Statistics on German PoW hauls in latter October 1941?

Discussions on WW2 in Eastern Europe.
Ружичасти Слон
Member
Posts: 488
Joined: 24 Jan 2020, 17:31
Location: Изгубљени

Re: Statistics on German PoW hauls in latter October 1941?

#61

Post by Ружичасти Слон » 16 Aug 2020, 15:40

TheMarcksPlan wrote:
16 Aug 2020, 03:43

There are surely some data issues caused by delays in reporting and time-overlaps for unit/army subordination. Which is why it's IMO essential to look at the broad picture and not get too tied up in granular analysis.
Tmp not like granular analysis when datas suggest tmp imagination storys are not good.

TheMarcksPlan wrote:
16 Aug 2020, 03:43
To repeat, AGC tactically captured >1% of its enemies post Taifun and < 0.3% pre-Taifun. What's your argument that there's no significance to that discrepancy, or that it traces to something other than morale? [Aside from the Krivosheev reasoning chain and excluding the Toropets PoWs from tactical credit]
Germany army was collect much datas on Soviet Red army surrenders. You can to find much datas on documents in BAMA archives and TsAMO archives.

But there was be BIG problem with Germany army datas on 1941.year. They was not count correct by 500.000 mens and on december 1941.year was make total 500.000 down. They was not explain where was be error or when was be error. So impossible for to know what datas on Red army surrenders was be correct.

For to make imagination story on big numbers surrender must to mean low morale is anti-intellectual when not have no historical datas or evidences on why mens surrender.

For to make imagination story on big numbers surrender must to mean low morale when compare datas on different time and place but datas have big errors of 500.000 on not know time and not know place is BIG analytics mistake.

User avatar
TheMarcksPlan
Banned
Posts: 3255
Joined: 15 Jan 2019, 23:32
Location: USA

Re: Statistics on German PoW hauls in latter October 1941?

#62

Post by TheMarcksPlan » 17 Aug 2020, 05:07

Max Payload wrote:It seems to me that there is insufficient hard data to justify your assertion that Red Army morale declined markedly after mid-October. There was never even any true consistency on the post-Viazma Moscow axis on any particular day, some units fighting heroically, others performing abysmally.
The presence or magnitude of variance does not make out a valid critique of a macro trend.

For instance the macro temperature trend is for the second week of October to be cooler than the first (northern hemisphere) but there's immense variance between daily temperatures within each week and in many years October's second week is warmer than the first.

The macro causal factor in that example - declining insolation - can be swamped by variations in fronts, jet streams, etc.

The macro factor here is morale, specifically its component that is sensitive to perceptions of the war's course (i.e. soldiers who think "we're gonna lose" have generally show lower morale). But within the domain of morale there are many other factors such as unit cohesion, training, leadership, etc. As the Red Army's forces around Moscow varied from untrained militia to seasoned professionals shipped from the East, as their leaders varied from excellent to party hacks and over-promoted NCO's, it's entirely predictable that there would be immense morale variation around a moment's global mean.

But just as October 14 is generally cooler than October 7 on the top-line historical stats, so too will the macro trend of declining morale show in the top-line stats. If those top-line stats show a macro-trend (as I think they do) then variance is just a distraction.
Secondly -
You quoted Lopukhovsky, let me expand on the quote (p389)
“Thus approximately 200,000 men either avoided encirclement entirely, or had managed to make their way back to friendly forces. However, separate groups of military personnel continued to trickle back into friendly lines in November and even later.”
How many of those attempting to ‘make their way back to friendly forces’ in late October and early November were captured by Fourth Army, thereby inflating its prisoner haul above that of prisoners taken on the frontline?
That's a possible explanation but the same could be said of August/September when Soviets were streaming back from Minsk, Smolensk, and smaller encirclements. For example: https://en.wikipedia.org/wiki/Ivan_Boldin

One possible test of this theory is to look at PoW's caught by Army Group-level units such as security divisions and other rear area units. It seems all but certain that these units would capture proportionately more stragglers than front-line army units. AGC-level units captured 38k PoW up to Sep27 and 13k in the month after Vyazma, a similar daily rate. https://wwii.germandocsinrussia.org/ru/ ... ect/zoom/6 This seems good evidence against there having been a significant increase in stragglers, post-Taifun.

And there's good reason to expect proportionately fewer stragglers from Taifun than from Minsk/Smolensk: It was a much shallower operation that left far less "empty space" into which bypassed units could disappear.
Max Payload wrote:Firstly -
If there is a change in force ratio to the advantage of Fourth Army it might be expected that Red Army losses would increase proportionally. (Fourth Army had Fourth Panzer Group subordinated to it in Oct/Nov increasing its force ratio advantage still further). Whether that proportional increase in Red Army losses would be greater or less than total numerical losses at an earlier stage would depend on a range of circumstances.
The TDI study found no correlation between force ratios and % of enemy force captured. http://www.dupuyinstitute.org/pdf/e-4epw1and2final.pdf at 63.
Max Payload wrote:The question is, do you accept Krivosheev’s figure for NWF irrecoverable losses over the first 18 days of the war? If so, pick your own KIA figure -
20,000?
25,000?
30,000?
35,000?
I have very little confidence in Krivosheev for 1941 but because I believe he understated casualties in this period that works in your favor.

Let's try another means of getting a ballpark figure for KIA: attrition ratios compared to German losses. I don't have daily data for the opposing German forces (AGN and PzGr3) so for now let's estimate that as well.

Depending on which source you credit (OKW diaries or Overmans), the Ostheer had between 1,800 and 2,400 daily dead in June/July 1941. https://en.wikipedia.org/wiki/German_ca ... rld_War_II (reproducing tables from said sources).

AGN and PzGr3 had ~1/3 of Ostheer's divisions, so their average daily dead would be around 600-800.

If we assume a typical German-Soviet WW2 combat efficiency ratio of ~5 and rough numeric parity between the forces then we'd expect NWF to average 3-4k daily dead. Over 18 days 54-72k dead. Which would leave room for 2-20k MIA, most of which would have become PoW.

Take the high range of that estimate and say NWF had 20k MIA. That's entirely consistent with AGN's PoW report, as some of NWF's PoW went to AGC and as AGN had the most favorable battle outcomes in the first 18 days of war and therefore could easily have taken most of its prisoners in this period.

In fact it's feasible that (1) NWF lost more men than Krivosheev says and (2) relatively few of them were PoW.
https://twitter.com/themarcksplan
https://www.reddit.com/r/AxisHistoryForum/
https://medium.com/counterfactualww2
"The whole question of whether we win or lose the war depends on the Russians." - FDR, June 1942


User avatar
TheMarcksPlan
Banned
Posts: 3255
Joined: 15 Jan 2019, 23:32
Location: USA

Re: Statistics on German PoW hauls in latter October 1941?

#63

Post by TheMarcksPlan » 17 Aug 2020, 06:35

Max Payload wrote:Krivosheev quotes for Northwestern Front in 1941 a total of 182,264 irrecoverable losses of which 142,190 were Missing in action, POWs - 78%.
Krivosheev also quotes 73,924 irrecoverable losses for Northwestern Front in the first 18 days of the war alone.
Are you sure you're reading Krivosheev - or whichever author cited him - correctly?

From Liedtke's Enduring the Whirlwind I have the following from FN29 of Chapter 3:
By 7 July, the Northwestern Front had sustained 90,000 casualties, and lost 1,000 tanks and 4,000 guns and mortars. Glantz, The Battle for Leningrad, p.32 and 36
I don't have Glantz's better Leningrad book, only have Siege of Leningrad (just ordered Battle though). But Glantz usually cites Krivosheev so I'm guessing he does so here as well.

Plus it seems implausible that NWF suffered 42% of its casualties in the first 18 days of 190 days of fighting.

Are you sure you're not citing a figure for total casualties through July 8th, rather than "irrevocable losses?"
https://twitter.com/themarcksplan
https://www.reddit.com/r/AxisHistoryForum/
https://medium.com/counterfactualww2
"The whole question of whether we win or lose the war depends on the Russians." - FDR, June 1942

Richard Anderson
Member
Posts: 6347
Joined: 01 Jan 2016, 22:21
Location: Bremerton, Washington

Re: Statistics on German PoW hauls in latter October 1941?

#64

Post by Richard Anderson » 17 Aug 2020, 07:33

Losses, HG-N, 22 June-6 July 1941/6 July-3 August 1941

AOK 16 - 3,907/12,927
AOK 18 - 5,610/10,352
PzGr 4 - 3,279/15,290
Richard C. Anderson Jr.

American Thunder: U.S. Army Tank Design, Development, and Doctrine in World War II
Cracking Hitler's Atlantic Wall
Hitler's Last Gamble
Artillery Hell

Max Payload
Member
Posts: 574
Joined: 21 Jun 2008, 15:37

Re: Statistics on German PoW hauls in latter October 1941?

#65

Post by Max Payload » 17 Aug 2020, 09:48

TheMarcksPlan wrote:
17 Aug 2020, 06:35
Max Payload wrote:Krivosheev quotes for Northwestern Front in 1941 a total of 182,264 irrecoverable losses of which 142,190 were Missing in action, POWs - 78%.
Krivosheev also quotes 73,924 irrecoverable losses for Northwestern Front in the first 18 days of the war alone.
Are you sure you're reading Krivosheev - or whichever author cited him - correctly?

From Liedtke's Enduring the Whirlwind I have the following from FN29 of Chapter 3:
By 7 July, the Northwestern Front had sustained 90,000 casualties, and lost 1,000 tanks and 4,000 guns and mortars. Glantz, The Battle for Leningrad, p.32 and 36
I don't have Glantz's better Leningrad book, only have Siege of Leningrad (just ordered Battle though). But Glantz usually cites Krivosheev so I'm guessing he does so here as well.

Plus it seems implausible that NWF suffered 42% of its casualties in the first 18 days of 190 days of fighting.

Are you sure you're not citing a figure for total casualties through July 8th, rather than "irrevocable losses?"
1997 Greenhill/Stackpole edition page 110/111
“The Defensive Operation in Lithuania and Latvia 22 June - 9 July 1941
N.W. Front (whole period)
Losses during the operation
Irrecoverable losses 73,924
Sick and wounded 13,284
Total 87,208
Average daily losses 4,845”

On page 166 the Front’s losses for 1941 are given as 270,087, of which 182,264 are irrecoverable (31,511 KIA or died during evac, 142,190 MIA - POWs, 8,563 non-combat) and 87,823 sick and wounded.

During the first 2-3 weeks AGN was advancing at around 30km/day. Subsequently it’s advance towards Leningrad and the Lovat averaged around one-tenth of that. This is the probable reason that sick and wounded account for a relatively low proportion of total losses in the first 2-3 weeks (one-sixth) compared to the year as a whole (one-third), a greater majority of the total losses during that initial phase being POWs.

Max Payload
Member
Posts: 574
Joined: 21 Jun 2008, 15:37

Re: Statistics on German PoW hauls in latter October 1941?

#66

Post by Max Payload » 17 Aug 2020, 12:41

TheMarcksPlan wrote:
17 Aug 2020, 05:07
The presence or magnitude of variance does not make out a valid critique of a macro trend.
But 'statistical noise' is one of the factors that makes identifying a statistically meaningful macro trend problematic.
TheMarcksPlan wrote:
17 Aug 2020, 05:07
Secondly -
You quoted Lopukhovsky, let me expand on the quote (p389)
“Thus approximately 200,000 men either avoided encirclement entirely, or had managed to make their way back to friendly forces. However, separate groups of military personnel continued to trickle back into friendly lines in November and even later.”
How many of those attempting to ‘make their way back to friendly forces’ in late October and early November were captured by Fourth Army, thereby inflating its prisoner haul above that of prisoners taken on the frontline?
That's a possible explanation but the same could be said of August/September when Soviets were streaming back from Minsk, Smolensk, and smaller encirclements. For example: https://en.wikipedia.org/wiki/Ivan_Boldin

One possible test of this theory is to look at PoW's caught by Army Group-level units such as security divisions and other rear area units. It seems all but certain that these units would capture proportionately more stragglers than front-line army units. AGC-level units captured 38k PoW up to Sep27 and 13k in the month after Vyazma, a similar daily rate. https://wwii.germandocsinrussia.org/ru/ ... ect/zoom/6 This seems good evidence against there having been a significant increase in stragglers, post-Taifun.
PoW's caught by Army Group-level units may only show that they were achieving similar outcomes with what were probably similar resources. Meanwhile, at Smolensk the escapees in late August were escaping into both the Fourth and Ninth Armies zones whereas post-Viazma in late October they were escaping towards Fourth Army's lines only.
TheMarcksPlan wrote:
17 Aug 2020, 05:07
And there's good reason to expect proportionately fewer stragglers from Taifun than from Minsk/Smolensk: It was a much shallower operation that left far less "empty space" into which bypassed units could disappear.
I don't understand your reasoning here.

TheMarcksPlan wrote:
17 Aug 2020, 05:07
The TDI study found no correlation between force ratios and % of enemy force captured. http://www.dupuyinstitute.org/pdf/e-4epw1and2final.pdf at 63.
No demonstrable correlation - to which they added the rider, 'although this was expected' 'because combat is simply too complex for such a simplistic formulation'.
They are not saying that if force A doubled in size it would have no impact on the prisoner take from Force B.
TheMarcksPlan wrote:
17 Aug 2020, 05:07
I have very little confidence in Krivosheev for 1941 but because I believe he understated casualties in this period that works in your favor.

Let's try another means of getting a ballpark figure for KIA: attrition ratios compared to German losses. I don't have daily data for the opposing German forces (AGN and PzGr3) so for now let's estimate that as well.

Depending on which source you credit (OKW diaries or Overmans), the Ostheer had between 1,800 and 2,400 daily dead in June/July 1941. https://en.wikipedia.org/wiki/German_ca ... rld_War_II (reproducing tables from said sources).

AGN and PzGr3 had ~1/3 of Ostheer's divisions, so their average daily dead would be around 600-800.

If we assume a typical German-Soviet WW2 combat efficiency ratio of ~5 and rough numeric parity between the forces then we'd expect NWF to average 3-4k daily dead. Over 18 days 54-72k dead. Which would leave room for 2-20k MIA, most of which would have become PoW.

Take the high range of that estimate and say NWF had 20k MIA. That's entirely consistent with AGN's PoW report, as some of NWF's PoW went to AGC and as AGN had the most favorable battle outcomes in the first 18 days of war and therefore could easily have taken most of its prisoners in this period.

In fact it's feasible that (1) NWF lost more men than Krivosheev says and (2) relatively few of them were PoW.
Krivosheev has his critics for good reason, but I didn't quote him because his figures work in my favour but because they are the most detailed available. Your calculations are so full of caveats as to render them of little value. What you are trying to do is justify Glantz's figure. What is the actual source of the 35,000?

Ружичасти Слон
Member
Posts: 488
Joined: 24 Jan 2020, 17:31
Location: Изгубљени

Re: Statistics on German PoW hauls in latter October 1941?

#67

Post by Ружичасти Слон » 17 Aug 2020, 13:34

TheMarcksPlan wrote:
17 Aug 2020, 05:07

The presence or magnitude of variance does not make out a valid critique of a macro trend.

For instance the macro temperature trend is for the second week of October to be cooler than the first (northern hemisphere) but there's immense variance between daily temperatures within each week and in many years October's second week is warmer than the first.

The macro causal factor in that example - declining insolation - can be swamped by variations in fronts, jet streams, etc.
Ok.
TheMarcksPlan wrote:
17 Aug 2020, 05:07


The macro factor here is morale, specifically its component that is sensitive to perceptions of the war's course (i.e. soldiers who think "we're gonna lose" have generally show lower morale). But within the domain of morale there are many other factors such as unit cohesion, training, leadership, etc. As the Red Army's forces around Moscow varied from untrained militia to seasoned professionals shipped from the East, as their leaders varied from excellent to party hacks and over-promoted NCO's, it's entirely predictable that there would be immense morale variation around a moment's global mean.

But just as October 14 is generally cooler than October 7 on the top-line historical stats, so too will the macro trend of declining morale show in the top-line stats. If those top-line stats show a macro-trend (as I think they do) then variance is just a distraction.
Another time tmp try to mislead mens with lots of words and anti-intellectual thoughts and analysis on tmp imagination story.

1.
Meteorology mens have much data on much years. Meteorology mens can to see macro trend only after studys on micro datas on many years. Can to make macro trend on study many years.

Tmp have little data on one year. Tmp was invent macro trend because he was not have much micro datas and only on one year. How to make macro trend when not have much datas on only one year ?

2.
Meteorology mens have much data on much years. Meteorology mens can to see macro trend only after studys on micro datas on many years. Trend was decline.

Tmp have little data on one year. Tmp was invent macro trend because he was not have much micro datas and only on one year. Trend was decline. But tmp was invent tactical surrenders idea for to make theory on trend was increase.

3.
Meteorology men have much data on much years. Meteorology mens can to see micro causal factor because have very much micro datas on many years on insolation. Meteorology mens have mush micro datas on insolation on many years.

Tmp have little data on one year. Tmp was invent macro causal factor because he was not have no datas on micro or macro causal factor. Tmp have no micro datas or evidences why mens was surrender.


It is only tmp invention that tmp theory is same as meteorology anaysis and explain.
It is not same because tmp was write it is same.
It is tmp on anti-intellectual thoughts and analysis for to mislead mens.


TheMarcksPlan wrote:
17 Aug 2020, 05:07

The macro factor here is morale, specifically its component that is sensitive to perceptions of the war's course (i.e. soldiers who think "we're gonna lose" have generally show lower morale).
4.
Meteorology mens was start on micro datas and was make conclusions and explains on micro datas.

Tmp was start on tmp imaginary theory he waa want to be true and was invent ideas and was invent explains.

Ружичасти Слон
Member
Posts: 488
Joined: 24 Jan 2020, 17:31
Location: Изгубљени

Re: Statistics on German PoW hauls in latter October 1941?

#68

Post by Ружичасти Слон » 17 Aug 2020, 22:18

Max Payload wrote:
17 Aug 2020, 12:41

TheMarcksPlan wrote:
17 Aug 2020, 05:07
I have very little confidence in Krivosheev for 1941 but because I believe he understated casualties in this period that works in your favor.

Let's try another means of getting a ballpark figure for KIA: attrition ratios compared to German losses. I don't have daily data for the opposing German forces (AGN and PzGr3) so for now let's estimate that as well.

Depending on which source you credit (OKW diaries or Overmans), the Ostheer had between 1,800 and 2,400 daily dead in June/July 1941. https://en.wikipedia.org/wiki/German_ca ... rld_War_II (reproducing tables from said sources).

AGN and PzGr3 had ~1/3 of Ostheer's divisions, so their average daily dead would be around 600-800.

If we assume a typical German-Soviet WW2 combat efficiency ratio of ~5 and rough numeric parity between the forces then we'd expect NWF to average 3-4k daily dead. Over 18 days 54-72k dead. Which would leave room for 2-20k MIA, most of which would have become PoW.

Take the high range of that estimate and say NWF had 20k MIA. That's entirely consistent with AGN's PoW report, as some of NWF's PoW went to AGC and as AGN had the most favorable battle outcomes in the first 18 days of war and therefore could easily have taken most of its prisoners in this period.

In fact it's feasible that (1) NWF lost more men than Krivosheev says and (2) relatively few of them were PoW.
Krivosheev has his critics for good reason, but I didn't quote him because his figures work in my favour but because they are the most detailed available. Your calculations are so full of caveats as to render them of little value. What you are trying to do is justify Glantz's figure. What is the actual source of the 35,000?
You can to find different datas on different place on same topic. How to know what datas is correct ?

Kinzel was report on OKH Datas on Red army captures.
22.6.41 - 31.7.41 HGN = 56.320
22.6.41 - 31.7.41 HGM = 580.910
22.6.41 - 31.7.41 HGS = 162.685

User avatar
TheMarcksPlan
Banned
Posts: 3255
Joined: 15 Jan 2019, 23:32
Location: USA

Re: Statistics on German PoW hauls in latter October 1941?

#69

Post by TheMarcksPlan » 17 Aug 2020, 22:25

Max Payload wrote:
17 Aug 2020, 12:41

TheMarcksPlan wrote:
17 Aug 2020, 05:07
And there's good reason to expect proportionately fewer stragglers from Taifun than from Minsk/Smolensk: It was a much shallower operation that left far less "empty space" into which bypassed units could disappear.
I don't understand your reasoning here.
Compare the battlefield area in Minsk/Bialystok and Smolensk vs. Viazma.

In both earlier battles the mobile pincers met up approximately 300km from their starting points and the operating front (distance between outer wings of the PzGr's) was ~250km.

For Viazma the pincers converged only ~100km from their starting points and the encirclement's front was ~200km wide:

Image

...so Viazma's battlefield area was ~1/4 of the earlier big kessels.

As discussed in many sources - I can dig up cites but I'm assuming your're familiar - during AGC's deep encirclements much of the gap area between mobile units and following infantry became chaotic regions in which bypassed Soviet soldiers were able to harass German communications, move/hide fairly effectively, and later melt into the forests/swamps.

Given the much smaller mech-foot gap in Viazma and the smaller area, it seems inevitable that fewer of these anarchic zones would have been created. And indeed the stats on the percentage of forces destroyed seems to support this: Lopukhovsky estimates 80% destruction of Soviet forces in Typhoon (probably a higher percentage for the Viazma sub-battle), whereas Minsk and especially Smolensk did not destroy as high a proportion of its enemies.

Of course I'm not disputing that many Soviets were able to hide out and return - Viazma was still a large battlefield even if smaller than Minsk/Smolensk. But there's little reason to think that Viazma created more stragglers behind 9th/4th armies - and therefore more captured stragglers in the following weeks/months - than did Minsk/Smolensk.

BTW - it seems clear to me that Hitler's preference for smaller/shallower Kessels such as at Viazma was absolutely right.
https://twitter.com/themarcksplan
https://www.reddit.com/r/AxisHistoryForum/
https://medium.com/counterfactualww2
"The whole question of whether we win or lose the war depends on the Russians." - FDR, June 1942

User avatar
TheMarcksPlan
Banned
Posts: 3255
Joined: 15 Jan 2019, 23:32
Location: USA

Re: Statistics on German PoW hauls in latter October 1941?

#70

Post by TheMarcksPlan » 17 Aug 2020, 22:27

Ружичасти Слон wrote:
17 Aug 2020, 22:18

Kinzel was report on OKH Datas on Red army captures.
22.6.41 - 31.7.41 HGN = 56.320
22.6.41 - 31.7.41 HGM = 580.910
22.6.41 - 31.7.41 HGS = 162.685
Not familiar with Kinzel. What's the source?
https://twitter.com/themarcksplan
https://www.reddit.com/r/AxisHistoryForum/
https://medium.com/counterfactualww2
"The whole question of whether we win or lose the war depends on the Russians." - FDR, June 1942

User avatar
TheMarcksPlan
Banned
Posts: 3255
Joined: 15 Jan 2019, 23:32
Location: USA

Re: Statistics on German PoW hauls in latter October 1941?

#71

Post by TheMarcksPlan » 17 Aug 2020, 22:44

Max Payload wrote:No demonstrable correlation - to which they added the rider, 'although this was expected' 'because combat is simply too complex for such a simplistic formulation'.
They are not saying that if force A doubled in size it would have no impact on the prisoner take from Force B.
The rider was in reference to previous "simplistic" research. In this study, TDI took the additional step of attempting to control for outcome by using its outcome-based data sub-sets for correlation analysis.

That analytical move attempts to separate the effect of outcome when analyzing your reasonable intuition that a superior force usually captures more of its enemy. But is that because a superior force wins more often, and winning causes more PoW, or because something inherent to force superiority causes more PoW? TDI didn't find any evidence of the latter.

This discussion returns me to a top-line point about AGC's post-Taifun PoW's: these were taken while AGC advanced slowly. While we don't have TDI-level data on engagements, the outcomes associated with greater PoW - penetration and envelopment - cannot have been dominant for AGC during this period. It was more of a slow slog through mud and then against stiffening numerical odds.

Were we able to control for outcomes at the tactical level, I would expect AGC's post-Taifun PoW performance to appear even more anomalous.
https://twitter.com/themarcksplan
https://www.reddit.com/r/AxisHistoryForum/
https://medium.com/counterfactualww2
"The whole question of whether we win or lose the war depends on the Russians." - FDR, June 1942

User avatar
TheMarcksPlan
Banned
Posts: 3255
Joined: 15 Jan 2019, 23:32
Location: USA

Re: Statistics on German PoW hauls in latter October 1941?

#72

Post by TheMarcksPlan » 17 Aug 2020, 22:56

Max Payload wrote:But 'statistical noise' is one of the factors that makes identifying a statistically meaningful macro trend problematic.
True but just as in weather/climate, the problematics of detecting the signal in the noise is the intellectually valid fun of the analytical chase, regardless of noise level.
Meanwhile, at Smolensk the escapees in late August were escaping into both the Fourth and Ninth Armies zones whereas post-Viazma in late October they were escaping towards Fourth Army's lines only.
True* but IMO this is pulling us back into the noise and away from the signal. There's probably more noise in the variable morale of the units facing each army than there is in the differential rates of straggler capture, to name just one factor. So I'd suggest analytical retrenchment to the Army Group level.

*EDIT - on reflection your contention is true only if stragglers always headed east towards Western Front's lines rather than north towards Kalinin Front. Not sure that's true. Surely many Soviets were smart enough to realize that German troop density would be higher in the east than in the north, and that the AGN-AGC juncture contained more forested land through which escape was easier.
Your calculations are so full of caveats as to render them of little value. What you are trying to do is justify Glantz's figure. What is the actual source of the 35,000?
Glantz cites an AGN communique dated August 6, 1941. I don't have the original document but I'm inclined to trust Glantz's relation of its content.

If [member with Cyrillic user name] is correct, then AGN's communique was an undercount. It may be that AG HQ issued the count before all units sent their PoW figures up the chain. That would be a serious challenge to my thesis.
https://twitter.com/themarcksplan
https://www.reddit.com/r/AxisHistoryForum/
https://medium.com/counterfactualww2
"The whole question of whether we win or lose the war depends on the Russians." - FDR, June 1942

Max Payload
Member
Posts: 574
Joined: 21 Jun 2008, 15:37

Re: Statistics on German PoW hauls in latter October 1941?

#73

Post by Max Payload » 18 Aug 2020, 01:16

TheMarcksPlan wrote:
17 Aug 2020, 22:27
Ружичасти Слон wrote:
17 Aug 2020, 22:18

Kinzel was report on OKH Datas on Red army captures.
22.6.41 - 31.7.41 HGN = 56.320
22.6.41 - 31.7.41 HGM = 580.910
22.6.41 - 31.7.41 HGS = 162.685
Not familiar with Kinzel. What's the source?

In 1941 Lt-Col Kinzel was in command of a branch of the German General Staff responsible for the collection of data on the Red Army. His POW figure, which I hadn’t seen before, can be reconciled with Krivosheev’s NW Front data relatively easily -
Irrevocable losses from 10 July to 31 December
182,264 - 73,924
=108,340 over 144 days= 752/day
For the year as a whole, 78% of the irrevocable losses were MIA.
Let us assume 78% of the irrevocable losses from 10 July to 31 July (that is, after the rapid advance period of AGN) were MIA, that is .78x752x21=12,318
Subtracting that from Kinzel’s figure of 56,320 POWs to the end of July suggests there should have been 44k POWs in the first 18 days (with around 30k KIA).
Taking the calculation to 8 August requires a further .78x752x8=4,692 POWs giving a total of 61k, not 35k.

Max Payload
Member
Posts: 574
Joined: 21 Jun 2008, 15:37

Re: Statistics on German PoW hauls in latter October 1941?

#74

Post by Max Payload » 18 Aug 2020, 01:37

TheMarcksPlan wrote:
17 Aug 2020, 22:25
Lopukhovsky estimates 80% destruction of Soviet forces in Typhoon (probably a higher percentage for the Viazma sub-battle), whereas Minsk and especially Smolensk did not destroy as high a proportion of its enemies.

Of course I'm not disputing that many Soviets were able to hide out and return - Viazma was still a large battlefield even if smaller than Minsk/Smolensk. But there's little reason to think that Viazma created more stragglers behind 9th/4th armies - and therefore more captured stragglers in the following weeks/months - than did Minsk/Smolensk.
I’m not sure of the relevance of Minsk, by August it was 400km behind the frontline.
Differences between Viazma and Smolensk are the numbers involved in the initial envelopments, the fact that for more than a week Rokossovsky was able to hold open an escape route from Smolensk, and, as previously mentioned, escapees from Smolensk after early August were as likely to try to filter through Ninth Army’s lines as those of Fourth Army.

Max Payload
Member
Posts: 574
Joined: 21 Jun 2008, 15:37

Re: Statistics on German PoW hauls in latter October 1941?

#75

Post by Max Payload » 18 Aug 2020, 01:57

TheMarcksPlan wrote:
17 Aug 2020, 22:44
Max Payload wrote:No demonstrable correlation - to which they added the rider, 'although this was expected' 'because combat is simply too complex for such a simplistic formulation'.
They are not saying that if force A doubled in size it would have no impact on the prisoner take from Force B.
The rider was in reference to previous "simplistic" research. In this study, TDI took the additional step of attempting to control for outcome by using its outcome-based data sub-sets for correlation analysis.

That analytical move attempts to separate the effect of outcome when analyzing your reasonable intuition that a superior force usually captures more of its enemy. But is that because a superior force wins more often, and winning causes more PoW, or because something inherent to force superiority causes more PoW? TDI didn't find any evidence of the latter.
Inherent force superiority won’t cause more POWs unless that force superiority is put to effective use. If it is put to effective use it doesn’t take a TDI study to conclude that enemy losses, including POW numbers, will be greater than if force superiority is absent.

Post Reply

Return to “WW2 in Eastern Europe”