Ukraine war: Part 15. The reality behind a peace deal

 In my last blog post I had ended my analysis with this:

Winning an attritonal war depends on the manpower reserves of each side (Russia is six times more and suffering fewer casualties) the ability to supply armaments - Russia is able to replace hardware losses. Ukraine cannot – even with NATO but not US support), and the will to fight. 

My sense is that the Ukrainian army is sufficiently attritted for Russia to start a spring offensive once the ground hardens after the spring thaw. Once they take most of the four provinces (with territory captured in other areas exchanged in an eventual peace deal), Russia will dig into defensive positions, from where they will offer a ceasefire – which they are under a lot of pressure to do, under the threat of more US sanctions. Once on the defensive, Russian casualties and consumption of armaments will be lower than while they have been attacking. Ukraine’s losses will be correspondingly higher, which will facilitate a de-facto ceasefire line for an eventual peace deal, as I argued in a previous article. 

In my blog post of Feb 25, I had suggested how the war might end:
How the war might end

My view was that a peace agreement would precede a ceasefire. It would end on the current contact line (delays by NATO in accepting that proposal will result in Russia improving their position).
No NATO for Ukraine, but an acceptable security guarantee and a buffer zone to separate the armies, with a removal of multilateral sanctions and free movement of people to either side. 

There have been updates since then which validate these conclusions:

Ukrainian desertions have been increasing at an increasing pace:
These are cases filed in Ukrainian courts, under section 407 (temporary absence from place of duty) or
Section 408 (deserted and not traceable). Hence, this is a number from official Ukrainian sources.

	Section 407 Temporary	Section 408 Permanent	Total deserted.
2022	6641		6641
2023	16615	7771	24386
2024	67840	23343	91173
2025 (Till 1 Aug)	90216	35000 (est)	125216

While the total number in this table is 247416, pending cases are listed as 238348, as of 30 Jul 25,
as per the Ukrainian prosecutor general’s office. The Russian estimate of Ukrainian desertions for the same period is 318000.
When considering these numbers, one should keep this in mind:

Commanding officers don’t like to report section 407. It is procedurally cumbersome, when units are at the front. His commander will prefer to cut a deal with a AWOL case to return, or for being absent, avoiding the larger bribe he would have to pay the courts.

Many cases filed under article 407 were subsequently closed. Hence the total for 2023 & 2024, will not correspond to a higher figure I mentioned in an earlier blogpost. 
Conversely, a lot of 407 cases will eventually become section 408, but will continue to be recorded as section 407 (temporary absence).

A lot of 408 cases are actually killed in action, but by reporting them as deserted (which happens when
the body cannot be recovered) the family does not get compensated. 
There is an amnesty scheme for deserters, but only about 1800 deserters (mostly section 407) reported back for duty.

The total number of desertions under article 407 & 408 has been consistent since Oct 2024. The size of the Ukrainian army at the front since Oct has been slowly declining – Ukrainian sources also acknowledge that irrecoverable casualties are higher than replacements. Presumably when desertions start, the soldiers least willing to fight desert first. There should typically be a declining trend in desertions, if units are left with a core of motivated soldiers, but that is not happening. 

If there is a peace deal which involves surrender of the remaining territory in the Donetsk province,
soldiers will question the need to defend difficult positions, with very high casualty rates, rather than
retreat a few km, outside the area claimed by Russia.

KIA: There is also an update for `Killed in action’, as per the website ualosses.org
This has names, with date of birth and death and region they belong to.
It has 153751 losses, upto 7 Aug. There is usually a time lag in reporting deaths. For e.g. there are 615
deaths in the last 6 weeks (102 per week), against an average of 242 per week for the first half of the year. This covers cases where day of death has been recorded.
This does not include foreign fighters. As I have analysed in a previous blog article, it may not include some territorial army or national guard units, which were not part of the regular army, or soldiers
who hailed from districts now under Russian control.

Update: A closer look at the Ukrainian confirmed dead, shows the following:
- The figures have not been updated for the past month ( no change from 1 to 27 Aug 25). 
- Units in the thick of fighting, have very high confirmed dead. For e.g
The 36th Marine Brigade has 1496 confirmed dead. 
The Azov brigade has 1389 dead. 

Given that the starting strength of an average brigade is around 3500, with a combat strength of around
2500 and assuming 2 seriously wounded for every dead, plus one more deserted or missing, the combat 
strength of each brigade has over 5000 irreplaceable casualties. That means each brigade has been wiped out twice. The current manpower is mostly inexperienced conscripts and the brigade is very 
understrength. The Presidential brigade, considered a unit of `last resort' has 408 confirmed dead.     

Conversely, territorial army or national guard units have relatively few casualties. This may be the result of less interest in compiling figures, or more desertions.  

There is a separate category of `missing in action’ (usually presumed dead). This number is around 70,000, It is reduced when a missing person is reported dead. Getting on the `missing’ list requires the
unit to confirm the soldier has not deserted.
 
Assuming 153751 dead + 70000 missing presumed dead, there would be two seriously wounded for
every man killed. That gives 447500 seriously wounded and unable to fight. This is consistent with a figure of 120000 amputations of Ukrainian soldiers (a norm of 1 amputation for every 4 seriously wounded).  Adding 67000 deserted under article 408 and assuming half of those under 407 who have effectively deserted, gives another 88656 deserters.
There are also around 10,000 Ukrainians taken prisoner.
The total number of irrecoverable casualties is therefore 836907

The total number of Ukrainians mobilized was:
250,000 (regular armed forces at the start of the war).
125000 (National guard).
  75000 (Armed and border police).
600000 (mobilized. Priority was volunteers with experience, other volunteers and reservists who did
                not volunteer on their own).
300000 (mobilized in 2024 – mostly inexperienced and did not volunteer.
60000 (estimated volunteers in 2025 – this total has been 10-12000 per month since last Oct), since
            the last cohort of 25-27 y.o men and previously exempt men were mobilized.

Thus of the 1410000 personnel mobilized, Ukraine has lost 837000, including most who had any experience in combat units. Many of the remainder are lightly wounded and returned to combat.
As there is an increasing trend of people trying to leave the country, the size of the border guard cannot be reduced, nor can those in supply and maintenance functions, given the number of diverse weapon systems Ukraine has, or engineers, since fortifications have to constantly be built and minefields laid.

Russian figures of killed in action: From Mediazona (the most accurate source of casualties) was 121507 dead as of 1^st August. I have dealt at length with Mediazona’s attempt to use a new methodology to give a higher estimate of Russian casualties. Even that estimate, which has flaws in methodology, has an estimate of 165000 dead.  There are a negligible number of missing, deserted, or prisoner. All considered, Russian irrecoverable casualties would be around 400,000.

Russian dead include 18000 convicts, Wagner mercenaries (2800), foreign fighters (546) and fighters
from the Donetsk region of Ukraine.  

In terms of exchange of bodies, since Dec 24, 10744 bodies of Ukrainian dead were exchanged for
281 Russian. While an advancing side (Russia) will recover more bodies than a retreating side, this is one indicator of the loss ratio. In  2024, this ratio was approx. 5 Ukrainian bodies for 1 Russian.
Update: On 18 Aug, there was an exchange of 1000 Ukrainian bodies for 19 Russian.

Russian frontline strength (the strength of all combat units at the front) is now over 600000 – both Russia and Ukraine have similar estimates – President Zelensky has officially put this number at 700,000. Ukrainian frontline strength is under 200,000. Apart from having a 3:1 superiority for the first time in this conflict, all Russians at the front are volunteers and are being periodically rotated, while the Ukrainians are now mostly conscripted and are unable to be relieved from their positions as there are no replacements. 

Closely related to Ukraine’s desertion problem is the increasing fatigue with the war among Ukrainians
and a reduced will to fight. Opinion polls show that support for the war (fight till victory) has fallen from 75% in 2022 to 24% today. Those who want a negotiated settlement has increased from 20 to 69%
(74% of those who expressed an opinion).
The survey excludes districts close to the fighting or occupied by Russia (no telecom facility), so the true number who want a negotiated settlement is close to 80% and rising each month. 

Ukraine’s weapons problem:
The sites lostarmour and Oryx provide visually verified details of Ukrainian hardware losses.
If one compares the total amount of Western aid promised to Ukraine, against the number confirmed to be destroyed, Ukraine has started seeing a reduction in total hardware available to it.

Losses of hardware are exceeding replacements, whereas for Russia, as I had concluded in an earlier blog post, production has exceeded losses/consumption since the beginning of 2024.

Weapon system	Quantity promised by West**	Delivered by west / Ukraine pre-war stock	Confirmed loss
Western tanks	351	271	111
Soviet design tanks	562	457	347
Ukrainian tanks		898	784
Western IFV’s	1230	1126
Ukrainian IFVs		1341	936
APCs	2791	2234	1113
Western Self Propelled Artillery	588	408	211
Ukraine SP Artillery		785	405
Western fighter aircraft	82	55	26
Ukrainian fighter aircraft		123	84

* Figures from Lostarmour. Oryx. (incl. captured/abandoned)
** Western commitments to supply as of 1 July 

Losses are based on photographic evidence. A lot of Ukrainian equipment lost behind the front lines will not be recorded by western sources. Actual losses would be higher than verified numbers.
At best, 2/3rd of the remaining number would be battleworthy. Spares for many of the Western platforms are not being produced. For e.g of the 160 Western tanks remaining with Ukraine, 142
are the obsolete Leopard 1, or early Leopard 2 designs – the production of the Leopard 2 ended in 1994.
Tanks committed, but not supplied is because of the difficulty in making obsolete tanks operational, before shipping them to Ukraine – not a lack of will. 

What this suggests is that Ukraine (and NATO) lack the hardware to carry out any large scale offensive to regain any of their lost territory. They also lack experienced and motivated infantry.

The most critical weapon system is the Patriot anti-aircraft missile. However, US stocks are
at 25% of required levels. Against a US requirement of 13733 missiles, the US has barely 3500.
Lockheed Martin makes barely 500 missiles a year – a capacity that is being increased to 650/year.

Against the depleting supply of anti-aircraft missile systems and missiles, Russia has increased the
production of long range (Geran) drones 10 fold in the past year. More drones have to be intercepted by a diminishing number of missiles.
Between Jan and June 2024, Russia launched an average of 400 Geran drones per month. In June 2025, Russia launched 5400 drone strikes. The drones are being upgraded (to jet drones).
My last blogpost covered the drone war:
Ukraine war - part 14 
I had said:

The use of drones by Ukraine has increased from roughly 2000 in Jan 23, to 20,000 in Jan 24 to
200,000 drones in Jan 25, as estimated by both sides. Ukraine is estimated to increase drone use to
500,000 per month by Dec 25. An Estimated $ 2.3 billion was spent on Ukrainian drones in 2024 and $ 2.6 billion already announced in 2025. (Source: War Quants). Russian use of FPV (hand held) drones has gone up at a similar rate, but most likely below that of Ukraine, which is sourcing drones from all NATO countries. Russia's FPV drone production target for 2024 was 1.4 million.     

Russian use of long range (Geran/Shahid) drones has also increased tenfold between Mar 24 and
Mar 25. Ukraine recorded 4198 Geran drones used by Russia in March, against 356 in Feb 24.

Russia has also acknowledged that over 80% of all its casualties (dead + wounded) are inflicted by drones. An attacking side (Russia has been attacking since Nov 2023, barring a few days Aug 24 when Ukraine entered Kursk) loses more men to drones, because attackers are visible and have to move in the open, whereas defenders are dug in, in trenches, or built up areas, which reduce the chance of drone strikes.

As a proxy for Russian armament production, consider the T-90 M tank.  
At the start of the war, Russia had 370 older T-90 tanks and 75 T-90M (newer version) tanks.

In 2022, Russia produced around 70 T-90M Tanks. In 2023 it was around 160, in 2024, 200.
Till July 2025, T-90 production is estimated at 140. (capacity has been increased to 300 per year) Thus Russia produced 570 T-90M during the war, while it has lost an estimated 160.

In the first 6 months of 2025, Russia lost 205 tanks Only 12 of these were T-90s (model unknown).
34 Tanks were obsolete T-64s.
Against this, Russia is estimated to have produced 150 new (120 T-90M and 30 T-80 tanks (restarting an
old production line) and another 500 refurbished T-72 and T80s.

Note: Losses of Russian tanks from open source data, correspond to the losses of tank crew in Mediazona. 

Similarly, the Russian air force has more frontline SU-34 & 35 aircraft today than when the war started. The same goes for all categories of long range missiles. Against a loss of 11-12 SU-34 & 35 aircraft since 2024, Russian factories have delivered 25-30, SU34 & 35 aircraft, between Jan 24 and Jun 25. Russian glide bomb usage has increased from 50 per day in 2024 to 120/day currently, suggesting an increase in the number of sorties from the Russian air force – though aircraft availability rates should have fallen after three years of war.

What keeps Ukraine going ?
What has changed in the war is the widespread use of drones, which I wrote about in a previous blog post. Drone usage has increased more than 10 fold in the past year. Ukraine uses 6-7000 FPV drones a
day (Russia almost the same number) to target any vehicle or groups of people. Coupled with hundreds of NATO satellites over the battlefield, the defender is at a big advantage, when artillery can start hitting the assembly points of any attacking group and drones can start targeting individual vehicles or even soldiers when they start advancing.

Advances have to be made by groups of 3-4 individuals (not even a single section) who either use motorbikes, or a APC protected against drones. This does not lend itself to `big arrow’ advances, but a series of small advances, each a couple of hundred meters, typically from one tree line, or building to another.  

That said, Russia has been taking more territory each month since March 25. The advance has been without any surge in casualties that characterized earlier advances. One can conclude that Russia is casualty averse (no evidence of any human wave tactics) and is not advancing based on a time table dictated by negotiations.

Russia's stated goal since the beginning of its special military operation is the demilitarization of 
Ukraine. As long as Ukrainians are drying trying to defend difficult positions with a favourable loss ratio for Russia, there is no need for Russia to take more casualties by advancing more aggressively. 
As discussed in earlier posts, the loss ratio swung in Russia's favour since 2023 and this ratio is becoming more adverse for Ukraine as time passes. Russia is also producing more weapons than it is losing, so it can afford to dictate the pace of operations. The only other variable is the Russian economy. Secondary sanctions might cause it real trouble and seek a ceasefire.

What’s next: I will cover progress on the various fronts in my next blog post. At this stage I would argue that Russia is close to reaching its territorial objectives. They have 100% of Luhansk and about 2/3^rd of Donetsk (which they need the whole of) and 2/3^rd of Kherson and Zaparozhye  - which they probably do not need more of, hence the offer to freeze the front in these two districts. They also have a small portion of Kharkov, Sumy and Dnipro districts ,which equal the territory they need in Donetsk. I believe Ukraine’s choice is either to surrender the Donetsk now, as part of a peace deal, or lose it completely in a few weeks – with another 50,000 casualties.

A clarification on casualties: This is an emotional subject and even trying to explain it with logic and facts, got me banned from several western forums. I'd like to clarify the following:

Casualties can be crossed checked by a simple accounting principal: 

Starting strength + recruitment - losses - current strength = irrecoverable casualties. 

When both sides agree on each number left side of the equation, the result must be correct. 

Both side's reporting of each other's losses, use the Soviet term `sanitary losses'. These includes 

all lightly wounded to return to combat. 2/3rd of the wounded return to duty. The rest form irrecoverable losses. 

I use Western data to report Russian losses. The methodology ensures it gets more accurate over time. 
Ukrainian data is cross checked from other sources - e.g. losses of specific units in combat reported

by the unit, amputations, new graves in a town with a known population etc. 

Further viewing:

A Russian soldier's account of life at the front and the challenges of drones, training and surviving. 
(dubbed in English)
https://www.youtube.com/watch?v=B1QHxbpfD44&t=180s

A Russian video on Russia's use of drones (English subtitles), From 28 mins onwards, a view of Russia's Geran drone production facility which gives an idea of the scale of production. This facility is now increasing capacity. https://www.bitchute.com/video/aYCAXILeMU2O

Russia has started using Chinese sourced anti drone lasers, which can fry the electronics of a drone 
2.5 km away. Here is an earlier thread by Michael Kofman, a generally pro Ukraine source who describes Russian tactics.

Russian drone usage

An article from the German press on Russia's jet drones (Geran upgrades).
Russian jet drones (in German)

Mapping Russian air strikes on the frontline:
https://x.com/clement_molin/status/1955602490994401453

Update: On 7th Sept, a month after this post, deaths reported by ualosses.org, increased by 6137,
Russian deaths in that same period as per Mediazona increased by 4167. 
However, the number of Ukrainians deserted (using the numbers above) are estimated at 11444 with 
a negligible number on the Russian side. This is an irreplacable casualty ratio of over 4:1 in Russia's favour, though an attacking side can generally be assumed to incur high casualties, particularly against a dug in opponent, with the attackers not having any surprise. This is what attrition warfare is about. taking territory is a consequence of inflicting casualties.