement vehicles sent to 12th SS it looks like no new MkIV’s made it to the division before 8 July.

Can I assume that MkIV losses between 1 Jun and 26 Jun are:

98 combat ready\short term repair on Jun 1 minus 72 combat ready\short term repair on Jun 26 equals 26 KO’s during this time period?

How valid do you feel this approach is for other Panzer Divisions strength returns indicated in your book?

quote:

---

Originally posted by Jeff Duquette:
[b]Niklas:

I have a question regarding tank strength returns indicated in Normandy 1944 for the 12th SS Panzer. Strength returns for 1 June indicate 91 Panzer IV’s Combat Ready, and 7 in Short Term Repair. The next return for 16 June indicates 52 MkIV’s combat ready and a question mark in short term repair. Can I assume that information is just not available for vehicles in short or long term repair for this period?[/B]

---

If there was a report, it always gave the number of operational tanks. Quite commonly it gave tanks in “short term repair”, occasionally it gave tanks in long term repair. Hence, your assumption is correct.
BTW, it seems that the major factor deciding whether all three categories were reported was the formation itself. To exaggerate slightly, those units subordinated to 47th Panzer Corps did not report anything. The 10th SS-Panzer Division, on the other hand, almost always reported operational and the two categories of repairable.
The fact that 47th Panzer Corps reported poorly is not an impression caused by the loss of its records. The main source I have used for the tank strength returns is reports compiled by the 5th Panzer Army, based on information sent to it from its subordinate corps.

quote:

---

The 26th June strength return indicates 60 Combat Ready MKIV’s and 12 in short and long term repair.

In your section on replacement vehicles sent to 12th SS it looks like no new MkIV’s made it to the division before 8 July.

Can I assume that MkIV losses between 1 Jun and 26 Jun are:

98 combat ready\short term repair on Jun 1 [b]minus 72 combat ready\short term repair on Jun 26 equals 26 KO’s during this time period?
[/B]

---

On 1 June the 12 SS-Panzer division had no tanks in long term repair, hence the overall strength of the division was 98 Pz IV. As you see in the table there is an asterisk at the figure for the number of tanks in repair on 26 June. This figure gives both short and long term repair. Before 26 June the 12. SS-Pz.Div. had a period of lull, which enabled it to repair most of its lightly damaged tanks. But the reason why I gave a combined total for the damaged, and no separation between short and long, is that the report actually the number of operational and completely destroyed. Hence I actually arrived at the number of damaged by simple arithmetic’s.

quote:

---

How valid do you feel this approach is for other Panzer Divisions strength returns indicated in your book?

---

You have to include vehicles in long term repair, else the calculation will be flawed. Since figures on vehicles in long term repair are less easy to find, this can make the approach difficult to follow, but look for example at 10. SS-Pz.Div.
The method is of course not fool proof. It is necessary to have data on deliveries of new tanks, but very few new tanks were sent to Normandy, and I think I have found most of the shipments, even though there occasionally can be some hesitance on the date they actually arrived. A few tank may also have been sent from schools in France, for example Mailly-le-Camp. I have no records on the latter, only an order from August that said that the Panthers with the units training at Mailly-le-Camp should be available as a reserve to OB West. If any of them actually made it to Normandy is unknown.
Another potential problem is that tanks sent to rear facilities for repair need not have been sent back to its parent unit. Generally the Germans strove to do that, but they need not have succeeded in every single case. Hence you can conclude that my figure on tanks in repair on 26 June is not absolutely accurate. Nevertheless, I think it is perhaps within a very small interval.
Similarly, the objections given here do not suggest that the methodology is seriously inaccurate, rather I would say that usually it is within perhaps +/- 5 % correct.

Niklas Zetterling

quote:

---

Originally posted by Jeff Duquette:
[b]Niklas:
How valid do you feel this approach is for other Panzer Divisions strength returns indicated in your book?
[/B]

---

I can add that when Anders and I did our research on Kursk, we found that the 4th Panzer Army records contained reports giving the tank strength for the 1st, 10th and 20th of each month (I am quite convinced Chris has copies of these reports). These were in file BA-MA RH 21-4/450. These reports gave the strength according to T/O&E and actually on hand. The latter was divided into three categories: combat ready (category 2a), in repair with the unit (category 2b), and in repair facilities at army level (category 2c).
Also the reports showed losses within the preceding ten days. These losses were divided into vehicles sent to facilities well to the rear (could even be in Germany) (category 3a) and “Totalausfälle” (captured by the enemy, destroyed or damaged to such an extent that it was not economical to repair) (category 3b).
Also the reports gave the number of tanks that been added to the unit in the preceding ten days. There were three basic categories of the latter: returning from rear (cat 3a above) repair facilities (which was called category 4a), new tanks received from OKH (category 4b) and captured enemy tanks that could be put into use (category 4c).
Had I found such reports in the records pertaining to Normandy, I would have been saved much work, but unfortunately I have not found any such reports in the records of 7th Army or 5th Panzer Army. In general, staff work seems to have been sloppier at Normandy than at Zitadelle.

Niklas Zetterling

quote:

---

Originally posted by Niklas Zetterling:
I can add that when Anders and I did our research on Kursk, we found that the 4th Panzer Army records contained reports giving the tank strength for the 1st, 10th and 20th of each month (I am quite convinced Chris has copies of these reports). These were in file BA-MA RH 21-4/450.

---

Yes, I have them, and they are available in the National Archives. They were very useful. The German daily ready-for-action reports on tank strength tended not to report on command tanks, observation tanks, panzers I & II, lumped together all Pz IIIs, regardless of gun, etc. They were useful for seperating these out and getting some record of losses to them.

They were also the only source of strength and losses to halftracks and armored cars.

Finally, they actually provided the best soruce for the TOE for the German armor for that division at that time (and German TOE changes regularly and is never the same from one division to the next).

Chris: I think my last post on page three of this thread got lost in the shuffle to pg4. I would be interested in your thoughts on the following...

quote:

---

Chris Lawrence Said: In the case of the Kursk air campaign, where I had both German and Soviet claims and Soviet and German kills from unit records, I did do this test. When I compared claims to kills on a daily basis, they varied widely from day to day. Furthermore, I could see no correlation between German losses and Soviet claims. There was however, a pretty clear correlation between actual Soviet losses and the Soviet claims of German losses.
There problem is that one cannot assume a direct relationship, becuase 1) it varies widely and 2) what is causing the "overclaiming" may have nothing to do with actual enemy losses.

---

Interesting. Although it is not particularly surprising that claims to kills would very widely on a day-by-day or engagement-by-engagement basis. What was the over-all claims to kills for the entire operation (time line -- say – kick off of the German Attack up through cancellation of the southern pincer of the attack), and how was either side investigating\confirming ground kills relative to air claims? Surely there isn’t anything equivalent to ORS studies generated by the German Army or Red Army?

Secondly as you have brought up the Russian Front, I am curious how coordination of various “tank busting aircraft” (ala HS-129B-2, JU-87G-1 or Shturmoviks) was conducted between ground and air elements. Moreover were the Soviets and\or Germans employing Air liaison Officers operating from tanks\AC’s ala the Anglo-American Armies in NW Europe? Or were tank-busting sorties employed more along the lines of a roving sorties over the battlefield? Assuming the former, what sort of reaction times were typical between a request for CAS and aircraft appearing over\engaging the target?

And Lastly; I vaguely recall an incident during Kursk in which it is claimed that either Ju-87G’s (or perhaps they were HS-129’s) destroyed a large portion of a Soviet Tank Brigade which had been assembling for a counterattack. Based upon your research is there any truth to this story? (Sorry the finer details of this engagement somewhat allude me…I can provide more information on this event if it is unclear to you what I am asking).

quote:

---

Originally posted by Jeff Duquette:
Interesting. Although it is not particularly surprising that claims to kills would very widely on a day-by-day or engagement-by-engagement basis.

---

Not only does is vary widely from day-to-day, or engagement-by-engagement, but also becuase of units involved, nature of combat, type of attacker, type of target, commanders, traditions and habit, weather, etc. The point is that it is so inconsistent that one cannot rely on it, even when deflated by some "average deflating factor", which I gather is what you were hinting at.

The second part of my comment was equally signficant, which is that "what is causing the 'overclaiming' may have nothing to do with actual enemy losses."

The general rule of thumb is to never rely on intelligence claims of enemy losses. The fact that many writers on the Eastern Front have done this has lead to many books being fundamentally flawed and as a result, poor historical works. I have seen this error made by Soviet, German, US and UK writers.

Now in some cases, there are exceptions. For example, in WWII, after the taking of an island from the Japanese, one is pretty comfortable that the body count of Japanese, added to the few who surrendered, less any civilians that may have also been there, is relatively close to their actual casualties.

I also find sometimes that reports of the size of an enemy attack in the units records is of value in getting a sense of the intensity (or percieved intensity) of a combat. Still these are often in error by an order of magnitude.

There is no substitution for looking at the records of both sides in a combat...although for more modern combat, this is sometimes difficult.

quote:

---

What was the over-all claims to kills for the entire operation (time line -- say – kick off of the German Attack up through cancellation of the southern pincer of the attack...

---

The data we have in hand is the following:

German losses by day and plane type
German claims of Soviet losses
Soviet losses by day and plane type
Soviet claims of German losses

Based upon an analysis of these, covering the period of 4 to 18 July for the VIII Air Corps and the Soviet 2nd and 17th Air Armies we showed that:

The ratio of Soviet claims to actual kills: 9.01 to 1

The ratio of German claims to actual kills:
1.02 to 1

Ratio of Soviet sorties count versus actual German day-time sorties: 1 to 1.2

This is all from my briefing "Historical Analysis of the Air War at Kursk".

quote:

---

and how was either side investigating\confirming ground kills relative to air claims? Surely there isn’t anything equivalent to ORS studies generated by the German Army or Red Army?

---

For ground action, we do have the list of tanks by type killed for each unit for each day (usually). We also have the claims of tank kills by the various ground units (usually by Corps or Army, sometimes by division), often aggregated over several days. Furthermore, we do have claims for each day by the German and the Soviet air forces of how many tanks and other vehicles they killed. We also have many individual incidents describing the effects of air attacks (both claims and reports by the people attacked). The only OR-type report we have is the 1st Tank Army's summary of cause of loss, which I believe has already been posted earlier in this thread.

With this material, we could probably come up with some measurement of the effectiveness of air support on the ground combat, and something regarding the accuracy of the claims. We have not done such an analysis and it would a little time consuming.

Thanks for the erudite response. I am looking forward to reading your commentary on my other questions.

quote:

---

Chris Lawrence Said: The point is that it is so inconsistent that one cannot rely on it, even when deflated by some "average deflating factor", which I gather is what you were hinting at.

---

This is sort of what I was implying…moreover the statistical "noise" present in "day to day", "engagement to engagement", or "target type to target type" is somewhat drowned out as the sample size increases.

To switch gears somewhat; What concerns me (or more properly interests me) regarding ORS studies are the very small samples of air inflicted AFV casualties. There is a great deal of room for slop occurring in these numbers (statistical noise) even if only a very few AFV's were either not found by ORS ground teams, or destruction causation was simply misidentified (i.e. 20mm armored piercing round through the back deck engine grills …tank looks in perfect condition yet the engine block has a ¾" hole through one of the cylinders). When we are talking about examining 100 tanks in Joint Report No.1 which identifies only 1 definite air to ground kill, and 6 possible kills, the statistics are quite easily skewed if only 3 air inflicted kills are not represented in the sample population.

On the surface Joint Rpt 1 might lead the reader to the conclusion that less than 7% of the AFV casualties examined in the Ardennes were probably air inflicted kills. However, (Hypothetically) what if during the blizzards, snow, and poor visibility conditions described by ORS Group 2 as impedances to their ground investigations result in 3 or 4 air killed AFV's either not being seen, or misidentified as to causation. Now we are potentially looking at a 10% to 12% air to ground kill ratio.

If one were interested in modeling air to ground inflicted kills of AFV's what sort of statistical range are we really talking about here? Mean = 4%, 5%, 7% ? stand dev = 1%, 2%, 5% , 7%? Presumably this range of probable error is strongly a function of # of ground attack sorties being flown.

quote:

---

Chris Lawrence Said:
This is all from my briefing "Historical Analysis of the Air War at Kursk".

---

I am looking through the various TDI reports listed on:

http://www.dupuyinstitute.org/tdipub3.htm

and I can't find your briefing. Is it available from TDI? Specific Kursk studies are catalogued K1 - K11? None of these seem to match the title you indicated above.

quote:

---

Originally posted by Jeff Duquette:

To switch gears somewhat; What concerns me (or more properly interests me) regarding ORS studies are the very small samples of air inflicted AFV casualties. There is a great deal of room for slop occurring in these numbers (statistical noise) even if only a very few AFV's were either not found by ORS ground teams, or destruction causation was simply misidentified (i.e. 20mm armored piercing round through the back deck engine grills …tank looks in perfect condition yet the engine block has a *" hole through one of the cylinders).

---

Whether the sample is large enough or not depends partly on what question it is supposed to answer. Also, to me it would seem very unprofessional if the teams did not recognise a victim like the one you describe above. It would require a quite sloppy investigator to miss such a hit, especially considering that the pilot hardly fired a single round against the tank. Rather it is to be expected that there are marks from shells or bullets around the tank if it has been hit by automatic cannon or machine gun fire from an air craft.

quote:

---

Originally posted by Jeff Duquette:

When we are talking about examining 100 tanks in Joint Report No.1 which identifies only 1 definite air to ground kill, and 6 possible kills, the statistics are quite easily skewed if only 3 air inflicted kills are not represented in the sample population. On the surface Joint Rpt 1 might lead the reader to the conclusion that less than 7% of the AFV casualties examined in the Ardennes were probably air inflicted kills. However, (Hypothetically) what if during the blizzards, snow, and poor visibility conditions described by ORS Group 2 as impedances to their ground investigations result in 3 or 4 air killed AFV's either not being seen, or misidentified as to causation. Now we are potentially looking at a 10% to 12% air to ground kill ratio.

---

In PRO file WO 171/153 there is a report on claims by allied ground forces in Normandy. It shows that units under command of 2nd British Army (including Canadians up to 23 July) reported that 854 German tanks were knocked out 6 June – 12 August, while 1st Canadian Army knocked out 39 enemy tanks 23 July – August. Also the report shows that these two formations reported 365 and 9 respectively German tanks to have been destroyed as found by captured (destroyed, damaged or undamaged) tanks. The report also shows that the US 1st Army captured (destroyed, damaged or undamaged) 195 German tanks 6 June - 6 August.
Despite the fact that the periods are different for these three armies, the general impression is that the figures on captured tanks is close to the number reported as lost by the Germans. Hence, it seems that the destroyed tanks usually were found, at least by someone on the allied side.
It should be noted that claims by air units are not included in the figures, but victims by air attack, that subsequently were captured by advancing ground troops, are.

Niklas Zetterling

quote:

---

Niklas Said: Whether the sample is large enough or not depends partly on what question it is supposed to answer. Also, to me it would seem very unprofessional if the teams did not recognise a victim like the one you describe above. It would require a quite sloppy investigator to miss such a hit, especially considering that the pilot hardly fired a single round against the tank. Rather it is to be expected that there are marks from shells or bullets around the tank if it has been hit by automatic cannon or machine gun fire from an air craft.

---

Perhaps. My profession is civil engineering and I can speak from experience that it is not uncommon for professionals (ala Architects, Engineers and Builders) to “miss” things. That is one of the reasons we build things with healthy factors of safety. It dilutes the possibility that human error will result in catastrophic failures. Humans are prone to making mistakes. Regarding this aspect of human nature one really doesn’t have to go any further than looking at people’s postings on an internet site. To proof read or not to proof read. That is the question. I am as guilty as the next fellow in not proof reading posts occasionally.

ORS Group No. 2 consisted of a relatively small group of folks that – in many cases -- functioned as Scientists during their civilian lives. Several members of Group No. 2 seem to have been; given uniforms, had some very basic military training, and sent on their way to the front. At times they began studies on seemingly whatever they fancied. We have reports from No. 2 group ranging from effects of dust on aircraft engines at Carpiquet Airfield, to the investigation of Radar Performance along the Dover Coast. That few, that happy few, that band of brothers making up No. 2 group were certainly professionals in civilian life and undoubtedly brought their own particular professional skills with them to ORS. But their range of studies while with No 2 Group might easily plop them into the category of “jacks of all trades” but “masters of none”.

Your definition of “Unprofessional” and “not recognizing an air attack victim” as being “sloppy” has already been demonstrated in the Kampfgruppe Pieper example I provided several pages back. But I certainly wouldn’t apply either of your characterizations to No. 2 Group’s Joint Report No 1. I would look at the circumstances during the ORS ground investigation – i.e. snow, blizzards and poor visibility -- as extenuating, and certainly contributed to potential error in the study. And again ORS Group No. 2 indicated that it was not particularly happy with their results on Joint Report No. 1. They acknowledge that errors may have been present in their own Study. In my mind that admission is professionalism…and if nothing else it certainly is an indication of their professional integrity. You make a mistake, have the balls to admit it...than roll up your sleeves and try to fix the problem.

quote:

---

Obviously poor visibility could result in victims to both ground and air weapons being undetected, thus it is not clear that such problems would result in skewed statistics.

---

Agreed.

Question: Jentz indicates the available panzer\assault gun strength of all German formations on the Western Front on Dec 15, 1944 was 1695...operational 1216. What was the percentage of this strength that was focused in the Ardennes?

Question: What was German and US ARMY tank\assault-gun\SP-TD strength loss in the Ardennes? The sources I have on this information are seemingly contradictory.

Thanks for any information.

quote:

---

Originally posted by Jeff Duquette:

Your definition of “Unprofessional” and “not recognizing a victim” of an air attack as being “sloppy” has already been demonstrated in the Kampfgruppe Pieper example I provided several pages back. But I certainly wouldn’t apply either of your characterizations to No. 2 Group’s Joint Report No 1. I would look at the circumstances during the ORS ground investigation – i.e. snow, blizzards and poor visibility -- as extenuating, and certainly contributed to potential error in the study. And again ORS Group No. 2 indicated that it was not particularly happy with their results on Joint Report No. 1. They acknowledge that errors may have been present in their own Study. In my mind that is professionalism…and if nothing else it certainly is an indication of their professional integrity. You make a mistake, have the balls to admit it.

---

First of all, what I would regard as “unprofessional” is for the ORS team to stand on or beside a tank, with the task to decide what made it a loss. If they would then miss what you described (auto cannon hits in the engine room) in your post dated 09-21-2001 06:03 PM it would seem sloppy to me. I have not seen the evidence that this happened in the Peiper example.
Second, your entire argument (concerning the Peiper incident) seems to rely on Peipers statement that three tanks were destroyed or damaged by air attack. When did he make this statement, and did he make it out of memory? It must be remembered that Peiper probably experienced several hundred days in combat. There is a definitive risk that his memory may confuse actions after a few years. I have worked at a project were we investigated the capabilities of Swedish brigades in various situations, using war games. I have been responsible for the combat results outcome of the various engagements that occurred. We played at a speed that was roughly 1:1 compared to real time. I had one or two colleagues occupied with this task only. It constantly struck us that after a few days, it was usually impossible to recall how many tanks that were lost in the individual engagements. Had we not made notes constantly, we could not have told this at the time that we wrote the reports. Possibly we were not as involved in the events as Peiper were in combat, but on the other hand we had the advantage of getting a clear overall picture, due to our role in the game. All orders made by the opposing staffs went through our hands, as did all reports.
Since you have my book on the German units in Normandy, you may have noted that I do quite often reference manuscripts by German officers. It is very common that they state the impossibility of giving such details from memory after a few years. Possibly Peiper (who was younger) had better memory. I don’t know, and perhaps he had some notes or anything else to base his statement on, perhaps it was something he wrote in a report immediately after the battle, but to me it seems that exact figures on losses and strength are among the first things that fall out of memory.
If, on the other hand, the three tanks were hit, but subsequently repaired, no blame should be placed on the ORS teams. But, as I have said before, should the tanks then really be classified as kills.
If the weather made observation difficult, it is obviously possible that the ORS may have missed knocked out vehicles. There is no disagreement on that and I do not imply that poor weather is an indication on “professionalism”. But I can not see that you have, in previous posts, showed that the ORS teams have made any mistakes in the Peiper action.

In fact my academic background is not in history, but in engineering science, thus I am not a “professional” historian. Hence I am quite familiar with the principles you describe for constructing things with a safety margin, to compensate for human errors (among other uncertainties). However, I think there is one fundamental difference here. If you build a bridge, you make it stronger than what, most likely, would be enough. But if the ORS reports suggest that – say – 7 % of all tanks were knocked out by air power, you can not just raise this figure. What could be done is to say that the actual outcome may have been – perhaps – between 4 and 10 % of all tanks knocked out fell victim to air power.
After all, tanks knocked out by ground fire may have been towed away and repaired, tanks knocked out by ground fire may have been missed by the ORS, etc. It is fully conceivable that the ORS reports may have exaggerated the number of victims to air. I have not read the ORS report on the Ardennes, but from some of the quotes given on this site, it seems that they may have read the claims by the air units and gone to places where air attacks had taken place, which may have resulted in a skewed sample. I don’t know this for sure, but perhaps this was the case.
In short, I do not think that the ORS reports are completely free of error, but I am not convinced that the errors made air power look less efficient than it actually was.

Niklas Zetterling

quote:

---

Niklas Said: I have not read the ORS report on the Ardennes, but from some of the quotes given on this site, it seems that they may have read the claims by the air units and gone to places where air attacks had taken place, which may have resulted in a skewed sample.

---

Check your email. I scanned in the report and sent it along to the address indicated on your profile. Hopefully I am not violating any international copyright laws here by sending you a copy of the thing.

US ARMY Tank and TD losses were 733 between Dec 16, 1944 and January 28th, 1945.

German losses seem a bit more vague. Parker puts German Panzer\Assault gun losses during the Bulge at 600 to 800.

quote:

---

Originally posted by Jeff Duquette:
Danny Parker in “Battle of the Bulge” indicates:

US ARMY Tank and TD losses were 733 between Dec 16, 1944 and January 28th, 1945.

German losses seem a bit more vague. Parker puts German Panzer\Assault gun losses during the Bulge at 600 to 800.

Do these numbers seem reasonable?

---

Hmm....

From transcripts of unit histories of 11 Armored Divisions and 27 Tank Battalions active in the ETO in Dec. 1945, I get the figure of 576 total writeoffs in total. Rich or Chris could probably make a better 'guesstimate' than I can, but the number of knocked out tanks for December could be as high as 2,900 (4:1 knocked out to writeoff ratio).

As for total Allied losses, Daniel Kane claims:

"...By January 1945, tank losses were 800 per month for the US and 200 per month for the British. The total figure for US Army Shermans of all types lost between 20 November 1944 and 20 January 1945 was 1,137. Tank fighting was costly."

quote:

---

Originally posted by Alex H:
Jeff,

From transcripts of unit histories of 11 Armored Divisions and 27 Tank Battalions active in the ETO in Dec. 1945, I get the figure of 576 total writeoffs in total. Rich or Chris could probably make a better 'guesstimate' than I can, but the number of knocked out tanks for December could be as high as 2,900 (4:1 knocked out to writeoff ratio).

As for total Allied losses, Daniel Kane claims:

"...By January 1945, tank losses were 800 per month for the US and 200 per month for the British. The total figure for US Army Shermans of all types lost between 20 November 1944 and 20 January 1945 was 1,137. Tank fighting was costly."

---

OH MY GOD! Who is Daniel Kane and why are you using him to torture me?

US First Army tank losses (total loss) from 13 December 1944 (inclusive) to 19 January 1945 (inclusive) were recorded as:

M5 Light 108
M4 75mm 299
M4 76mm 113
M4 105mm 20

Third Army losses from 17 December 1944 to 19 January 1945 were:

M5 Light 37
M4 75mm 73
M4 76mm 158
M4 105mm 3

Ninth Army losses from 13 December 1944 to 19 January 1945 were:

M5 Light 2
M4 75mm 2
M5 76mm 2
M5 105mm 0

I do not have data for the US Seventh Army, French First Army, or Commonwealth and Allied forces. However, I doubt that the Seventh Army losses would be sufficient to substantiate Kanes's claim.

The monthly losses for the 12th Army Group 20 December 1944 to 20 January 1945 (they did not in fact record losses by 'month') were as follows:

Light Tanks (M5 and M24 series) 208
Medium Tank M4 (75mm and 76mm gun) 585
Medium Tank (105mm) 62
GMC M10 69
GMC M18 27
GMC M36 26

I expect that Kane obtained his 'data' from some bowdlerized version of the above. However, 20 December 1944 to 20 January 1945 was a peak period for losses. Compare the total of tank losses (855 light and medium tanks) for that period to the earlier and later periods below):

6 June to 20 June 223
to 20 July 150
to 20 August 760
to 20 September 563
to 20 October 403
to 20 November 341
to 20 December 657
to 20 January 855
to 20 February 168
to 20 March 404
to 20 April 374
to 20 May 99

Total recorded losses was 4,875 or 406.25 average per month. Some of these were in fact recovered and may have been eventually repaired (in the First Army tank loss statistical report there was at least one or two noted -- out of 900-odd -- as "burned out" that were also designated as "repairable") However, as I noted in my earlier post, it appears that the ratio of 'write-offs' to 'repairable battle damage' was much lower in the US and British Army than you indicate and may have been as low (on average) as 1 to 1. Unfortunately, the sample of data to work with is not extensive so I cannot be sure, but there is no indication that the ratio would have been as high as the 1 destroyed to 4 damaged that you assume. In any case, if that assumption was followed, then the 4,875 losses indicated above would have been associated with almost 20,000 damaged vehicles, plus approximately 5 percent of the on hand in repair for maintenence, a figure that would probably have overstressed the already overextended US Army Ordnance Maintenence units too exhaustion.

Sorry, keep forgetting stuff. At a 5% 'normal' daily maintenence rate, an average of 296 of the above tanks would have been in repair facilities on a daily basis between 20 November and 20 December (on hand as of 20 November was 5,091, as of 20 December it was 6,741, giving a rough average of 5,916 for the month). However, overall I think the maintenence average was probably under 2 or 3 percent.

[This message has been edited by Rich (edited 09-25-2001).]

And yet on examination, I find that US medium tank losses from 6 June to 22 July (48 days) were 336, an average of 7 per day. But, about 40 of those were DD tanks lost on D-Day (evidence I have indicates that as few as one-half dozen tanks were lost to other causes than swamping on D-Day). Subtracting those gives an average of 6.3 per day lost, which is less than one-half the rate lost during the breakthrough and 'easy' pursuit (591 lost 23 July to 2 Sept (41 days) or 14.4 per day). Even if we look at the true 'pursuit' phase (when everyone 'knows' that the GIs were just rolling along handing out candy ) from 6 August to 2 September (1 August for Third Army) the totals were 266 lost in 28 days or 9.5 per day.

quote:

---

Originally posted by Rich:
This appears to be straying from the Tac Air thread, but what the heck. The last few posts regarding US Army tank losses in the Bulge got me to thinking about what we all 'know' happened in different events in the war. We all know that losses in the fighting in Normandy were extreme and caused everyone to realize that the US tanks were overmatched - right?

And yet on examination, I find that US medium tank losses from 6 June to 22 July (48 days) were 336, an average of 7 per day. But, about 40 of those were DD tanks lost on D-Day (evidence I have indicates that as few as one-half dozen tanks were lost to other causes than swamping on D-Day). Subtracting those gives an average of 6.3 per day lost, which is less than one-half the rate lost during the breakthrough and 'easy' pursuit (591 lost 23 July to 2 Sept (41 days) or 14.4 per day). Even if we look at the true 'pursuit' phase (when everyone 'knows' that the GIs were just rolling along handing out candy ) from 6 August to 2 September (1 August for Third Army) the totals were 266 lost in 28 days or 9.5 per day.

As I said, just musing about what we 'know.' Odd, isn't it?

---

As a comparison, I can add that the Germans lost about 2.5 Panthers per day 6 June - 27 July in Normandy, while the corresponding figure for Pz IV 4.3.
AFter that losses soared. They can only be estimated, but losses of Panthers 28 July - 31 August must have exceeded ten per day, with Pz IV losses slightly higher.
Perhaps this says something about the value of controlling the battlefiled.
All figures of course refer to fighting against both US and Commonwealth forces.

Niklas Zetterling

Thanks for information. Also interesting commentary regarding loses relative to period of campaign in NW Europe which is being considered.

quote:

---

Originally posted by Jeff Duquette:
Rich:

Thanks for information. Also interesting commentary regarding loses relative to period of campaign in NW Europe which is being considered.

You reckon Commonwealth tank loss\day stats for Normandy tell a different story relative to the Commonwealth pursuit period?
I am still curious as to any insights you, Niklas or Alex (or Chris if he gets the opportunity to pop in here again) may have on German tank losses in the Ardennes. My sources seem fairly vague as to this aspect of the campaign. Any recommended references that could help (aside from a full-on Archive dig)?

---

Rich gave the following tank losses for the Germans a few weeks ago:

16-20 Tiger
191-194 Panther
141-158 PzIV
179-182 StG (I assume you include JagdPz in these figures)

This would suggest that the Germans lost between 527 and 554 tanks and assault guns (I am assuming Rich meant the perid 16 December - 16 January).

As to British tank losses in Normandy, I have seen a couple of PRO documents on this, but they are quite contradictory. One shows losses in August that were about four times higher than in July, whihc seems strange indeed. It is worth noting that the British forces possessed more tanks than the US forces initially, but in August they had approximately a similar strength.
In any case I think that the British may have been slow at classifying tanks as destroyed. The day a tank was written off can have been been much later than the day it was actually hit by enemy fire. Or else the reports at higher level may have had had some other delays. I have discussed this in note 33 on page 75 and 76 in my book on Normandy.

quote:

---

Originally posted by Jeff Duquette:
Check your email. I scanned in the report and sent it along to the address indicated on your profile. Hopefully I am not violating any international copyright laws here by sending you a copy of the thing.

I will make an attempt to reply to the rest of your last post when I have more energy. It’s been a long day here

---

Thanks for the report. I have read it and my suspicion that the ORS teams went to places were air units hade made claims seems justified. Depending on the question, this can mean that the sample is biased. If the question is the percentage of the German tanks lost that were victims to air power, such a sample will show air power better than it actually was. On the other hand, if the question is the ratio between air claims and actual kills, the sample need not be biased.

Niklas Zetterling

Jeff brought up the A2D2 report that our good friend and associate Jay Karamales was involved in, specifically, the action of the 823rd TD Battalion supporting the 30th ID at Mortain on 7-10 August 1944. This area was also the subject of an ORS Report, No. 4 "Air Attacks on Enemy Tanks and Motor Transport in the Moartain Area, August 1944."

Well I did something so simple I can't belive that I hadn't done it 10 years ago (getting old I guess). I plotted the locations of the wrecked German vehicles on the Moratin map and discovered something curious. For all their evident care, the mass of wreckage on the so-called 'Highway 3' (that was actually a name given to the road by the GIs, based upon its traffic class as given on their maps) alluded to in Jeff's post was not there! Instead of the 24-odd tanks, AFVs, and motor vehicles commonenly described by many participants, the investigative teams found two assault guns and one SPW.

Now, the Germans did not haul them away -- that can be confirmed since the 823rd gun crews had the site under observation and fire the whole time until the Germans withdrew. So where did they go? I can't answer that for sure, but I think I can make some fair assumptions. The road in question parralleled the main roaid from Sourdeval to Mortain and then on to Barentan, essentially the main axes of advance for the right wing of First Army. As such, it would have been critical to clear that roadway for US use as soon as possible after the battle. However, the roadway was hemmed in by hedgerows on one side and a drop to the river Seinne on the other (which is why it proved to be such a 'shooting gallery' for the 823rd). It appears likely that US engineers would have been forced to move the bulk of the destroyed vehicles to make the road passable (it is unlikely that they would have bulldozed the wreckage into the river, which was narrow and swift due to the heavy rainfall of that summer -- it could have created a dam threatening the security of the bridge just upstream at L'Abbayye Blanche). Now just about 1.5 kilometers to the north and about one kilometer to the east are two other wreckage sites that may have been used as dumping grounds by the engineers. In any case I'm looking at the wreckage descriptions there to see if my theory may make sense.

By the way, the other major road block sote, further north on 'Highway 2' at St. Barthelemy, had a mass of wrecked German vehciles clustered to the northwest and north of the American positions -- most of them off the roadway. Again, however it appears that many of the lost German vehicles described as lost between the two roadblocks are missing. However, in this case the position was eventually overrun, making it possible that the Germans would have pulled out damaged but repairable vehicles (many of the wreck sites around the roadblocks had vehicles the ORS teams described as being repaired, but abandoned) but if that was the case, why didn't they evacuate some of the mass of vehicles to the north of St. Barthelemy?

quote:

---

Niklas Said: As to British tank losses in Normandy, I have seen a couple of PRO documents on this, but they are quite contradictory.

---

WO291/1218 and WO295/961?

TDI report “The Effects of Combat Losses and Fatigue on Operational Performance” indicates that during Goodwood Commenwealth losses in tanks was in excess of 12.9% per day. Presumably the Goodwood loss rate represents one of the higher ended tank loss rates for a specific operation.

quote:

---

Thanks for the report. I have read it and my suspicion that the ORS teams went to places were air units hade made claims seems justified. Depending on the question, this can mean that the sample is biased. If the question is the percentage of the German tanks lost that were victims to air power, such a sample will show air power better than it actually was. On the other hand, if the question is the ratio between air claims and actual kills, the sample need not be biased.

---

quote:

---

Originally posted by Jeff Duquette:
Thanks for the erudite response. I am looking forward to reading your commentary on my other questions.

---

Thanks for the complement. It will still be a couple of days before I answer the other two parts of your question.

quote:

---

To switch gears somewhat; What concerns me (or more properly interests me) regarding ORS studies are the very small samples of air inflicted AFV casualties.

---

I agree in principal with your comments on this, not just for these studies but for all OR studies. I have actually heard senior US OR people state that what is needed for study is a few well-researched examples. This, I believe, is entirely incorrect. The study of warfare is not an engineering or technical study, but is a "quantitative social science" study. As such, one needs large robust data bases with plenty of examples from different environments, so as to be able to ferret out the truth among all the noise, variables and cross-correlations. You will note, from our data base section of the web site, we are the only ones developing such data bases.

quote:

---

I am looking through the various TDI reports listed on....and I can't find your briefing.

---

We do not list our briefings among our reports, except where the breifing was the final deliverable. For most projects, we produce a briefing at the start of the project, outlining what we are going to do, we often provide an interim breifing, showing status, and then we produce a final breifing that sumarizes the final report. So most projects have two or three breifings.

I also have provided a number of independent breifings, not tied to any project, as part of various conferences. Also, I gave a tour of the battlefield in 1996, and so prepared a series of breifings and slide shows on Kursk for that (Battle Briefing, Casualties, Armor, Air, Terrain walk, etc.). The Kursk Air Campaign briefing was created at that time from work we were doing on the Kursk project and the Dupuy Air Campaign Model (DACM). I revised it and last presented it May 20, 1999 at the "Thunder User Group". Thunder is the Air Force's favorite air combat model. There are also other briefings I have done for other conferences and meeting ("Estimating Casualties in Contingency Operations", "Measuring Human Factors in Combat", etc.).

Most of this data, in various forms, is scattered in most of our reports. In the case of Kursk, there is material in my briefings that is not in any of our reports. Much of these material (including a lot of the graphics) will appear in my book, to the extent allowed by the publisher.

So, the question is....should we make our briefings available for purchase?

Perhaps.....I had not really thought about it. They often have color slides in them (read $$$) and we do not have them assembled in one particular file nor have we made a list of them. Furthermore, the briefings were not really designed to stand alone without the briefer.

Added to that, we also have hundreds of briefing slides of Trevor Dupuy, that he used to make up 50+ briefings on various subjects.