From the moment that Vladimir Putin launched this illegal, unprovoked invasion, open source intelligence from hundreds of sources has been vital for both reporters and the general public when it comes to understanding events in Ukraine. That applies to the politics within each nation, the shuffling of roles within the militaries, and the vital behind-the-scenes logistics.
There are many different ways in which open source intelligence, known as OSINT, has helped to create a measuring stick for the relative strength and success of Russian invaders and Ukrainian defenders. Often that measurement has come in the form of maps and the calculation of distances as the two militaries vie over control of territory, and that may well be the most effective measure. However, those numbers have changed rapidly on occasion.
The most common and most familiar measure of day-to-day action may be the one that is posted by the General Staff of the Ukrainian military. Their summary of Russian losses gives an estimate of Russian equipment and personnel going back to the start of the invasion on Feb. 24, 2022. It also breaks out the losses for the latest day. These numbers get repeated widely on social media, and (with some caveats) on Daily Kos. But how well do they measure real Russian losses, and how much does that matter?
The second-best-known set of values when it comes to how much Russia has lost in this conflict is the OSINT site Oryx. The founder of the site put down his metaphorical pen on Oct. 1, and for months in advance of that date it seemed that this would be the end of updates. However, so far other analysts have continued to maintain Oryx’s list, so it’s (hopefully) up to date.
The numbers from Oryx are documented losses. These are pieces of equipment that have been imaged, usually in videos, and can be confirmed as newly reported losses. The Oryx team does its best to eliminate instances in which the same loss is recorded from multiple angles, by different forms of imaging, or when older equipment is revisited days or weeks later. The Oryx numbers will not have everything. Not all losses are imaged, and not all images are of sufficient quality to ensure that they represent new losses. The numbers from this source are minimum losses, and even that is probably going too far.
On the other hand, it’s understood that the numbers provided by the Ukrainian military are going to be on the high end. They’re not the maximum that could be estimated, but in many cases they are compilations of claims from individual soldiers and units. Those losses may falsely record the same piece of equipment. They may log the same loss twice because some of those pieces of equipment involved came off the line at different times. It’s not necessarily that the General Staff is deliberately overestimating Russian losses, but they’re also not holding up every reported loss to a cross-examination.
Comparing the two sets of numbers can also be a bit difficult because of categorization. For both sources, a tank is a tank, but beyond that it gets slippery. What’s called an armored personnel vehicle in the daily military report is likely to overlap the categories of armored fighting vehicles, infantry fighting vehicles, and armored personnel carriers at Oryx. What shows up as generically labeled artillery systems at General Staff probably includes artillery support vehicles, towed artillery, self-propelled artillery, and self-propelled anti-tank missile systems at Oryx. But even when you’ve made these kinds of mappings, there are a lot of categories that are difficult to determine. Does that broad “armored” category in the military estimate also include engineering vehicles? Are radar installations considered part of anti-aircraft weapons?
Using my best guesses, this is how things look at the moment:
|Oryx||General Staff||% Verified|
In the thankfully easy category of tanks, Ukrainian military estimates exceed those recorded by Oryx by 3,756. Or, from a different perspective, Oryx records 47% of the military’s estimated losses as confirmed. The armored vehicles category gives a similar result, with 44% of all reported losses being confirmed by Oryx.
These seem like entirely reasonable values. In fact, having nearly half of all lost equipment be converted into confirmed losses seems very good. It makes the numbers reported by the Ukrainian military seem as if they’re pretty darn realistic.
Things are uglier when it comes to artillery and MLRS. Even taking a generous view of how to map categories, just 15% of estimated artillery losses have been converted into confirmed images. MLRS systems are better, at 36% confirmed, but still well below the big front-line equipment.
But of course, tanks and armored vehicles are just that: big and front-line. That makes it easier to confirm their loss. A piece of towed artillery parked a dozen kilometers from the front is unlikely to get the same level of attention. But then, much of the artillery lost in this conflict has either been taken down directly by FPV drones, or located by surveillance drones. That would seem to suggest that there should be imagery. That only 15% of those guns are ending up as confirmed losses is worrisome, especially considering how important artillery has been both in Russian advances and in locations where Russia has held against the Ukrainian counteroffensive.
One item that Oryx doesn’t track is that most vital of military resources: people. Over the weekend, U.K. Defence Intelligence gave its most recent estimate of Russian losses. The numbers they posted (150,000-190,000 “permanent casualties” and 240,000-290,000 “temporarily wounded”) fall below the 294,700 estimated by the Ukrainian military, but not by much. Since the U.K. numbers do not include Wagner losses around Bakhmut, which have been estimated at 20,000, the numbers look remarkably good.
Using the maximum U.K. estimate plus the Wagner numbers, that would be 210,000 Russians killed or wounded to an extent that they are unlikely to ever return to combat. That’s 71% of the number that Ukraine provides, and Ukraine’s number includes at least some wounded. Overall, this seems like a strong expression of confidence in the values provided by Ukraine.
It’s easy to throw out the values that the Ukrainian military has provided, especially when it comes to places like Avdiivka, where they reported an astonishing 120 armored vehicles and as many as 45 tanks lost in a day. But the history of Ukraine’s reporting doesn’t suggest that these are gross exaggerations. It suggests that as more information comes in, many of these values will look very much in the ballpark, especially when it comes to personnel, tanks, and armored vehicles.
But it would be interesting to know how Ukraine is calculating its estimates of artillery losses. Maybe we’ll just ask them.
Does any of this matter? Yes, it matters. Because irrespective of what number or estimate you use, it’s clear that Russia can’t produce or repair some of these systems at the rate they are being lost. If conditions hold true, there will be a point where Russia can no longer proactively deploy equipment to attack or defend Ukrainian positions. That time isn’t here yet, but at least we have a pretty good clock.
Speaking of the daily report from the Ukrainian General Staff:
The latest situation update indicates that Ukrainian forces beat back three attacks near Ivanivka, east of Kupyansk. Another attack took place near Nadiya, west of Svatove. Another 10 attacks took place along the line from Bakhmut south to Avdiivka. Then another 10 attacks were carried out at both Avdiivka and Marinka. After two weeks of attempted Russian attacks involving heavy losses, there remains a high level of activity.
More amazing footage out of Avdiivka. Seriously, it’s not so hard to believe those losses.
Some Telegram accounts are saying that the trucks in this failed wave at Avdiivka are ancient GAZ-AA, though they may just be more of the same Urals that have been beaten up through the whole war. In any case, a transport truck column seems like a strange way to follow up two weeks of blasted-apart armor.
Finally, after sending tanks, then armored vehicles, then a wild array of trucks, could this be Russia’s next big means of assault?
Have they tried bicycles?