Reliability of Sovjet military deaths

[Piotr Kapuscinski]

Thanks for your explanation Art. I have one more question:

Could someone explain why isn't Krivosheev listing Soviet losses in the Polish-Bolshevik War of 1919 - 1921 or even specifying this military conflict as a separate chapter of his book about Soviet military losses in 20th century?

Where are Soviet losses in that military conflict included? Are they even included anywhere in his book?

[Art]

Losses in the Soviet-Polish War are included in the overall Civil War losses.

[Piotr Kapuscinski]

There were two Soviet fronts involved in the Polish-Soviet war of 1919 - 1921 - Western and South-Western.

So according to Krivosheev (table 8) Soviet losses in the 1920 campaign of the Polish-Soviet war were:

Western Front - 72,391 irrecoverable losses, 72,032 sanitary losses
South-Western Front - 57,677 irrecoverable losses, 29,887 sanitary losses

Correct? But according to Krivosheev these figures are incomplete as he writes in the note below table 8.

He also writes that - besides of this general incompleteness (because of which this data doesn't match analogous information about the irrecoverable and medical losses of the Armed Forces for 1920 from CSASA, file 7, cat. 6, d. 802, p. 86, 86a and 99) - there are also no losses for January of 1920 listed. And these losses in January were:

Western Front - 1268 irrecoverable, 20,724 sanitary
South-Western Front - 1068 irrecoverable, 38,790 sanitary

He didn't list them in the table as he didn't have data about those losses for all armies.

I think that this data from CSASA, file 7, cat. 6, d. 802, p. 86, 86a and 99 is in table 9 - am I right? The difference between what is given in the table 8 and what is given in the table 9 is very considerable so losses of Western and South-Western Fronts were in fact much greater (table 8 is very incomplete). The best way to calculate real losses of these fronts would be to calculate what percentage of losses from table 8 they constitute and then take similar percentage of total losses listed in table 9 - the result would be close to real losses suffered by these fronts.

Aditionally we shoul add to those numbers given above losses of Western Front in 1921:

14,602 irrecoverable, 26,372 sanitary (table 15 and footnote 2)

And we still don't have Soviet losses against Poland during the 1919 campaign of that war.

Then Krivosheev writes something about POWs and interned in Germany (and there are like dozens of thousands of those who returned from Polish captivity + dozens of thousands of those who returned from Germany - so they weren't really irrecoverable loses) and I'm not sure whether he counted these losses as irrecoverable in tables 8 & 9 or not. There are also over 20,000 desertions in 1920 listed (much more than in other years) - certainly many if not most of them deserted on the Polish-Soviet front.

So it is quite hard to estimate Soviet losses in that war basing on Krivosheev. A lot of information to anaylze.

It would have been easier if he had written a separate chapter on that war...

[Art]

Continuing to discuss reliability of existing sources on Soviet military losses: I was interested in the question so when primary documents were available I made a sample study of losses of the Leningrad Front in 1944. It is presented in the table below:

Numbers are taken from TsAMO f.217, op.1244, d.643 (Leningrad Front HQ, "Summary reports on losses submitted to the General Staff"). Data for all months are based on monthly reports except April, where they are calculated as a sum of ten-days reports, and December, where there are calculated from ten-days reports again for the period 1-20 December 1944. The sum for the year is calculated by me. Numbers in the "Krivosheev" row are taken from the book "Soviet Casualties and Combat Losses in the twentieth century". In general it looks like there is a good agreement. It must be said that in some cases the summation of ten-days and monthly reports gives very different results, that is when large-scale shuffling of forces between different fronts occurred, for example in February 1944 (the most part of the former Volkhov front transferred under the Leningrad Front's command) or April (3rd Baltic Front is formed by splitting the LF). It's not clear how exactly data in the official publication were calculated, summation of monthly reports gives the sum which is considerably small, I don't have all the ten-days reports for the whole year, probably they would give a closer fit to Krivosheev.

[Vaeltaja]

How does that reflect or correspond with the several known errors in Krivosheyev's data regarding the Soviet offensive against the Finns in June - August 1944? Were the losses set under 'wrong' header? As we know there are errors in the data, and we now see that sums of the casualties still came up roughly equal.

EDIT: http://forum.axishistory.com/viewtopic. ... a#p1502064
- Number of formations seems too low (ie. other sources (victory.mil.ru, July 1 1944) state number to be considerably higher)
- Timeframe of the offensive is off (though in accordance with 'traditional' Soviet histography)
- (slightly off topic) As number of formations is wrong and the also the timeframe is off wonder if the manpower numbers also contain errors?

[Art]

How does that reflect or correspond with the several known errors in Krivosheyev's data regarding the Soviet offensive against the Finns in June - August 1944?

In no way. Annual losses and losses in operation are two different sets of data. There are certain questions regarding Krivosheev's numbers for operations, in particular I don't like the numbers for the Pskov offensive operation. What should be repeated here is there were two different types of reports on the level of fronts: monthly and ten-days report (also there were numbers reported on a daily or five-days basis, but there were not complete). Because of units shuffling they actually give different sum, so the question "what were losses of X during the period Y according to documents?" is not as self-obvious as it may seem. My point is that it is not clear what was the algorithm used by Krivosheev and team. They could use summation of ten-days reports or they could sum up ten-days report. Theoretically speaking that could affect distribution of losses between different fronts but the sum for all fronts must be roughly the same. Also they could use a combination of both methods. In this case it is rather easy to make mistakes and either miss a part or duplicate it. It is hard to get a definitive answer without knowing all that details.

Were the losses set under 'wrong' header?

What is that exactly?
I do remember about that thing with Vyborg operation, I would like to check some thing before posting on it.

[Vaeltaja]

What is that exactly?
I do remember about that thing with Vyborg operation, I would like to check some thing before posting on it.

Just wondered since the total number of losses were in the same ballpark that could some of losses omitted from the entry of the Leningrad Front from Krivosheyev's data from Soviet offensive against Finns be placed under some roughly simultaneous operation undertaken by the Leningrad Front in his data. That is still present in the data just 'mislabeled'.

[Art]

Krivosheev's losses in operations are based on the part of the data, while losses by years and quarters are calculated by summation of all of it. So the second is always more representative than the first.

[tramonte]

I guess it was Manninen who using Soviet sources got loss figures of Leningrad Front in Karelian Isthmus 9 June - 19 July 1944: at least 105 000 while Dr. Tapio Tiihonen found serious conflicts in Soviet loss reports and estimated losses been 189 000. Even Manninen himself saw 105 000 clearly too low.

When it comes to Soviet Offensives of 7th and 32th Army in Karelia (21 June - 11 August) the loss figure of 63 600 has not been criticized much by Finnish historians. It's worth to compare Finnish loss rate between Isthmus and Karelia: around 68% in Karelian Isthmus versus 31% in Karelia while around 1% in Rukajärvi-Uhtua, which might hint that Soviet losses in Karelian Isthmus might have been around 140 000 - 145 000 meaning the truth might have be between those official Soviet loss figures and those of Tiihonen.

[Art]

I guess it was Manninen who using Soviet sources got loss figures of Leningrad Front in Karelian Isthmus 9 June - 19 July 1944: at least 105 000

Yes, that was something close to reality.

[tramonte]

Continuing to discuss reliability of existing sources on Soviet military losses: I was interested in the question so when primary documents were available I made a sample study of losses of the Leningrad Front in 1944. It is presented in the table below:

Numbers are taken from TsAMO f.217, op.1244, d.643 (Leningrad Front HQ, "Summary reports on losses submitted to the General Staff"). Data for all months are based on monthly reports except April, where they are calculated as a sum of ten-days reports, and December, where there are calculated from ten-days reports again for the period 1-20 December 1944. The sum for the year is calculated by me. Numbers in the "Krivosheev" row are taken from the book "Soviet Casualties and Combat Losses in the twentieth century". In general it looks like there is a good agreement.

I mostly agree when focusing for instance 3 month period from June to August 1944. 9 June to 20 July the bulk of losses came from Karelian Isthmus while after that 25 July to 10 August are losses came from Tannenberg Line battle in Estonia. Those official figures of Leningrad Front are debunking part of claims of Dr. Tiihonen (Leningrad Front losses 189,000 in Karelian Isthmus) and much much more of those fantastic claims of Mart Laar (Leningrad Front losses of 170,000 during Tanneberg line battle).

When it comes to battles in Karelian Isthmus ( 9 June - 20 July 44) the most realistic losses of Leningrad Front seems to have been around 130,000 - 140,000 (including sick). Actually it has been my long period guess for a rather long time. I have no idea how heavy losses Leningrad Front took during period of 25 July-31 July in Estonia but those relatively moderate loss figures of August (around 8,500 killed, MIA, WIA) are not suggesting something very high losses during the last week of July.

[tramonte]

I guess it was Manninen who using Soviet sources got loss figures of Leningrad Front in Karelian Isthmus 9 June - 19 July 1944: at least 105 000

Yes, that was something close to reality.

total losses of Leningrad Front June-July: 160,059 of which 138,787 where killed, MIA or WIA. Now the question here is: how many of them not in Karelian Isthmus. According report of 10 - 20 June 44 less than 5% of killed, MIA and WIA were those of 8th and 2nd Shock Army (total 1 433). If other 10 days periods have around same killed/MIA/WIA numbers for 8th and 2nd Shock Army then their losses in June-July have been around 8,600 and bulk of losses, some 130,200 have been those of Leningrad Front in Karelian Isthmus (21st, 23rd, 59th Army + front + 13th Air Army+ navy). Plus those of sick and other cases (some 7,000 - 8,000?).

However, i have no idea how high were combat losses of 8th and 2nd Shock Army during first stage of Tannenberg line battle (25 - 31 July 44). These ("extra") numbers will of course reduce share of Karelian Isthmus.

[Art]

Battle losses of the 2 Shock and 8 Armies in June 1944 were about 3000 men and in July 1944 - some 25,000. That would give you a rough idea.

[tramonte]

Battle losses of the 2 Shock and 8 Armies in June 1944 were about 3000 men and in July 1944 - some 25,000. That would give you a rough idea.

So just 8th and 2nd Shock army took 25,000 battle losses in July 44 but whole Leningrad Front in August:
KIA..............1,702 KIA
WIA.............6,536 WIA
MIA................196 MIA
combat loss...8,432

Most of studies are timing Battle of Tannenberg line 25 July - 10 August but these loss figues are suggesting that daily combat losses have been 4 times higher in late July compared to early August. Really???
The Battle of Vuosalmi (Eastern Karelian Isthmus) was literally over in 18th July.

[Reigo2]

If compared to the casualty reports at obd-memorial.ru then one can conclude that Leningrad Front's casualty summaries for KIA/MIA are not complete. The reason was probably nothing sinister, rather usual problems related to reporting (mainly part of the data not arriving on time etc).

Sum of KIA/MIA losses of 2nd Shock Army's and 8th Army's rifle divisions, fortified regions, armored units and penal units 24 July - 10 August day by day according to reports at obd-memorial.ru:

24.07 - 1030
25.07 - 1210
26.07 - 557
27.07 - 577
28.07 - 321
29.07 - 537
30.07 - 537
31.07 - 353
01.08 - 401
02.08 - 179
03.08 - 725
04.08 - 498
05.08 - 321
06.08 - 210 [the last day of the Soviet offensive against the Tannenberg line - 10th August is not the correct date]
07.08 - 85
08.08 - 72
09.08 - 48
10.08 - 47
Sum: 7708 KIA/MIA

To this about few hundred KIA/MIA can be added mainly from artillery and engineers units directly under corps or army subordination.