Wagner's strength at the beginning of 2022 was rather murky, and AFAIK no good estimates existed. But based on older estimates, it was a minimum of 5K globally, perhaps much more.
Wagner as I understand typically maintains several thousand, say 1-2K personnel, in Russia at any time as its training, recruitment, and admin base, as well as personnel on rotation. Some thousands are present at any time at Wagner's bases of operation in Syria and throughout Africa. I do not expect a significant level of rotation between Africa/ME and Ukraine, because of the difference in built-up expertise between branches, as well as the complications of long-distance travel. I have always excluded Wagner personnel outside Ukraine from my calculations.
Wagner first appeared publicly in Ukraine in March, right around the seasonal transition. AFAIK Wagner was considered to have 5K personnel in Ukraine by the height of the spring offensive in May (when it was relying on LDPR conscripts for bulk infantry, somewhat akin to the future role the convicts would fill). Wagner was fundamentally implicated in the Popasna salient, the outflanking of Lysychansk, the grind towards Bakhmut - more or less all RuFOR progress on the central front since the start of the invasion.
IIRC public information on the convict recruitment campaign indicates that Prigozhin was visiting prisons from June, clearly looking ahead. It took until late summer for recruitment to really pick up, during which time the separatists were given some form of reprieve and the Russian military's draft was initiated. July and August in general were the quietest months of the war up to then, as both sides regenerated their expended resources and HIMARS caused a temporary shock to Russian logistics. UFOR also fully integrated 155mm artillery platforms and RuFOR focused on preparing to meet future UFOR offensives.
Throughout the fall Wagner escalated its activity in Donetsk again, maintaining a heavy presence along the central front from Toretsk/Mayorsk all the way to Bilohorivka. Progress was very slow. The absorption of and transition to penal assault waves was not complete and UFOR still had some space to trade along the Bakhmut Line.
By December, US intel releases claimed that Wagner fielded 10K pros and 40K convicts. This is one of those curiously-low figures that comes out of US intel. For example, I seem to recall a US estimate last summer of 5K Wagner KIA, a number I can no longer locate cite for, and which I suspect includes casualties of separatists commanded by Wagner. As of early February this year, the estimate was at least 9K KIA total, most of which over the winter, obviously inconsistent with the earlier elusive estimate. This week's estimate was of around 10K Wagner KIA from either the beginning of the year or the beginning of December (unclear). 1K Wagner KIA in almost 3 months of fighting in Bakhmut is absurdly low and well below Prigozhin's intimations anyway, even if over the course of the year Wagner's area of responsibility in the central front has shrunk and shrunk until by the early spring they were pretty much only actively fighting in Bakhmut itself and its immediate outskirts, with regular military and a few minor/new PMCs (which may or may not themselves be offshoots of Wagner) supplementing them. 10K Wagner KIA January through April would at least be more consistent with the US February estimate. As of late March, Milley attested that "It's probably about 6,000 or so actual mercenaries and maybe another 20 or 30,000 recruits that they get."
But in the end we do know US intel relies heavily on media reporting and OSINT of varying quality to generate these reports, and as this article points out, the product is often flawed or even figmentary. So I only use those figures as another guidepost.
https://ridl.io/lies-damn-lies-and-s...lly-recruited/
Here are the figures I use that I think fit best with the totality of the evidence:
In December/January I entertained higher estimates of Wagner convict recruitment, such as 65-70K, but by now I've settled on 60K, with the last 6-month contracts signed in January.
Over the course of 2022, a total of 70-75K
unique individuals served in (we could say "passed through") Wagner. 60K of these were convicts, 10-15K pros (Wagner has been recruiting heavily throughout the war). It should be clear that this does not mean that Wagner's available strength at any
given time was 75K. I estimate 12-15K of the convicts were already 'irrecoverable' casualties consequent particularly to the fighting in October-December, along with 3K pros. Thus for 2022 I estimated at least 15K irrecoverable Wagner casualties (including desertions but excluding pros who declined to reup their contracts, who are discussed later).
I have always used a heuristic of two irrecoverable WIA for every KIA with RuFOR losses btw. Assessments from the US and many other sources use a 4x multiplier, which I halve to crudely exclude those returned to duty after recuperation. There may be a better argument for a 1.5x multiplier, the consequences of which I explore briefly below, but that's a matter for another day.
Based on the February US estimate of 9K Wagner KIA in total, a contemporaneous Russian media investigation documenting through paperwork a floor of 8K Wagner dead, and ground-level reports from both belligerents, I figure that on average in January-April there were 2.5K convict KIA and 0.5K pro KIA per month, for a total of 10+2K KIA and 20+4K WIA.
Total Wagner losses: Convict: 14-15K KIA; 28-30K WIA
Pro: 3K KIA; 6K WIA
KIA: 17-18K
WIA: 34-36K
SUM: 51-4K
Round to 55K to include short-term pros passing through the group, as I assume an outflow of 10-33% of pros per year of combat (Wagner contracts are usually 3 or 6 months and evidently more flexible than those with the MoD). Captures are negligible.
Subtraction leaves us with a range of 15-20K Wagner in Ukraine. So how did I get 10K at Bakhmut? Well, it is known that Prigozhin and/or Russian higher-ups like to keep some Wagner detachments scattered throughout the theater of war as security forces and perhaps a strategic reserve. For example, IIRC two Wagner battalions (500 men?) were dispatched to help stabilize the Luhansk front in September. Moreover, some proportion of Wagner forces will be involved in garrison and support duties, including Wagner's independent artillery branch. Wagner does predominantly receive its logistical and artillery support from the MTO and SVRF respectively, but it is known to have its internal capabilities.
So I rounded it down to 10K "bayonet" strength at Bakhmut.
An interesting question is whether the new information in Prigozhin's figure of 26K completed convict contracts includes those with significant, or even disabling, injuries. The output is affected quite a lot depending on the answer. Obviously 26+55 is a lot more than a cumulative strength of 70. We might compromise by assuming that convicts past some level of injury serve as support, transport, or other miscellaneous personnel, but that's highly speculative. Would Wagner bother to keep around or invest in individuals who may need weeks of rehabilitation just to become available for heavy manual labor? In this case, 30-26=4K heavily injured convicts retained by Wagner.
At any rate, if the 26K figure is truthful, it makes it unlikely that Wagner recruited only 40K convicts between June and January, although it is possible if the figure does include all the seriously wounded, and the very smallest KIA estimates are used. We can pretty much rule out that a light/moderate wound is grounds for amicable release on these terms, as that would only be compatible with very small KIA or recruitment figures, given that the majority of all convicts would have passed the 6-month threshold by now even had none of them died (half a year ago was already November).
Separately, I realized we should also add a few thousand, maybe 5K, to the number of unique Wagnerites to account for 2023 recruitment. This includes their extensive public campaign as well as some number, likely between 1-10%, of convicts who decided to join as pros following completion of their term of service (these not properly added to cumulative unique individuals but to active strength at a point in time).
If we try again with 1.5x WIA ratio this time:
Total Wagner losses: Convict: 14-15K KIA; 21-22.5K WIA
Pro: 3K KIA; 4.5K WIA
KIA: 17-18K
WIA: 25.5-27K
SUM: 42.5-45K
If we subtract all convict WIA from the cumulative fulfillment figure, we get a range of 3.5-5K, which we add to the sum as representing a small set of convicts outlasting their term without grievous harm, for the new range of 46-50K, comfortably rounding up to 50K to fudge the short-termer pros. In this case though there would be 25-30K total Wagner in Ukraine as of now (70 or 75 + 5 - 50 = 25-30). For reference, Milley's late March range of 26-36K would transform to about 16-26K as of early May if applying my cumulative attrition numbers.
According to the foregoing too, Wagner has run through 2/3 of all its convicts, and no more than 10% of convicts can expect discharge without death or significant injury. Sounds about right.
What I take away from this is that despite the wide plausible ranges in which we can tweak most of the variables (total recruitment, KIA, WIA ratio), my baseline estimates of at least 60K convicts recruited, and around 15-20K Wagner remaining active in Ukraine, are at least plausible. What I have previously been most tentative about is my treatment of the pros; it is admittedly daunting to conclude that anywhere from a third (7.5/20) to 90% (9/10) of everyone who passed through the core of Wagner in Ukraine left as a casualty in the war so far. The latter should be dismissed; I would bet the true figure is between 25-50%.
I would update my current estimate to
20K total Wagner in Ukraine in light of the additional information presented by Prigozhin altogether. But it doesn't greatly affect my assessment of Wagner strength around Bakhmut, especially as one to two thousand are still being drained by the week to all causes.
Bookmarks