How Hot is Ukraine Gonna Get?

[Battlefront.com]

12 minutes ago, Vanir Ausf B said:

 

That's a terrific podcast but I did not remember Kofman saying that. After reading your post I went back and re-listened to his sections and he never mentions the UA needing better training, or anything else about UA training. Maybe it was Lee or Alperovitch? If you can find it give me the time stamp, please.

Kofman has been openly skeptical of the efficacy of western-style combined arms training for the UA since it's inception, so the idea that what the UA needs is more of it would be a rather un-Kofman thing to say no matter what his traveling companions may think. Here's something Kofman really did say about UA training by NATO forces back in December:

• Understand US is trying to find ways to improve outcomes and reduce UA dependence on high rates of arty fire. Less attrition, more maneuver. Training to do combined arms at company/battalion level is good in and of itself, but it won’t necessarily solve this problem.
• I have no doubt UA can learn combined arms maneuver, and saw elements of this at Kharkiv. However, without USAF air superiority, US logistics, C4ISR, etc it’s a bit hard to ‘fight like Americans.’ How well would we do without airpower?
• More importantly, it misses that attrition is what enabled maneuver in UA offensives. Against a well prepared defense, with sufficient density of forces, it wasn’t nearly as successful and casualties were high. This is why Kherson was so difficult compared to Kharkiv/Lyman.
• UA way of war depends on fires, exploited by maneuver. It is a successor military to the Soviet military, which was arty centric, and in that respect is much closer to the Russian military than our own.
• You have to work with what has proven successful for your partners. Deep strike, precision, better ISR, can help improve UA performance. My bias is that I’m  wary of seeing a solution that implies trying to turn that military more into us.
• That said, there’s no easy answer here. The US is not optimized to support a protracted artillery-driven war in Europe. Folks can also judge for themselves, looking at the history how good we are at converting other militaries to ‘fighting more like Americans'

https://twitter.com/KofmanMichael/status/1606637882994819072

Thanks for that clarification.  So it seems the group went to the front and came away with different, and quite divergent, points of view.

Kofman's, as posted above, is inline with what I believe to be the most realistic and pragmatic view of what's going on.  Ukraine could definitely benefit from improved small unit training, but in the end it isn't likely going to be the thing that wins them this war.  Giving them ways to overmatch Russia's primary means of defending itself is, such as better ways to challenge Russian air power is and striking ever deeper, are more pragmatic and likely effective ways to improve Ukraine's near future prospects.

What I'd be interested to see is a detailed analysis by current serving military personnel explicitly tasked with sorting out what the Hell is going on at the tactical level.  Do they think there are better ways for the Ukraine that exists today to breach minefields or tackle trenches?  If so, what would they recommend?

From what I can see of Ukraine's small unit tactics they are effective, but can be costly.  I'm not sure if there's any real practical way to improve this right now, other than Ukraine passing along Lessons Learned quickly and efficiently so that the next unit doesn't have to learn for themselves.  I have no idea how effective Ukraine is at doing this under current conditions.  I suspect it's at least OK, maybe better.

Steve

[_Morpheus_]

AMX-10 at the east. Soldiers really like the gun. Wheels are not protected and get flat under the shelling easily so manoeuvrability become problematic.  
PS: I would assume Leo 1 will be also warm welcomed in the troops

 

Edited July 19, 2023 by _Morpheus_

[Maciej Zwolinski]

23 minutes ago, Vanir Ausf B said:

That's a terrific podcast but I did not remember Kofman saying that. After reading your post I went back and re-listened to his sections and he never mentions the UA needing better training, or anything else about UA training. Maybe it was Lee or Alperovitch? If you can find it give me the time stamp, please.

My apologies, the statements of Kofman which are very similar to Gady's Twitter thread are in a different podcast - that's "War on the Rocks" podcast http://warontherocks.libsyn.com/assessing-ukraines-three-axes-of-advance.

They start at around 8.38 time mark.

I am sorry for the confusion. I listened to both podcasts on consecutive days, so when I heard some of Kofman's remarks on the "Geopolitics Decanted" they reminded me of the earlier podcast and got mixed up.

 

[Teufel]

1 hour ago, hcrof said:

It is possible to model a blast with a decent degree of accuracy, and this is done a lot when for example designing things you expect to get attacked, or to forensically determine what happened at a blast site. An example of some software that does it is viper, but there are others and they are all validated by physical testing.

https://www.viper.as/

The key, as you mention, is to get your inputs right and we just don't have the information to make an accurate assessment!

Thank you for the reference! Interesting to see what this thing can do with some known practical results as desired output. Input random parameters and see what it suggests is required to achieve desired results. 

On the topic of what predictive software is able to do and if that accuracy is good enough is down to probability and more importantly the application. Tried from experience to express it humbly - that violent chemical reactions are difficult to predict in theory and less practice. Then we exclude all external factors. If we include those then we are down to even more uncertainty and tons of different circumstantial parameters.

Being no expert in software simulation, my experience and responsibilities were in other fields. I would still maintain that systematic predictive simulations is difficult. At least for non-military R&D purposes. Reason why empirical, “blow things up and see what happens” testing is primarily method even in 2023.

Must keep in mind that lots of the R&D in the field is strongly standardized with plenty of system requirements. There can be thousands of requirements to test and verify before production can begin to put products into service. It’s not uncommon that testing what appears to be almost identical key parameters results in experimental outliers even when compared to larger sample sizes of almost identical tests.

Requirements could state that no more than 1-2% standard deviation is acceptable. Or in plain English, requiring results from predictions and reality to align with 98-99% statistical probability. That’s difficult to even achieve within practical testing due to randomness of reactions and standard error of measurements.

Not suggesting theoretical modeling is useless, quite contrary, wish it could predict reality better than empirical testing. Until then “blow it up and see what happens” still trumps applied mathematics, from perspective of industrial R&D. And still that occurs under well known conditions and parameters.

I dare not assume anything about field experiments even from few tests under known conditions. Less predictions with plenty unknown conditions and circumstances.

[DesertFox]

Lets see what happens - could be the end of the black sea fleet, if they make one false move...

 

https://twitter.com/ragipsoylu/status/1681683779054714881?s=20

[womble]

19 minutes ago, DesertFox said:

Lets see what happens - could be the end of the black sea fleet, if they make one false move...

 

https://twitter.com/ragipsoylu/status/1681683779054714881?s=20

They've already walloped the Odessa grain terminal and stated that the grain deal is over. Looks like they don't care what the potential purchasers of Ukrainian grain think of them. And now they're declaring unrestricted warfare against traffic into or out of Ukrainian ports... Trying to gain escalation dominance? Hoping to prove that nothing they do will provoke Western escalation? Will they really fire into Turkish-flagged vessels? Maybe they're just reserving the right to board and inspect.

I'm curious what the Russians mean by: (as the article linked above puts it) "...demands for Russian exports had not been honoured." Were some sort of concessions made on Russian exports as part of the grain deal?

The language is odd, since demands are not honoured, commitments are. Demands are acceded to, or permitted or some such, not "honoured". Maybe that's just linguo-cultural translation, or maybe it's indicative of the Kremlin's solipsistic attitude that all of their wants are automatically promises made by other people to fulfill them...

 

[hcrof]

42 minutes ago, Teufel said:

Thank you for the reference! Interesting to see what this thing can do with some known practical results as desired output. Input random parameters and see what it suggests is required to achieve desired results. 

On the topic of what predictive software is able to do and if that accuracy is good enough is down to probability and more importantly the application. Tried from experience to express it humbly - that violent chemical reactions are difficult to predict in theory and less practice. Then we exclude all external factors. If we include those then we are down to even more uncertainty and tons of different circumstantial parameters.

Being no expert in software simulation, my experience and responsibilities were in other fields. I would still maintain that systematic predictive simulations is difficult. At least for non-military R&D purposes. Reason why empirical, “blow things up and see what happens” testing is primarily method even in 2023.

Must keep in mind that lots of the R&D in the field is strongly standardized with plenty of system requirements. There can be thousands of requirements to test and verify before production can begin to put products into service. It’s not uncommon that testing what appears to be almost identical key parameters results in experimental outliers even when compared to larger sample sizes of almost identical tests.

Requirements could state that no more than 1-2% standard deviation is acceptable. Or in plain English, requiring results from predictions and reality to align with 98-99% statistical probability. That’s difficult to even achieve within practical testing due to randomness of reactions and standard error of measurements.

Not suggesting theoretical modeling is useless, quite contrary, wish it could predict reality better than empirical testing. Until then “blow it up and see what happens” still trumps applied mathematics, from perspective of industrial R&D. And still that occurs under well known conditions and parameters.

I dare not assume anything about field experiments even from few tests under known conditions. Less predictions with plenty unknown conditions and circumstances.

All modelling is a simplification of reality, but physical testing cannot necessarily be done for every conceivable scenario. In reality a combination is used, along with standards etc. 

Depending on how critical the scenario is there will be an emphasis on one or the other. Protecting a stadium from a hypothetical bomb threat will be done with modelling, standards and safety factors. Protecting a warship from known torpedo threats would include the above as well as physical testing. But no-one is going to torpedo their brand new warship for real to check if it works as intended!

Edit: here is a full ship shock trial. Note how they are not actually blowing up the ship directly - they are mostly checking all the fragile stuff like pumps and electronics can withstand a big bang

https://www.bbc.co.uk/news/av/world-us-canada-57547885

That is really expensive so is rare and not often done for smaller ships afaik. As well as this, computer modelling would be used for other scenarios, including direct hits. 

Edited July 19, 2023 by hcrof
More info

[dan/california]

 

3 hours ago, Battlefront.com said:

We have wondered how well NATO forces would do if they were in Ukraine's shoes.  With the same equipment restrictions, I'd wager they would do a little better.   Maybe keep casualties down a bit, but even then I'm not totally sure about that.  So call that a hunch.

The talk we've had about how NATO needs to figure out how to better survive this sort of war are all valid, however it seems we sometimes forget that the whole point of NATO's overmatch capabilities is so they don't have to.  Whether it be prolonged expenditure of munitions, loss of armored vehicles, air defenses, etc. it all comes down to NATO having designed its forces around the strategy of being able to deliver a "knock out punch" in the opening phase of conflict.  Therefore, this sort of war would only be fought by NATO if its primary strategy failed miserably.

Based on everything I've seen from this war up to this point, I have a high degree of confidence that NATO's  "knock out punch" strategy would work.  The best boxer in the world probably doesn't need to worry about how he's going to fight exhausted in the 10th round if his opponent is an egotistical bully with no track record of winning a fair fight.

Steve

 

2 hours ago, Vanir Ausf B said:

 

That's a terrific podcast but I did not remember Kofman saying that. After reading your post I went back and re-listened to his sections and he never mentions the UA needing better training, or anything else about UA training. Maybe it was Lee or Alperovitch? If you can find it give me the time stamp, please.

Kofman has been openly skeptical of the efficacy of western-style combined arms training for the UA since it's inception, so the idea that what the UA needs is more of it would be a rather un-Kofman thing to say no matter what his traveling companions may think. Here's something Kofman really did say about UA training by NATO forces back in December:

• Understand US is trying to find ways to improve outcomes and reduce UA dependence on high rates of arty fire. Less attrition, more maneuver. Training to do combined arms at company/battalion level is good in and of itself, but it won’t necessarily solve this problem.
• I have no doubt UA can learn combined arms maneuver, and saw elements of this at Kharkiv. However, without USAF air superiority, US logistics, C4ISR, etc it’s a bit hard to ‘fight like Americans.’ How well would we do without airpower?
• More importantly, it misses that attrition is what enabled maneuver in UA offensives. Against a well prepared defense, with sufficient density of forces, it wasn’t nearly as successful and casualties were high. This is why Kherson was so difficult compared to Kharkiv/Lyman.
• UA way of war depends on fires, exploited by maneuver. It is a successor military to the Soviet military, which was arty centric, and in that respect is much closer to the Russian military than our own.
• You have to work with what has proven successful for your partners. Deep strike, precision, better ISR, can help improve UA performance. My bias is that I’m  wary of seeing a solution that implies trying to turn that military more into us.
• That said, there’s no easy answer here. The US is not optimized to support a protracted artillery-driven war in Europe. Folks can also judge for themselves, looking at the history how good we are at converting other militaries to ‘fighting more like Americans'

https://twitter.com/KofmanMichael/status/1606637882994819072

NATO's knockout punch strategy would work against the current state of Russian air defenses. But I think the the west needs to have a very hard think about how long it current model of airpower based on large expensive platforms is viable. There are a number of things happening technologically that certainly aren't going to help Russia in this war, but they might matter a very great deal in Taiwan Straight in 2032, just to pick a number.

Edited July 19, 2023 by dan/california

[DesertFox]

I am sure we learn about the details in the next hours/days

https://twitter.com/PStyle0ne1/status/1681721639698874376?s=20

 

 

[DesertFox]

News from the Bunker-Clown...

 

https://twitter.com/NOELreports/status/1681725231075868677?s=20

[Teufel]

22 minutes ago, hcrof said:

All modelling is a simplification of reality, but physical testing cannot necessarily be done for every conceivable scenario. In reality a combination is used, along with standards etc. 

Depending on how critical the scenario is there will be an emphasis on one or the other. Protecting a stadium from a hypothetical bomb threat will be done with modelling, standards and safety factors. Protecting a warship from known torpedo threats would include the above as well as physical testing. But no-one is going to torpedo their brand new warship for real to check if it works as intended!

Edit: here is a full ship shock trial. Note how they are not actually blowing up the ship directly - they are mostly checking all the fragile stuff like pumps and electronics can withstand a big bang

https://www.bbc.co.uk/news/av/world-us-canada-57547885

That is really expensive so is rare and not often done for smaller ships afaik. As well as this, computer modelling would be used for other scenarios, including direct hits. 

There is no contradiction in any of your statements in relation what I just said. Perfectly reasonable examples of applied mathematics don’t necessarily contradict the absence of such elsewhere?

You are listing examples of system level acceptance - from perspective of sub-system test and verification. As in, test individual components and if they pass suggested criteria then it is assumed the system passes. To which I can suggest the existence of the opposite; failure of sub-system requirements doesn’t disqualify acceptance at system level. Both examples have the same results, but they have different levels of complexity.

Blowing up stadiums, aircrafts or ships - system level - is obviously unreasonable. My example was from fundamental physics and chemistry on molecular level of explosives. Suggested that from experimental experience that’s how explosives are studied. Just because it’s not feasible to do so with stadiums and ships does it mean that such logics applies across all fields?

We can all have differences in opinions, or agreements in such, but we can’t have our own facts.

[DesertFox]

Interesting look into the state of affairs of vehicles activated from storage...

 

https://twitter.com/wartranslated/status/1681569162684932096?s=20

[The_Capt]

6 hours ago, Battlefront.com said:

We have wondered how well NATO forces would do if they were in Ukraine's shoes.  With the same equipment restrictions, I'd wager they would do a little better.   Maybe keep casualties down a bit, but even then I'm not totally sure about that.  So call that a hunch.

The talk we've had about how NATO needs to figure out how to better survive this sort of war are all valid, however it seems we sometimes forget that the whole point of NATO's overmatch capabilities is so they don't have to.  Whether it be prolonged expenditure of munitions, loss of armored vehicles, air defenses, etc. it all comes down to NATO having designed its forces around the strategy of being able to deliver a "knock out punch" in the opening phase of conflict.  Therefore, this sort of war would only be fought by NATO if its primary strategy failed miserably.

Based on everything I've seen from this war up to this point, I have a high degree of confidence that NATO's  "knock out punch" strategy would work.  The best boxer in the world probably doesn't need to worry about how he's going to fight exhausted in the 10th round if his opponent is an egotistical bully with no track record of winning a fair fight.

Steve

The question facing military professionals everywhere out of this war are:

- "What is unique to this war?"

- "What is universal to all future wars?"

We had a whole thread going on the General Forum on development of warfare over the 19th century and I believe modern militaries are facing a similar conundrum.

"How would NATO do in this war?"  Well it depends which side we are going to be.  As Ukraine against Russia we would likely have seen a shorter sharper war but the costs would have been a serious shock to the western world. [note: let's not get dragged into another nuclear equation discussion, we can just put that one to the side]  We are talking likely tens of thousands of casualties and a lot of expensive kit lost.  Why?

- Air superiority.  I do not know what this means in a modern context.  A2AD capability is rapidly becoming distributed and highly portable.  We may have been able to gain air superiority over 20,000 feet but below that we would have been taking serious losses as there is not such thing as SEAD for MANPADs basically everywhere.  Modern MANPADs and IADs can operate independently all over the battlefield.  Further they can deny airspaces at much higher altitudes and higher ranges.  Why?  Because while we were stonking Iraq, Libya, Serbia and a bunch of dirt farmers in Afghanistan competing states were taking notes and investing heavily in the tech.  Take away our air supremacy and the western way of warfare is immediately in trouble.  And, shocker, places like Iran really don't like us and do not want to be invaded.

Below 20,000 feet it is the freakin wild west right now.  I do not care how many lasers we strap on every tank, IFV or truck.  I do not care how much EM is pumped into space - birds f#ucked up for the next 20 years.  Unmanned systems are 1) cheap, 2) highly effective and 3) everywhere.  Whether they are doing ISR or strike they have changed the fabric of warfare between about 3 to 20,000 feet...and they are just getting started.  Air superiority below 20,000 feet does not exist as a concept right now.  Hell we lost it below 2000 feet in Iraq to freakin ISIL, who were basically the lowest bar one can get with respect to conventional warfare.   If we were fighting the RA the UAS problem would be extremely costly...as in freakin nations pulling out after losing too many people, costly.  Can anyone imagine if the Taliban got their hands on this tech and started dropping old cluster munitions right on our heads back in the COPs and FOBs?  I slept for weeks about 200m from a 50,000 gallon fuel bladder that was resting under an open sky ...let that sink in.

So what?  Well "wither goest Air Superiority" is one of the biggest questions of this war, and as you can see it is a multi-dimensional one.

- C4ISR.  Russia does not have a world class C4ISR architecture.  But even with what they do have the principle of "making them go dark" to establish C4ISR superiority - far more important in this day and age then any domain superiority - is also in question.  With everything being a sensor hooked into crazy comms and networks - hell with hotspotting everything can be a node in a comms network.  So I am not even sure how to make an opponent go dark anymore (see unmanned).  I am sure we got people working on it but the fact that an even poorly armed opponent can see me tens of kms out makes me nervous.  Worse, they can see my logistics train as well.  The fact they can record all this and stream it all over the planet in real time turns really concerns me.  A half decent opponent would be broadcasting every screw up and horror show, which makes sustainment of national will a big problem.

- PGM.  Artillery, ATGM...insert whatever nightmare comes next.  No one is ready to face this.  I cannot begin to imagine trying to do an obstacle crossing when my opponent can hit me at 3-4kms with an 80% success rate with ATGM.  "Oh that is ok, we have APS"...fantastic, right up until someone comes up with workarounds like sub-munitions or EFP.  And even if we do magically put bubble wrap around ourselves, nothing on earth can stop artillery round that can land directly on my head.  Oh and this is while I am still trying to deal with old stuff like mines, and new stuff like UAS.

All of that  adds up to some very disconcerting calculus.  As in "is combined arms dead as we know it?" type of calculus (someone is going to try and answer this, someone always does...just don't bother.  I do not post my mil quals for some very good reasons but trust me when I say no one has this figured out yet).

Now here is the punchline: this is all if we were fighting Russia.  I, frankly, am far less concerned about fighting Russia - now more than ever.  I am very worried about fighting Ukraine.  If we get stuck on the wrong side of a proxy war and our opponent is armed with C4ISR, PGM and A2AD like Ukraine is right now, we are in very serious trouble. 

"Well we just won't fight those wars."  Ya, that is not how it works.  We don't get to choose the wars we decide to fight, gawd that is a post-Cold War perception that needs to die, and fast.  This is the nightmare scenario and I do not know if you guys have been paying attention but we kinda been doing a lot of expeditionary operations in all sort of places to push the brand.  What happens when Chinese space based ISR start lighting us up?  We wind up in a hybrid fight with the other side armed with HJ-12s?

I do not know.  This is a big reason when [insert talking head] says "Ukraine needs to do this"  "We need to give them that"...and the war will be over in a week.  My advice is to stop listening.  No one in the west has been in a war like this since Korea and the freakin needle has moved miles since Korea.  I say this without hyperbole, we are going to be spending the rest of this century trying to figure this all out as things like UGVs start coming online.  Tell your (grand) kids to get into the sciences of killing because it is a major growth industry.  For now, the best we can do is watch and learn.  Both the UA and RA are feeling their way through a war unlike any we have seen before.   

[cesmonkey]

https://www.defense.gov/News/Releases/Release/Article/3463890/biden-administration-announces-additional-security-assistance-for-ukraine/

Quote

The capabilities in this announcement, which totals $1.3 billion, include: 

Four National Advanced Surface-to-Air Missile Systems (NASAMS) and munitions;
152mm artillery rounds;
Mine clearing equipment;
Tube-Launched, Optically-Tracked, Wire-Guided (TOW) missiles;
Phoenix Ghost and Switchblade Unmanned Aerial Systems (UAS);
Precision aerial munitions;
Counter-UAS and electronic warfare detection equipment; 
150 fuel trucks;
115 tactical vehicles to tow and haul equipment; 
50 tactical vehicles to recover equipment;
Port and harbor security equipment; 
Tactical secure communications systems; 
Support for training, maintenance, and sustainment activities.

 

[dan/california]

1 minute ago, The_Capt said:

We had a whole thread going on the General Forum on development of warfare over the 19th century and I believe modern militaries are facing a similar conundrum.

Could you drop a link to that?

 

[DesertFox]

How´s Private Mobiks family doing at home? Get an impression here...

 

https://twitter.com/wartranslated/status/1681711838537318411?s=20

[dan/california]

Now ask yourself what would be a higher priority than a system almost perfectly suited to knocking down drones? Because my read is that everything else coming out of Russian depots, and factories is probably worse.

[The_MonkeyKing]

2 hours ago, DesertFox said:

Lets see what happens - could be the end of the black sea fleet, if they make one false move...

 

https://twitter.com/ragipsoylu/status/1681683779054714881?s=20

As far I have understood Ukraine has been under total naval blockade since the start of the invasion.

Russian approved and inspected grain shipments were the short lived exception to this.

Nothing new

[sburke]

14 minutes ago, The_Capt said:

"Well we just won't fight those wars."  Ya, that is not how it works.  We don't get to choose the wars we decide to fight, gawd that is a post-Cold War perception that needs to die, and fast.  This is the nightmare scenario and I do not know if you guys have been paying attention but we kinda been doing a lot of expeditionary operations in all sort of places to push the brand.  What happens when Chinese space based ISR start lighting us up?  We wind up in a hybrid fight with the other side armed with HJ-12s?

Trying to look out just maybe a decade and the question of what conflicts we can expect is a guaranteed migraine, however there are some things we can see right now that are sure bets to get worse.  Water, food and migrations away from places that are unable to sustain human existence are the projections we can expect.  That in turn sets expectations on the potential adversaries and types of conflicts.  I don't know that we will see another war like this current conflict. However unless we can find some global agreement on how to face this future, it is going to be freakin chaotic as hell.

 

[FancyCat]

7 minutes ago, The_MonkeyKing said:

As far I have understood Ukraine has been under total naval blockade since the start of the invasion.

Russian approved and inspected grain shipments were the short lived exception to this.

Nothing new

short lived? this deal existed for a year according to wiki, considering the war has been ongoing for a year and 5 ish months, most of this war has had the deal as a staple, not a short lived exception. 

[The_MonkeyKing]

8 minutes ago, FancyCat said:

short lived? this deal existed for a year according to wiki, considering the war has been ongoing for a year and 5 ish months, most of this war has had the deal as a staple, not a short lived exception. 

One growing season is not long. But semantics.

Grain is only a portion of the whole Ukraine sea cargo and that even that part was further constrained by the Russians.

Ukraine has been in total sea blockade except grain shipping that was "just" constrained.

[Hapless]

1 hour ago, dan/california said:

Could you drop a link to that?

[JonS]

18 hours ago, chris talpas said:

An interesting and educational video recently released by Veritasium on fireworks illustrates the impact of confining an explosive quite nicely (see around 6-8 minutes and the 12:34)

 

One you can do at home:

Mentos + coke in a shallow bowl

Vs.

Mentos + coke in a coke bottle

At home, but I do reccomend doing it outside ...

[Carolus]

Wagner associated telegram channel released numbers about casualties of the PMC:

• totally employed 78.000 fighters, 49.000 of which were convicts and 29.000 were regular mercs
• 22.000 killed, 40.000 wounded, so ca. 62.000 casualties of 78.000 in total (79,5% casualty rate, 28% death rate)
• 25.000 are now "alive and healthy", so 16.000 who were never wounded and ca. 9.000 who recovered (so 32% of deployed force still available)
• 10.000 of those supposedly already in or going to Belarus as per agreement after the mutiny / coup attempt 

Convict troops had 90% casualties, so 44.000 either killed or wounded from 49.000, and the rest of the casualties, being 18.000, must have come from the 29.000 mercs (ca. 62% casuality rate mercs only).

Recovery rate returning to service roughly 1 in 4. Deaths to wounded ratio ca. 1:2.

Does not account for troopers who were wounded multiple times, so rates might be lower.

Also I suppose the surviving convicts would get released and reduce the number of PMC members by 5.000, unless they sign up again on their own, so ca. 20.000 PMC members left.

Source is Wagner-related so all regular caveats apply etc.

Edited July 19, 2023 by Carolus
corrections and additions

[Carolus]

5 minutes ago, Carolus said:

Wagner associated telegram channel released numbers about casualties of the PMC:

• totally employed 78.000 fighters, 49.000 of which were convicts and 29.000 were regular mercs
• 22.000 killed, 40.000 wounded, so ca. 62.000 casualties of 78.000 in total (79,5% rate)
• 25.000 are now "alive and healthy", so 16.000 who were never wounded and ca. 9.000 who recovered (so 32% of deployed force still available)
• 10.000 of those supposedly already in or going to Belarus as per agreement after the mutiny / coup attempt 

Convict troops had 90% casualties, so 44.000 either killed or wounded from 49.000, and the rest of the casualties, being 18.000, must have come from the 29.000 mercs (ca. 62% casuality rate)

Recovery rate returning to service roughly 1 in 4.

Does not account for troopers who were wounded multiple times, so rates might be lower.

Also I suppose the surviving convicts would get released and reduce the number of PMC members, unless they sign up again on their own.

(Edit: my apologies for the unintended double post - I quoted myself instead of editing)

Edited July 19, 2023 by Carolus