A Record October for Russian Losses, on Ukraine's Own Count

On 1 November 2026, Ukrainian Telegram channels carried a set of October figures attributed to Madyar, the name the source items use without a formal title. Units of the Unmanned Systems Forces (USF), according to those posts, hit an enemy target every 41 seconds on average and a Russian military target every 159 seconds, and the month was described as a record for Russian losses. One OSINT channel goes further, putting the Russian total at about 54,000 killed and wounded. Each figure is easy to repeat and hard to test, and none of the five items supplied for this story attaches a methodology to any of them.

This article treats the October figures for what the evidence supports: Ukrainian claims, relayed by channels that add little independent material of their own. Monexus analysis: a count of hits per interval suits a force whose effects are measured one engagement at a time. It reads more like a production line than a casualty list, which is why it travels well. A rate is only as useful as its definitions, and the source items supply few. The more productive question is not whether October was high, but what is being counted, and whether the same count will be published again.

One statement, five relays, and two totals

The most specific figures trace back to Madyar. In the posts reviewed for this article, he is reported to have said that October was a record month for Russian losses, that USF units hit a target every 41 seconds on average, and that a Russian occupier was struck every 159 seconds. WarTranslated carried the remarks in two posts dated 1 November 2026, and an osintlive repost in the same window reproduced the substance. The source items do not specify the wording of the original statement, so this article uses only the phrasing the channels themselves carry.

The five items are relays rather than separate evidence. The interval figures recur across them, and none supplies imagery, a unit-level log or a count from a second institution. The Ukrainska Pravda item is the only one in the set that carries the name of a national newspaper, and it adds a figure the others lack: a count of 64,807 enemy targets hit by the Unmanned Systems Forces group in October. The source items do not say how that count was compiled, so it adds a second Ukrainian number, not a check on the first.

The largest total comes from a separate osintlive post, which reports that “the Russian army lost about 54,000 killed and wounded in October, an absolute record since the full scale war began,” and pairs it with Madyar's claim that USF units alone inflicted 16,796 personnel losses. The thread does not show whether the 54,000 is Madyar's figure or the channel's own estimate, or how killed and wounded were counted. Monexus analysis: a combined killed-and-wounded category blends outcomes with different consequences, since a wounded soldier may return to duty and a killed one does not.

The 16,796 figure is where the items disagree. OSINTlive attributes it to Madyar. The Ukrainska Pravda item attributes the same number to losses of the Russian military and names Brovdi rather than Madyar. The source items do not establish how those two names relate, and they do not show which attribution is accurate. This article reports both readings and does not choose between them.

Hits are not kills, and the record needs a baseline

The strongest counter-reading starts with vocabulary. A hit on a target is not the same as a confirmed kill, a destroyed vehicle or a soldier removed from the fight. The cited posts use “hitting a target” and “personnel losses” without defining either term or describing how an engagement is confirmed. Monexus analysis: a record in claimed engagements can reflect more operations, tighter reporting or a wider definition of a target, rather than a change in outcomes at the front. That reading does not make the figures false. It makes them hard to compare across months.

The two intervals also use different units. A target every 41 seconds counts objects, while an occupier every 159 seconds counts people. The thread uses both without explaining whether one target can hold several occupiers, or whether the people figure is a subset of the target figure. The ratio between them suggests the two counts describe different things, and the items do not resolve the question.

The claim of an “absolute record since the full scale war began” needs a baseline the items do not provide. The cited items contain no monthly series against which October can be ranked. Verification searches for this article also returned outside headlines, not part of the five-item set, that give a lower October total and describe the month as the worst since the full-scale invasion. This article has not retrieved or verified those reports, and it does not rely on them. Monexus analysis: a gap between a lower total and the 54,000 in the osintlive relay may reflect different definitions, cut-off dates or sources, but the supplied items do not say which.

The Ukrainska Pravda count offers one arithmetic check. A 41-second average across the 31 days of October, if the interval held around the clock, works out to roughly 65,300 engagements by desk arithmetic. That is close to the 64,807 figure the same item reports. Monexus analysis: the closeness suggests the interval and the count were drawn from a similar base. The items do not state the method, and agreement between two Ukrainian figures is not independent verification.

The counting itself, in our assessment

Monexus analysis: when a war is described as a sequence of logged engagements, the logging itself becomes part of the contest. The thread's emphasis falls on USF units and on the number of targets they hit, which means the metric is built around the activity the force most wants to show. An operator who identifies, strikes and records a target produces a data point, and a unit that records thousands produces a rate. A rate is easier to publish than a body count, and it can be scheduled and reported monthly.

The same figures can be read in several directions. For Kyiv, a steady record in unit-level strikes is an argument for the force itself. For Moscow, the same numbers, if accepted, describe the cost of holding ground under constant observation. For Western governments that fund Ukraine's defence, the figures are a reason to read battlefield tallies closely before they inform procurement or judgments about whether pressure is working. Each reading is plausible, and each depends on counting rules the items leave out.

The format also rewards confident numbers. When a claim of record losses arrives with a precise interval attached, it travels faster than the definitions that would qualify it. This is not a claim that either side is fabricating. It is a claim that the format favours precision over method, and that readers who receive figures through channels relaying a single voice have few ways to separate the two.

What would change the reading

The stakes are concrete. If the rate holds into November, a single record becomes a trend, and analysts will need to explain it in terms of force structure rather than one month's luck. If the figures prove inflated by their definitions, the credibility of similar claims suffers, and that matters for audiences whose support rests partly on confidence in battlefield reporting. Both outcomes point to one test: whether comparable figures appear for later months, with definitions attached.

What remains uncertain is considerable. The items do not establish whether 54,000 and 16,796 measure the same thing, whether the unit-level intervals cover the whole month or selected days, whether the 64,807 count and the 41-second average describe the same set of engagements, or whom the 16,796 figure belongs to. Verification searches also returned a result describing a commander of Ukrainian drone forces, and a separate result tying a record to Madyar. Those may bear on whether Madyar and Brovdi are the same person. The thread itself does not say so, and this article does not resolve it. None of these gaps shows that the figures are wrong. Each is a reason to cite them as claims rather than as measured outcomes.

The desk's expectation, offered as a forecast rather than as guidance to readers, is that the November tallies from these channels will be the first serious test of whether October was an outlier or a baseline. If they arrive with their definitions attached, October's figures become checkable. If they arrive without them, the October numbers will remain a claim with precise intervals attached.

Desk note: Monexus has kept every October figure in attributed form, reported the 16,796 attribution conflict and the lower outside totals rather than resolving them, and has not treated the record as established, since the five items are Ukrainian channel relays with no independent count behind them.

Wire provenance

This editorial synthesis draws on the following public wire/social posts:

• https://t.me/osintlive/587621
• https://t.me/osintlive/587619
• https://t.me/wartranslated/19964
• https://t.me/wartranslated/19962
• https://t.me/ukrpravda_news/86042