Mystics & Statistics

What we did was a simple comparison of the Soviet General Staff study data on the air fighting in the south compared to the daily records we gathered from the Second and Seventeenth Air Armies. What we found was their were minor differences in the sortie counts, but overall that was close to what was reported in the unit records we had.

On the other hand, the reports on casualties was not. There were outrageously incorrect estimations of enemy losses, which is typical of Soviet accounts. But as significant, the reports of their own losses were low. In particular, our count of Second Air Army losses from 5-18 July was 481, their count was 371. This Soviet General Staff study only reported 77% of their losses. Does this mean that if I draw losses reports from the Sixteenth Air Army from the Soviet General Staff study (as I don’t have the unit records), should I “inflate” them by 30%? (the inverse of 0.77).

Added to that, they simply left out the Seventeenth Air Army losses (182 aircraft). It may have been an oversight or a deliberate effort to downplay their losses.

But, just to focus on the Second Air Army losses, the staff study has the total losses for the 5^th – 18^th as 371: 172 fighters, 31 bombers, and 168 assault. We have the Second Air Army’s losses for 5 to 18 July 1943, taken from their daily reports, as 481 (See Table IV.32 of Kursk: The Battle of Prokhorovka). This includes 248 fighters, 48 bombers, 180 assault and 5 night bombers. So actual losses of the Second Air Army were 30% higher than what was reported in the Soviet General Staff study, or 28% if one leaves out the night bombers.

One does wonder about the process where even the internal classified post-operation staff studies understate their losses (in addition to many other errors). They did have the unit records available to them. In particular, their table is vastly off on the 5^th of July when the Second Air Army lost 114 planes and the Soviet General Staff study reports only 78, but it consistently underreports for every single day. They also do not report the losses for the Seventeenth Air Army, which according to our count was another 182 or 221 planes lost (see Tables IV.34 and Tables IV.35). This does argue that the reported losses for the Sixteenth Air Army may be low compared to reality.

In the bigger picture, the Soviet General Staff studies are secondary sources, not primary sources. Furthermore they are secondary sources with considerable bias and errors. They invariably (grossly) overplay German losses and seemed to try to minimize their own losses. Furthermore their narrative of accounts often downplays certain aspects of their operations. They do have be used with extreme caution, as opposed to treating them as somewhat authoritative.

Now, Niklas Zetterling & Anders Frankson offer a similar discussion of the problems of relying on the Soviet General Staff studies in their book The Korsun Pocket: The Encirclement and Breakout of a German Army in the East, 1944. It is clear that these are secondary sources with biases that must be used with considerable caution.

This is the second part of my comparison of the data provided in the Soviet General Staff study on Kursk that was prepared in March-April 1944 compared to the Second and Seventeenth Air Army records that I have.

Losses:

            There are one table on losses in the Soviet General Staff study on Kursk that relate to the Second and Seventeenth Air Army. They are provided below. I have broken it into two tables for this blog:

The Air Struggle Along the Enemy’s Main Axis

                             Air          Enemy Losses:

                             Battles   Fighter   Bomber  Total

5 July                       81           71           83         154

6 July                       64           40           65         105 

7 July                       74           44           78         122  

8 July                       65           54           52         106

9 July                       62           49           22           71

10-14 July              152         112           93         205

15-18 July               43           45           27           72

Totals                     541         415         420         835

                            Second Air Army Losses:   

                            Fighter  Bomber  Assault   Total

5 July                       36           15           27           78

6 July                       23           —             22           45

7 July                       24           —             13           37

8 July                       24             1           16           41

9 July                       16             1           15           32

10-14 July                49           14           75         138

15-18 July              (the figures in the line above cover from 10-18 July)  

Totals                       172           31         168         371

            Now, these figures have been discussed before. The losses of the German VIII Air Corps was 111 planes, vice the 835 claimed here. The losses of the Second Air Army according to the records we reviewed was 481 planes from 5 to 18 July: see Appendix IV, Table II.32 (page 1424) of Kursk: The Battle of Prokhorovka), vice the 371 reported here. This report also does not include Seventeenth Air Army claims or losses. The Seventeenth Air Army’s losses were significant (182 planes). So, it does appear that the Soviet General Staff study basically leaves out 292 out of their 663 airplanes losses (44% of their losses), effectively under reporting their air losses by almost half.

       This is concerning, for it does appear that Soviet General Staff study is understating the Second Air Army losses, omitting the considerable losses from the Seventeenth Air Army and of course, grossly overclaiming the number of German aircraft shot down. This was in an internal classified report that was supposed to be an analysis of the battle. Hard to properly analyze if your data is not correct.

For my on-going Aces at Kursk book I was asked by the publishers to include a Chapter on the air war in the north from 5-11 July 1943. For the original Kursk project we were able to access the Second and Seventeenth Air Army records in the south. We did not attempt to obtain the Sixteenth Air Army records at that time (1993-1995). Therefore I was forced to rely on the Soviet General Staff study on Kursk that was prepared in March-April 1944 for the count of sorties and losses. As the staff study also reported the sorties and losses from the south, and I had the records for the air armies involved in that, I decided to do a little comparison and added a write-up of this to an appendix of the book.

Sortie Counts (I left out the table of the sortie count from Soviet General Staff study) :

          The Soviet General Staff study data on sortie counts is similar to the data we have assembled. The data we have for the Second and Seventeenth Air Armies operations are taken directly from the daily air army reports as drawn from the archives. The Soviet General Staff study may used these same reports, or used higher level reports or other assembled reports for their study. But there are minor differences between ours and their reports, so most likely they used other higher level or assembled reports for their study. For example, we have the Second Air Army flying 1,296 daytime sorties on 5 July. The Soviet General Staff study has them flying 1,274. There are also minor differences the next two days, but the two sets of counts are the same for 8 and 9 July and then vary slightly for most of the subsequent days (except for the 15^th and 16^th, where they again match). After the 5^th, the largest difference is on the 12^th, where our reports record 10 more daytime sorties. These are very minor differences. The Second Air Army nighttime sorties match in all cases between the counts we assembled from the air army daily reports and what the Soviet General Staff study reports.

            The Seventeenth Air Army is a little more complex as some of their missions were flown into the battle area while other of their missions were flown completely out of the battle area defended by the Voronezh Front. For the Kursk database project, I ended up reviewing each reported mission as to where it operated and made a judgment as to whether this mission was in the area of the Belgorod offensive or not. It does not appear that the Soviet General Staff study did that. For the 5^th through the 16^th, their estimate more closely matches with the total number of sorties flown by the Seventeenth Air Army than it does with my lower count of the number of sorties flown in the battle area. On eight of those 12 days in question, their totals matches the total we drew from the Seventeenth Air Army daily reports. The day they most differ was on 7 July when they reported 50 more sorties than we counted. We did re-check the original report and our total is 639. Suspect their number of 689 is a typo. As the Soviet General Staff study may have been drawn from a later aggregate report, there are multiple opportunities for typos.

           On the other hand, in the table we assembled of Seventeenth Air Army daytime sorties we had a lower count for “only those that were in the Belgorod Area or attacked the VIII Air Corps” (see table in Chapter Four). It is consistently lower from the 5^th through the 16^th, which the worse variance being on the 7th, where we count 588 as valid sorties in the battle area, whereas the Soviet General Staff study reports 689. On the 17^th we count none in the area and on the 18^th we count 12 sorties.

           Still there are a couple of observations we can make from this comparison. First, is that the Soviet General Staff study reports of Soviet sorties flown is fairly accurate in that it matches with records we have from the Second and Seventeenth Air Armies. This is important to note as we rely on the Soviet General Staff study for the count of sorties for the Sixteenth Air Army.