Maps and stuff

Tsimbulova Drone launching site August 2023-2024-2025-2026

Drones season openning

Where it goes next The first problem is finishing the Zherebets line and making it useful. Reaching the river is already partly done, but stopping there leaves AFU on exposed high ground with the valley settlements below and russian positions still close enough to contest them. The useful change comes from denying Russia those settlements and pushing the immediate fight onto the eastern side of the valley. Torske and Makiivka matter because they sit on the obvious crossing axes and give something the bare riverbank does not: a place from which movement across the river can actually be sustained. The second problem is south of Lyman. Yampil does not need to become a clean Ukrainian possession to matter; keeping it contested is enough to interfere with the 25th CAA's movement through the Donets valley and put pressure on the 3rd CAA's Sloviansk approach. That whole forested strip is infiltration terrain, so the useful result is denial rather than another neat line on the map. The point is to keep Russia from turning its presence south of Lyman into a stable flank while the northern side is being pushed back. The third problem is Borova. The 20th CAA may be the weakest part of the russian chain here, but that does not make the northern flank safe. Russian positions are still close enough that any Ukrainian movement east has to be balanced against the risk of pressure back toward Borova. Retaking ground around Vyshneve and pushing the threat farther from the town would widen that margin and make the northern side of the operation much harder to disrupt. Those three problems are linked. Progress across the Zherebets pressures the 20th and 25th from the west; pressure around Yampil complicates the 25th and 3rd from the south; securing Borova prevents the 1st GTA or the 20th from turning the northern flank into the distraction Russia needs. None of them is decisive alone. Together they are what would turn a local counterattack north of Lyman into a change in the geometry of the whole operation. What Russia does about it The counterweight is the 3rd CAA at Mykolaivka, but Mykolaivka is not a gateway into Sloviansk in the conventional sense. It is almost a dead end. Its value is that it would put a durable Russian position immediately outside the city, with the power-plant complex and water features protecting one side and narrowing the ground Ukraine has to cover around it. From there the 3rd does not need a clean breakthrough. It can keep pushing small groups toward Sloviansk, infiltrate the city, expand the contested belt around the eastern approaches and support operations around the river from much closer staging areas. Mykolaivka would therefore matter less as a road west than as a protected pressure point: every kilometre Russia gains there increases the area Ukraine has to deny and makes further infiltration easier. That is how progress by the 3rd CAA can compensate for failure farther north. It does not solve the 25th's Lyman problem, but it forces Ukraine to keep spending attention and forces on the opposite side of the same operation. North of Lyman the area barely changed for months; almost all of the deterioration came in the south, driven by the 3rd CAA pushing the contested band along the Donets valley well west. The 20th and 25th stalled while the 3rd kept creeping, and one formation's progress covered the other two's failure — the trick that kept this operation funded for more than a year. Nothing is decided yet. Ukraine can break the geometry north of Lyman while the 3rd CAA keeps rebuilding pressure from the south. Every gain against the 20th and 25th therefore has to be measured against what the 3rd can establish around Mykolaivka and the eastern approaches to Sloviansk. Success on one side changes the burden on the other but doesn't erase it.

Lyman and Repeat Nothing has been announced, and the maps are split the way they've been split all campaign: DeepState still has it red alongside RIA, while Suriyak has the whole area cleared. It's time to ask what comes next, where, and why. It's the fractal siege again, and the chain is easy to name. The 20th CAA is the weak link and has been since it failed to close the Borova pocket. The 25th CAA, which owns Lyman axis, has spent the campaign waiting on the 3rd CAA to make its position mean something from the south. The 1st GTA is busy at Kupiansk and may not come. Everyone leans on the neighbour; nobody solves the problem. What can actually be held Start with what the river isn't. The Zherebets is no major water barrier, but neither is it nothing: infantry can cross it in places, while vehicles and logistics still get channelled into usable routes. The more important feature is the ground above it. The block between the Oskil and the Zherebets isn't a smooth ridge but a carved one, cut by ravines running down toward the valley. Below it sit the villages, the most durable positions here. Built-up ground gives shelter; bare ground doesn't. What terrain still does is channel. The ravines funnel foot traffic into predictable draws, the crossings do the same at the river, and predictable routes get hunted. Infantry moves through almost any of it given shelter at both ends; without that, the route is the whole problem. So the important relationship isn't simply high ground plus shelter; it's the asymmetry created by the river between them. For Russia, the villages in the valley are the useful footholds — places from which men and supplies can be pushed toward the higher ground. For Ukraine, the same villages work almost in reverse. Denying them creates a buffer in front of the plateau, keeping russian infantry away from the ground that matters and leaving the exposed valley in front of Ukrainian drone positions. The high ground then reinforces control of the settlements below, while the settlements deny Russia the proximity it would need to contest the high ground in the first place. Behind the valley the ground empties out and stays empty until Р-66, the Kreminna–Svatove highway feeding this whole flank. That's what Russia can't fix. Its supply spine runs a long way back, and everything forward of it crosses relatively open ground with no village to shelter in. There may be prepared positions out there, but an open field is still an open field. What is actually being won What matters isn't the river or the high ground in isolation, but the depth between them. Ukrainian control of the upland puts the valley under observation and drone pressure; denying Russia the villages below keeps its infantry from establishing the footholds needed to contest that ground. The villages matter less as prizes in themselves than as a buffer. Lose them and Russia gets closer, deny them and the exposed valley becomes the problem Russia has to cross. The larger consequence is what it does to the shape of the offensive. Lyman was meant to be the northern claw of the Kramatorsk encirclement, built by the 20th, the 25th and the 3rd together, and it never closed. If Russia loses control of that high ground, the claw stops being merely stalled — its current geometry stops working, and Russia is left approaching Kramatorsk frontally from Pokrovsk and Kostiantynivka, against a defence that can concentrate instead of facing pressure from two directions.

Flamingo stopped fail-maxxing and finally hit something within its reachable zone.

The Orange Geography

Russian reported 6 month progress. Sep 2024 → Mar 2025: +3 191.0 km² Mar 2025 → Sep 2025: +2 833.7 km² Sep 2025 → Mar 2026: +3 507.7 km² Mar 2026 → Sep 2026: +3 286.5 km² Fake it till you make it. One of the reasons behind all the Russian claims is probably Putin's winning rhetoric — Russia needs to claim more to strengthen its position at the negotiating table.

GSUA 1,5 lemons It's potential 500K Mediazona pixels, that will take years to count. All the numbers have their inflation coefficient. Based on USF statistics Only 30% of bikes, 20% of vehicles, ~10% of APC are destroyed, tanks and artillery below 5%.

The 20th, 25th, and 3rd CAA shouldn't really be analyzed separately — their combined job was building the northern claw of the Kramatorsk offensive. Ukraine didn't need to break that plan outright, just stop the armies from spreading outward into that claw and push them inward, toward Sloviansk instead. That's exactly what happened: 63rd Brigade and the units around it stopped the 25th CAA on the Lyman axis, 60th Brigade and its neighbors stopped the 20th CAA, and 3rd Assault Brigade started working down their flank once the corps had the brigades in place to do it. The 25th CAA's ability to turn its Zarichne breakthrough into further operational progress had already begun stalling almost a year ago — probably the right moment for russian command to reconsider the whole geometry. Instead, progress by the 3rd CAA farther south helped preserve the impression that the northern operation was still moving. One formation's apparent success concealed another's failure, and the investment continued. The clearest tell of what happened next wasn't an arrow on a map — it was vehicles. The moment AFU could start moving vehicles through the area, the battle was basically over; vehicles didn't win it, they showed up because the conditions that had made vehicle movement impossible were already gone. There will be plenty of threads on the tactics, but none of it will look different from Novopavlivka: drones isolate russian positions and cut movement between them while also keeping Ukrainian units supplied, and small assault groups clear what's left on foot, position by position. It takes time, but that doesn't mean there's still a coherent russian defense out there. Russian soldiers can stay scattered across the area for weeks — some hiding, some trying to move back, some waiting for reinforcement that never comes — and none of that means the russian army still controls the ground. AFU does the same thing during its own withdrawals: rear guards stay behind to defend and buy time, and nobody treats those remaining Ukrainian groups as proof that Ukraine captured the territory it just left. A unit's presence alone doesn't tell you who controls the area. Russia putting itself inside a deep, narrow pocket only made the final phase easier. Once AFU had enough control over the surrounding movement, every new russian mission sent into the area walked into a reinforced perception trap — troops committed to hold positions that existed far more convincingly in reports than as a functioning front. Everyone likes tactical arrows because they show action, but the arrows almost always show up near the end. The vehicle arrives when there's little left to stop it; the assault group clears a position after drones have spent weeks isolating it; the map turns blue after the formation that was supposed to be holding it has already lost the ability to function as one. The visible operation is usually just the collection of a result produced much earlier.

The Bill Comes Due GSUA — reports Lyman-area clashes every day. Сounteroffensive wasn't officially rported yet, but the hype train already left the station regardless, so here's my take. The operation probably started around the same time as Novopavlivka, and I don't expect the mechanics to look much different. The real difference is that 3rd Army Corps had a much better starting position. Almost a year ago I wrote that Lyman itself wasn't the real question — Russia needed the flanks secured, the Drobysheve-Yarova line taken, the supply routes cut, before its infiltrations inside the town could mean anything. Without that outer ring, people showing up on the map were just people showing up on the map. That outer ring never happened. AFU denied russia the Nitrius valley and the advance near Kopanky, and unable to secure the high ground there, russian forces got pushed onto the much less comfortable Ivanivka-Oleksandrivka axis instead. They never established durable control over Lyman or Yarova — small groups reached both at different points, and infiltrators pushed as far as Sviatohirsk at one point, which generated exactly the headlines you'd expect, but the operational situation behind those headlines never followed. This has effectively been a 20-month operation. Ukrainian forces spent that stretch building one of the biggest bunny-ears situations of the war: russian forces pushed in deep enough that abandoning the position became politically and institutionally difficult, but never wide enough to actually develop it. Too much invested to leave, not enough achieved to continue — a genuinely terrible place to be, made worse once reporting became part of the battlefield itself. Every infiltration got presented as another sign of Ukrainian deterioration, every flag and every stray group farther west read as evidence the collapse was coming, while the actual problems — the missing flanks, the stalled neighboring formations, the inability to turn infiltration into control — got far less attention. That built its own feedback loop: reported progress justified more resources, more resources created pressure to report more progress, and admitting the whole thing had stalled got harder with every month and every reinforcement poured in. Peak fake-it-till-you-make-it, except they never made it. Now the bill comes due — see Sviatohirsk this week: Putin claimed it fell on September 1, and six days later Biletsky filmed himself standing in the town center to point out the 60th Brigade holds it as a rear area. Same pattern that's been running under this whole operation, just compressed into a week. OPSEC was explicitly useful as a weapon here. Every russian infiltration was visible and got amplified, while Ukrainian actions were much harder to read — so the defensive picture looked worse than the underlying balance actually was, even after 3rd Assault Brigade had already shown what it could do against the 20th CAA well before the corps around it existed. Put simply: exaggerate your own advance long enough, and eventually you have to defend the territory you claimed to have taken. The 2025 Siversk disaster set my expectations for this sector low, and at the time it genuinely looked bad for Ukraine. But the longer the campaign ran, the more obvious it became that russia had made a mistake.

Mass missile attack on Kyiv and oblasts, also Odesa was attecked by drones. Attempted to hit 5th THP, destroyed some fuel stations and warehouses but nothing exceptional. 142/166 — Drones 31/32 — Kh-101 / Iskander-K 2/? — Ballistic 29 locations were hit, with 9 more damaged by debris. More than half of the 166 drones were reported as jet-powered.

Russian YoY progress in Donetsk Oblast Sep 2023 → Sep 2024: +1 243.3 km² Sep 2024 → Sep 2025: +3 958.3 km² Sep 2025 → Sep 2026: +1 370.5 km² Based on DeepState, out of Donetsk Oblast's total area of 26 328 km²: 21 093 km² painted red 563 km² painted grey population of an occupied area: 507 741

Rubikon vs Magyar Birds — August 2026. Rubikon: 17 485 targets across 124 720 combat sorties. Birds: 16 712 hits across 28 770 strike missions.

Rubikon vs Magyar’s Birds: August Target Comparison Birds are 4 times more efficient than Rubikon but Birds had +10 279 recon missions, hence efficiency ratio is closer to 3

Second ring of Saratov

Ceasefire glitch

Russian YoY net expansion 2023 Sep 5 → 2024 Sep 4 +1869.3 km² 2024 Sep 4 → 2025 Sep 5 +5213.6 km² 2025 Sep 5 → 2026 Sep 5 +2934.0 km²

Rubikon vs USF - drone usage

Witkoff-Kushner 4 day ceasefire layout 5500km2 of the "grey" zone

Drone reach zones. 0–190 km Mid-range strikes 190–390 km FP2 range 390–1010 km FP1 100kg range 1010–2510 km max FP1 60 kg range FP5 can operate within green zone, but can barely cross the orange.