Skip to content
The ColumnColumn· No. 7205

COLUMN: 23,000 Russian Losses You Can See—And the Larger Count You Cannot

more than 23,000 A May 2026 account drawing on Oryx’s photo-and-video methodology put visually confirmed Russian equipment losses above 23,000 items. Oryx methodology The count is formidable precisely because it calls itself a minimum, while a video from the Kupyansk direction added a rare 2S44 Giatsint-K on 3 August. 3 August 2026 The task is to separate what is established from what is asserted.

Premium reading
MadMax
Key takeaways
  1. more than 23,000 A May 2026 account drawing on Oryx’s photo-and-video methodology put visually confirmed Russian equipment losses above 23,000 items. Oryx methodology The count is formidable precisely because it calls itself a minimum, while a video from the Kupyansk direction added a rare 2S44 Giatsint-K on 3 August. 3 August 2026 The task is to separate what is established from what is asserted.
  2. more than 23,000 A May 2026 account drawing on Oryx’s photo-and-video methodology put visually confirmed Russian equipment losses above 23,000 items.
  3. Oryx methodology The count is formidable precisely because it calls itself a minimum, while a video from the Kupyansk direction added a rare 2S44 Giatsint-K on 3 August.
Transparency

Facts, quotes, and cited links remain in the body. Interpretations are framed as analysis or opinion according to the format.

Introduction

more than 23,000 A May 2026 account drawing on Oryx’s photo-and-video methodology put visually confirmed Russian equipment losses above 23,000 items. Oryx methodology The count is formidable precisely because it calls itself a minimum, while a video from the Kupyansk direction added a rare 2S44 Giatsint-K on 3 August. 3 August 2026 The task is to separate what is established from what is asserted.

A verified minimum is not a weak number. It is a number that refuses to counterfeit certainty. 3 August 2026 A list of losses can be both devastating and deliberately incomplete without becoming useless. Its restraint is the reason the count can be trusted: it refuses to turn an absence of images into an absence of loss.

The number that can be defended

A May 2026 public total

Oryx methodology The Insider reported on 26 May 2026 that Oryx’s visual-evidence methodology had confirmed more than 23,000 Russian equipment losses. public visual evidence What is documented is narrower than the loudest interpretation, and that narrowness is useful. The Insider The threshold is more than 23,000.

more than 23,000 losses The figure derives its value from publicly available photos and videos linked to individual losses rather than from an uncheckable aggregate claim. 26 May 2026 The practical consequence is a question of timing, resilience, or decision-making rather than a theatrical verdict. Russian equipment The evidence is visual.

The count is above 23,000

4,390 tanks The total belongs to late May, not to the narrower early-August window cited elsewhere in this dossier. 6,429 infantry fighting vehicles The stated limitation is part of the evidence, not an inconvenience to be hidden after the conclusion. 730 APCs The date matters.

The first discipline is to count only what can be seen. 2,399 other armoured vehicles It establishes a large verified baseline while requiring the reader to carry its date and its method at the same time. 416 air-defence systems What remains is a sharper question about accountability: who has promised, counted, approved, or adapted—and what follows from that act. 130 radars A dated floor is more useful than an undated certainty.

Why Oryx calls it a minimum

The evidence rule excludes unseen losses

Oryx methodology Oryx records only losses supported by public visual evidence, so destroyed or captured equipment that is not filmed or published does not enter the list. public visual evidence The wording fixes a time frame and an actor; it does not supply every operational detail a reader might want. The Insider The exclusion is deliberate.

more than 23,000 losses That rule necessarily makes the verified total lower than the real total of Russian material losses. 26 May 2026 For Ukraine and its partners, the operational meaning depends on whether this can be sustained, supplied, or converted into action. Russian equipment Visibility is the admission test.

An incomplete count can still be rigorous

4,390 tanks The method cannot tell readers how many unfilmed losses exist, and it does not pretend to do so. 6,429 infantry fighting vehicles Nothing in the assigned record supplies a substitute measurement for what it leaves unknown. 730 APCs The unseen total is unknown.

A tank is not a statistic when the image fixes what happened to it. 2,399 other armoured vehicles Its discipline creates a defensible floor: every confirmed item clears a proof threshold, while the unobserved remainder is left unclaimed. 416 air-defence systems The next test is whether the documented change holds when it meets the next constraint in the chain. 130 radars Restraint is what lets the visible count carry weight.

4,390 tanks

A category with direct combat significance

Oryx methodology The late-May Oryx-based count included 4,390 Russian tanks among the confirmed losses. public visual evidence Its value lies in the distinction between a stated capacity and a result independently audited from outside. The Insider The count is category-specific.

more than 23,000 losses Tanks form one category in a broader ledger that also lists infantry fighting vehicles, personnel carriers, and other armoured vehicles. 26 May 2026 In a war of attrition, a figure changes the argument only through the system that can repeatedly turn it into capability. Russian equipment The number is 4,390.

The tank number is not the full armoured picture

4,390 tanks The dossier does not assign a separate operational effect to each tank loss or calculate replacement rates. 6,429 infantry fighting vehicles That missing element cannot honestly be reconstructed from the headline figure alone. 730 APCs Replacement is not assessed.

Categories matter because an army does not lose capability in one undifferentiated pile. 2,399 other armoured vehicles The figure proves a visually documented burden in a core armoured category without pretending to describe every Russian armoured loss. 416 air-defence systems This is the consequence that travels beyond the statistic: it tells the reader where scrutiny belongs next. 130 radars The visible record makes material attrition concrete.

6,429 infantry fighting vehicles

A larger confirmed category

Oryx methodology Oryx’s tally listed 6,429 Russian infantry fighting vehicles, exceeding the confirmed tank count in the same summary. public visual evidence The record identifies a concrete development, not a licence to fill in the missing measurements. The Insider The item is an infantry fighting vehicle.

more than 23,000 losses The separate category matters because infantry fighting vehicles are not interchangeable with tanks in an army’s structure or tasking. 26 May 2026 The mechanism is important because it links the announced item to a real constraint rather than to a public-relations tableau. Russian equipment The tally is 6,429.

Mobility and protection are counted separately

4,390 tanks No source in this packet converts that number into a complete judgement about Russian unit readiness. 6,429 infantry fighting vehicles The source establishes one layer of the story and leaves another layer uncounted, unaudited, or undisclosed. 730 APCs Readiness is not calculated.

The blind spot is built into the method and openly admitted. 2,399 other armoured vehicles What it does show is that the confirmed loss ledger reaches deep into vehicles used to move and support troops, not only headline armour. 416 air-defence systems The record invites pressure for proof, schedules, methods, and outcomes rather than applause for the first number announced. 130 radars Disaggregation prevents a total from hiding the type of capability lost.

730 personnel carriers

A distinct vehicle class

Oryx methodology The same compilation recorded 730 armoured personnel carriers among confirmed Russian losses. public visual evidence Placed in sequence, the evidence shows movement without pretending to provide a complete balance sheet. The Insider The category is explicit.

more than 23,000 losses Keeping personnel carriers separate from infantry fighting vehicles preserves the categories used in the source rather than flattening them into one armoured number. 26 May 2026 This is where a single data point becomes strategically relevant: it identifies which part of the chain is being tested. Russian equipment The number is 730.

The total rejects one-size-fits-all accounting

4,390 tanks The record does not specify the circumstances of every confirmed personnel-carrier loss in this cumulative total. 6,429 infantry fighting vehicles Calling the uncertainty by its name prevents attribution from being laundered into certainty. 730 APCs Every event is not narrated here.

A radar loss tells a different story from an armoured vehicle loss. 2,399 other armoured vehicles The category strengthens the central point: Oryx’s ledger works item by item and class by class, which is why it is a floor rather than a speculation. 416 air-defence systems Its broader lesson is practical: strategic language becomes credible only when the underlying process can be examined. 130 radars The method values identification before grand totals.

2,399 other armoured vehicles

A residual category still needs a name

Oryx methodology Oryx counted 2,399 other armoured vehicles beyond the tank, infantry fighting vehicle, and personnel-carrier categories. public visual evidence This is a dated indication, which is stronger than rumour and weaker than a full technical audit. The Insider The subtotal is 2,399.

more than 23,000 losses That subdivision prevents the headline total from pretending that all armoured losses are identical in role or design. 26 May 2026 The material therefore describes a pressure point, not an automatic outcome on the battlefield. Russian equipment The category is residual but real.

The subtotal belongs inside the larger ledger

4,390 tanks The assigned account does not provide an exhaustive sub-list for this category. 6,429 infantry fighting vehicles The caveat has operational significance because decisions made from partial information can still carry a human cost. 730 APCs A complete subtype list is absent.

Aircraft and helicopters are separate burdens on a military system. 2,399 other armoured vehicles Its inclusion shows the breadth of the documented record while preserving the limits of what the summary actually breaks down. 416 air-defence systems The measurable item has a political afterlife, because it can be used to demand that institutions account for the gap it reveals. 130 radars The ledger’s granularity is part of its credibility.

416 air-defence systems

A loss category with its own consequences

Oryx methodology The Oryx-based summary included 416 Russian air-defence systems among visually confirmed losses. public visual evidence The available material supports the stated point; it does not prove every implication sometimes attached to it. The Insider The category is air defence.

more than 23,000 losses Air-defence systems differ materially from armoured vehicles, so the category signals that the ledger reaches beyond ground manoeuvre equipment. 26 May 2026 Its consequence is concrete but conditional, because adversaries adapt and because delivery, access, and use are separate steps. Russian equipment The number is 416.

A verified item is not a complete air-defence map

4,390 tanks The material here does not establish the wider performance or current disposition of Russia’s remaining air defences. 6,429 infantry fighting vehicles An allegation, an official statement, and an independently confirmed event occupy different places in the public record. 730 APCs The surviving inventory is not provided.

The sea count is small beside the ground total, not irrelevant. 2,399 other armoured vehicles The count confirms a substantial documented category of loss without licensing claims about complete control of the air or a final balance. 416 air-defence systems That is why precision is not passive. It identifies the point at which a promise becomes a responsibility. 130 radars The limitation keeps the conclusion operationally honest.

130 radars

Sensors are counted apart from launchers

Oryx methodology The tally also included 130 radars, separating sensor losses from air-defence-system losses. public visual evidence That is a measurable change in the public record, even where the wider battlefield effect remains unmeasured. The Insider The item is a radar.

more than 23,000 losses That separation is methodologically important because a radar is not the same piece of equipment as the wider system with which it may operate. 26 May 2026 The number matters most when read as a problem of throughput, verification, or endurance—not as a stand-alone trophy. Russian equipment The total is 130.

The method makes a capability distinction

4,390 tanks The summary does not describe the function or location of each radar. 6,429 infantry fighting vehicles This boundary does not erase the reported development; it defines the credibility of the sentence that reports it. 730 APCs Location is not summarised.

A rare artillery loss can matter more than its place in a large ledger suggests. 2,399 other armoured vehicles By preserving the category, the record demonstrates why visual documentation can offer more than a single dramatic total: it maps types of material loss. 416 air-defence systems The fact does not need embellishment. Its real consequence is to make an unverified shortcut harder to sell. 130 radars Category discipline makes later analysis possible without overclaiming.

184 aircraft

A verified aviation count

184 aircraft The late-May count recorded 184 Russian aircraft as confirmed losses. 175 helicopters The detail matters precisely because it makes the claim more specific, not because it makes it absolute. 33 vessels or submarines Aircraft form their own class.

minimum count This is a visually documented cumulative category, not a claim that every aircraft loss since the invasion has been captured by Oryx. not exhaustive An accurate reading asks what the datum changes in the chain of decisions, rather than treating it as a verdict by itself. Oryx The number is 184.

The date prevents inflation

2S44 Giatsint-K The figure should not be silently updated to August without new item-level evidence. 3 August 2026 The file gives no basis for pretending that a partial indicator is a final total. Kupyansk direction The cut-off is May.

A video is evidence of one event, not proof of every unseen event. second confirmed loss Its force lies in the documented number available at its stated date, not in an imagined current total. Yuriy Butusov The strongest conclusion is therefore the most specific one: this development changes the next decision, not every decision after it. video A firm date is part of the proof, not a technicality.

175 helicopters

A separate aviation burden

184 aircraft Oryx’s compilation listed 175 Russian helicopters separately from the 184 aircraft. 175 helicopters No public source supplied here turns that development into a universal rule for the whole war. 33 vessels or submarines Helicopters are named separately.

minimum count The separation matters because the source treats different platforms as different confirmed equipment categories. not exhaustive That framing keeps the focus on capability and responsibility, where an announced number either becomes real or does not. Oryx The total is 175.

The count does not merge platforms

2S44 Giatsint-K The dossier gives no claim about the complete present-day Russian rotary-wing inventory. 3 August 2026 The appropriate qualifier is not weakness. It is the discipline that lets the verified portion remain usable. Kupyansk direction The remaining fleet is unknown here.

Oryx’s restraint is a method, not a rhetorical pose. second confirmed loss The tally is evidence of a documented burden on Russian aviation assets, with the same minimum-count limitation that applies across Oryx’s work. Yuriy Butusov For a reader, the useful question is not whether the headline feels large, but whether the chain behind it can bear weight. video The category refuses the easy shortcut of one aviation number.

33 vessels or submarines

A maritime count inside a land war

184 aircraft The reported Oryx tally included 33 Russian vessels or submarines. 175 helicopters The evidence permits a limited conclusion, and the limit protects the conclusion from collapse. 33 vessels or submarines The count includes vessels and submarines.

minimum count The number is far smaller than the vehicle totals, but it records a distinct maritime dimension of Russia’s material losses. not exhaustive The resulting insight is modest but consequential: public commitments and public evidence do not all have the same weight. Oryx The total is 33.

Small does not mean negligible

2S44 Giatsint-K The supplied summary does not evaluate the strategic condition of the whole Russian fleet. 3 August 2026 Where the method is not published, the calculation cannot be independently reproduced from the material provided. Kupyansk direction Fleet-wide effects are not quantified.

The count is old enough to need its date carried with it. second confirmed loss Its value is categorical: the visual ledger can establish verified maritime losses without pretending to resolve every naval question. Yuriy Butusov The evidence leaves a clear civic task: insist that claims of scale remain attached to the proof that supports them. video The category stops the land-war total from becoming the whole story.

The rare Giatsint-K near Kupyansk

A video-recorded loss on 3 August

184 aircraft Video filmed by journalist Yuriy Butusov documented the destruction of a 2S44 Giatsint-K self-propelled gun in the Kupyansk direction on 3 August 2026. 175 helicopters It is possible to recognise the direction of travel while refusing to invent a final destination. 33 vessels or submarines The weapon is a 2S44 Giatsint-K.

minimum count The dossier describes it as only the second confirmed Giatsint-K destroyed since 2022, making it qualitatively notable even beside much larger categories. not exhaustive The point is not to drain the fact of force; it is to identify the exact force the fact can legitimately exert. Oryx The date is 3 August.

Only the second confirmed since 2022

2S44 Giatsint-K A single video proves that specific loss; it does not provide a refreshed total for every Russian artillery system. 3 August 2026 The same caution applies to any claim about scale beyond the period, location, or category described. Kupyansk direction The wider artillery total is absent.

Russia’s material losses are vast; the verified ledger is still a floor. second confirmed loss The incident demonstrates how the cumulative ledger grows: individual, visible events add proof without erasing the difference between a case and a complete count. Yuriy Butusov This is not a call for false symmetry. It is a call to make the documented side of the story do its full work. video Rarity makes the item noteworthy, not magically representative.

Proof before the total

A method that keeps its limits

184 aircraft The most important Oryx rule is simple: no visual confirmation, no entry. 175 helicopters The fact holds on its own terms; the next step is to resist making it carry more than those terms allow. 33 vessels or submarines The rule is proof first.

minimum count That discipline leaves the real Russian loss total necessarily higher than the published list, but it prevents the list from becoming an exercise in assertion. not exhaustive Every responsible conclusion begins with that separation between an observed development and an assumed result. Oryx The total remains a minimum.

The public ledger has a moral advantage

2S44 Giatsint-K The early-August Giatsint-K footage is a fresh, documentable addition; the 23,000-plus number remains a late-May minimum. 3 August 2026 The uncertainty should remain visible, because it is the difference between analysis and advocacy by assertion. Kupyansk direction Precision has a date.

The proper response to a minimum is not to guess the maximum. second confirmed loss The correct reading is severe and exact: Russia’s verified material losses are immense, and the actual total is higher without being calculable from this file. Yuriy Butusov The accountable reading is the durable one, because it remains standing when the unmeasured parts are challenged. video A count that declares its blind spots is harder to manipulate.

Conclusion

23,000 confirmed losses Oryx’s count does not give the whole scale of Russian equipment losses; it gives the defensible part, itemised through public evidence. The late-May total and the August Giatsint-K video belong to different dates but the same rule. visual evidence The integrity of the ledger lies in what it refuses to invent about the unseen losses. minimum The record is stronger when its limits remain in the sentence.

The images are not the whole war. They are the part that can be named and defended. Russia Russia’s losses do not become less grave because Oryx counts conservatively. The conservative method is why the visible record remains so hard to evade. Giatsint-K Its restraint is the reason the count can be trusted: it refuses to turn an absence of images into an absence of loss.

Get the geopolitics analyses

Conflicts, powers, alliances: the MadMax thread without the noise.

Cite this article

Maxime Marquette (2026). COLUMN: 23,000 Russian Losses You Can See—And the Larger Count You Cannot. MadMax. https://mad-max.co/en/article/23-000-russian-losses-you-can-see-and-the-larger-count-you-cannot

How does this piece make you feel?
MM
Maxime Marquette
Independent columnist

Maxime Marquette writes most of the analyses and columns published on MadMax — geopolitics, technology, and current events, no filler.

The Newsletter

Enjoyed this piece? Get the next one.

One chronicle a week, straight to your inbox. No noise.

Comments

0 / 2000

Be the first to weigh in.

This article was generated with AI assistance, under human supervision.

Column280 reads3298 words19 min read