Reddit reviews Soviet Casualties and Combat Losses in the Twentieth Century

Per Krivosheev, out of 1 836 562 of POWs who got back into USSR (and about the same number emigrated), 233 400 were sentenced to various terms in GULAG. That's ~12%.

You need to distinguish several major POW categories:

1. POWs, freed from extermination camps. Almost never got harsh sentences.
   
   
2. POWs who fled from camps and broke thru (or so they claimed) enemy lines or joined guerilla bands ('Partisans'). Were they turncoats, trained for spying and terrorists attacks? Or were they honest soldiers? That was the question for SMERSH ("СМЕРть Шпионам" - "Death to Spies") operatives. Some of Brandenburgers were indeed placed into camps, then staged breakthrough with 'real POWs' to gain credence - see Shellenberg memoirs:
   
   ---
   Thousands of russians were picked in POW camps, who were infiltrated by paradrops into russian territory. Their primary goal was political propaganda and diversions. Other groups were targeting Partisans (e.g. pro-Soviet guerilla)
   ---
   
   (I have access to russian-translated Shellenberg book, that's quick translation into English)
   
   
3. POWs who changed sides and were captured while fighting along with germans - Hiwi. Kinda self-indicted and well, real traitors - but not all of them were shot on spot by troops. Many were, yes - and they weren't counted.
   
   
4. (Not strictly POWs) soldiers, who were left behind enemy lines and who either broke thru enemy lines or were living on occupied territory which were freed and then all populace filtered. This is by far the largest category - and some of them served as Hiwi or took part in nationalistic guerilla bands (Ukranians, Chechen and Crimean tatars mostly), which were conducting acts of terrorisms. Brandenburgers were also mixing up with (окруженцы - encirclee)
   
   Toughness of interrogation and chance of being acquitted depended a lot on personality of an interrogator.
   
   >how accurate Soviet methods for assessing treason were?
   ---
   
   I am not aware of any 'accurate' methods for assessing treason...
   
   So it was inaccurate, of course - if interrogator was able to get in touch with your unit and it could provide a glowing recommendation - you're free. If your CO was alive and asshole or great, but KIA - you might get a long sentence - depending how well you were holding up on interrogations. Mostly circumstantial evidence was used in convictions - like lack of serious wounds, general attitude to USSR, any inaccuracies in accounts how you've been captured, whether personal papers were saved or lost etc.
   
   Sources:
   
   
5. Schellenberg, Walter (2000) [1956]. The Labyrinth: Memoirs Of Walter Schellenberg, Hitler's Chief Of Counterintelligence, translated by Louis Hagen. Da Capo Press. ISBN 978-0306809279
   
   
6. Russian Wiki - Soviet POWs in Great Patriotic War
   
   
7. G.F. Krivosheev - Soviet Casualties and Combat Losses in the Twentieth Century Hardcover
   

The most comprehensive source on this is Krivosheev's Grif sekretnosti snyat, which has figures for all Russia's/USSR's 20th century conflicts with WW2 of course being most prominent among them. Here is the text in Russian, the WW2 chapter is the fifth one; the wealth of information is summarized in the tables, and I think you'd be able to get at most of it via Google Translate. (The book has an English translation, but appears to be both abridged and highly expensive to boot).

As for your specific question, the answer is - not long. As was pointed out, 1941 was an extremely deadly year, accounting for slightly more than 25% of total irrecoverable despite being only half a year. 1942 - another 25%. So, ~55% in 1941-42. The statistically most likely fate for a soldier called up in 1941 was to be captured in one of the great encirclement battles, and die in a Nazi PoW camp. If he survived through to 1943, his risk profile would slowly converge and, from 1944, begin to look better than his equivalent in the Wehrmacht. In particular, his risk of capture would drop dramatically henceforth; the risk of being killed would substantially fall, though it would still remain extremely high relative to most armed conflicts; and his risk of getting wounded would start exceeding the risk of getting killed by several factors (a high WIA-to-KIA ratio is a sign of a well organized military).

Also worth pointing out that risk profiles differed quite radically for different branches of the armed forces. I don't recall the source, but I remember reading an estimate of "life expectancy" (that is, from induction until KIA/MIA/WIA/POW) in 1941 for them: It was around 3 months for infantry vs. 3 years for artillerymen, with intermediate numbers for tank men, airmen, etc.

Ugh, not at my fingertips. Wikipedia gets its source here. The battlefield was larger than many European countries and the scale boggles the mind. According to the estimates, the Soviets lost more tanks than they started the battle with.

To hang all tank losses on any one individual is rather trenchant tendentious, though I don't know who would be a likelier target than Zhukov.

I think of Kursk as the high point of "total conventional war" doctrine. It didn't really have anywhere to go after that and I notice that superpowers lost all vigor for fighting each other since then.

EDIT: Trenchant is the wrong word.

Hey I am back, and here are the sources you asked for. First one is from wikipedia though I would point out that apparently citations are needed for the section in question.

http://en.wikipedia.org/wiki/Equipment_losses_in_World_War_II#Land

The wiki gives about 42,000 losses of T-34s

Here is another from a blog that uses this book as a source.

http://chris-intel-corner.blogspot.com/2012/07/wwii-myths-t-34-best-tank-of-war.html

They give 44,900 losses

And another which also gives 44,900 as the total losses:

http://www.operationbarbarossa.net/the-t-34-in-wwii-the-legend-vs-the-performance/#Conclusions%20Regarding the T-34’s Overall Performance as a ‘War Winner’

Here is a discussion of casualty figures from Zaloga, and from the author of the book I linked:

http://www.dupuyinstitute.org/ubb/Forum5/HTML/000024.html

I admit I rounded up to 45,000 when I made the album.

Hope that was helpful!

https://en.m.wikipedia.org/wiki/Great_Purge

https://en.m.wikipedia.org/wiki/Purge_of_the_Red_Army_in_1941

https://www.historynet.com/stalin-attacks-red-army.htm

https://en.m.wikipedia.org/wiki/World_War_II_casualties_of_the_Soviet_Union

Incompetent leadership, disease, frostbites killed more people in the Eastern Front than soldiers killing each other. This book pretty much covered it