Russians can lose a battle but unlikely to accept the lose without a gun at the temple.

As usual, the AFU should push the enemy into direct attacks, but recent Kupiansk experience has made those attacks more potent than they used to be; hence the strategy may change.

Sloviansk–Kramatorsk, with a diameter of 21 km, should be too hard to take to even attempt; the main strategy should be the Izium and Barvinkove claws.

Kostiantynivka is a dangerous place to hold but large enough to fight for. Druzhkivka will protect the routes, and the complex terrain to the west should help in defense, pushing the enemy further north.

The Pokrovsk–Myrnohrad node is large and protects a huge surrounding area; this has been explained multiple times. Dobropillia, which is supposed to hold the next line, sits aside the main road and can act as a significant obstacle, but remains a viable point on the way toward Barvinkove.

Grind allows focusing on high ground and using ravines for occasional jumps. Vectors of attack usually follow the watershed, with rare attempts to cross multiple streams in a single operation (hello Dobropillia).

Beyond dense settlements, terrain remains the main factor. Roads are important but rarely used for assaults. Fortifications may slow down the advance but don't affect the vector of attack.

Rashanverse has no time dimension; Frontline isn't an exception, hence all time-based predictions consistently fail. With this and the Russians' recent dynamics in mind, let's explore how the Donbas offensive will evolve.

They have production, they dont rely on armor anymore.

We are witnessing some shades of future warfare.

September in two pictures

SBS September

APCs almost disappeared from the stats
Tanks are 2% and APC - only 1% of all the losses.
"Loafs" and Bikes replaced them completely.

Number of Attacks remained stable, but new russian tactics left to much grey, hence reduce the DS confirmed gains.
Monthly attacks distribution looks similar to August, but weekly data show significant force redistribution towards South.

Russia was able to capture 251km2, grey out 194km2 - extended by 446 km2 in total.
Making only 28420 ivanZ in process.
Half of all their gains come from Novopavlivka direction in 836 attacks.
55km2 were filled in 1683 attacks at Pokrovsk.

GSUA Summary
Russians launched 5391 assaults at 228 settlements, 58 - never appeared in Septembers reports, 37 -appeared for the first time.

For the first time reported russian FPV usage surpassed artillery barrages, slowly replacing artillery.
Bombs hasn't increased but MLRS bounced back, apparently driven by some cooperation.

147480 Barrages
175416 FPVs
3994 Bombs
3175 MLRS

There were 5585 drones - 2nd highest monthly result
with the record 810 drones in a single attack.
There is no stable attack pattern, with two most recent mass attacks happened once a week.

Striking September

Settlements constellation

from small villages with less than 1000 people to 100 000+ cities.

Settlements follow the rivers.
Donbas is dominated by big stars, in the void of fields.
The Haichun river is a natural Donbas border, with a better spread population.

Comparisson of DeepState and MoD RF layers with relative alignment despite russian map is a distorted masterpiece.

ModRF published a map with captured 4714km2 in 2025.
DeepState shows ~3500km2 for that period.
4727km2 for a year period and match the shape of the reported front.
The initial frontline that matches russian has never exist.

t.me/mod_russia/5...

Don't let russian slowdown fool you,
they are probably going to regrouping, additional attacks needed to consolidate and secure gains. then they will decrease activity for a while but will return better prepared for another "round".

Stable positions are required on the way to a super-node.
bsky.app/profile/mons...

Bottom line: stable positions are the prize that converts tactical success into operational control. They're essential for attackers - but only if taken with an operational plan that accounts for logistics, consolidation, and the defender's ability to exact a high price.

Economies of force: prioritize targets within the enemy’s supply web - choke points, forward supply dumps, routes - that, when taken, make a node unaffordable to hold. This lets you convert limited resources into disproportionate operational leverage.

Decision rule for commanders: capture + consolidate = success. If you can’t both seize and immediately sustain a stable position, accept the operational cost of returning to a prior line rather than letting a hollow advance become expendable dead weight.

Footholds matter as stepwise risk management: seize a small, defensible foothold, quickly consolidate, and use it to secure supply - but only if follow-on forces and logistics are ready. A poorly supported foothold becomes a magnet for counterattack and a drain on combat power.

Operational trade-off: if you rush for the node you risk overextension. If you delay to isolate it, you concede initiative and time for the defender to adapt. The correct choice depends on tempo, relative logistics, and strategic timetable - not just raw local advantage.

Two main attacker methods: direct seizure (shock, breach, clear) or indirect neutralization (sever logistics, isolate). Direct seizure buys the position immediately but bleeds resources; indirect isolation is cheaper in losses but slower and leaves the objective in enemy hands longer.

But the price is high. Defenders will have prepared fields of fire, concealed fortifications, local knowledge, and layered logistics. A direct blow to a stable position often requires concentrated fires, engineers, and sustained logistics - the very things defenders hoped to force you to expend.