Ukraine Military History Institute
Russia is recruiting enough men to replace losses, and not one more
The single most useful number in this chapter is not a number at all. It is a ratio.
Rob Lee's assessment is that Russia is roughly meeting its manpower targets but no longer exceeding them, as it did in previous years. This year the attrition rate is essentially level with the recruitment rate.
That equation explains a great deal about why the front looks the way it does. Russia has spent the last few years trying to form new units and has not managed it. Last year, 80% of new recruits went straight in as combat replacements, filling the places of men killed or wounded. An army whose intake is consumed by replacement cannot build anything new.
It also explains the narrowing. Because of manpower, Russia has concentrated on fewer places rather than the whole front. It was pushed back in Zaporizhzhia, lost ground in the Dnipropetrovsk direction, and was pushed back on the Lyman direction while advancing toward Donetsk Oblast.
Mobilisation is the variable that would break the equation
Volodymyr Zelenskyy has said Russia is planning mobilisation, and Lee treats it as the first of several X factors rather than a certainty.
The mechanism matters more than the headline. Mobilisation need not mean an enormous intake in a single month. It might simply mean raising the monthly number on top of ordinary recruitment. What it would buy is the thing Russia currently cannot buy: enough men to form new units.
Lee is careful about what those units would be. They would not be well equipped and they would not have armour. But they would be enough to open a new front somewhere else. He notes there has been talk about Chernihiv for some time, and a possible refocus on Sumy and Kharkiv.
Nobody can say where the breaking point is, and that is the honest answer
The chapter's most quotable passage is also its most cautious.
There is a reason Putin has not mobilised again since 2022: political risk. But Lee will not convert that into a prediction. All the talk about hitting refineries, about Wildberries, about the rest of it, imposes real costs on Russia. Where the breaking point sits, no one can say.
Russia has demonstrated it can absorb very heavy costs. It is not clear at what point the country would say this is enough. Is mobilisation that point? Maybe. Maybe not. And if it is not, that is a big problem for Ukraine.
This is the correct register for a claim about a threshold nobody has observed, and it is worth noticing how rare it is. A great deal of commentary about pressure on Russia assumes the breaking point exists at a knowable place.
North Korea is no longer only a supplier
Zelenskyy has mentioned 30,000 to 50,000 North Korean troops being deployed to Russia, and Lee is direct that their purpose is not clear.
What is clearer is the direction of travel. North Korea is sending more ballistic missiles and launchers and appears to be taking a more direct role.
His assessment of the troops themselves is worth quoting for its lack of dismissiveness. The soldiers committed earlier fought only in Kursk. They were tough, capable soldiers who were not ready for FPV warfare, and they have learned since. If they cross the border and are used for assaults again, that is not insignificant, because these are real infantrymen. Even 10,000 added to Russia's total in Sumy would be a significant change, and Ukraine would probably have to reinforce that direction to counter it.
The assault that showed Russia learning
The last X factor is the one Lee seems most alert to, because it undermines an assumption Ukraine has been relying on.
Ukraine has held the line with very little infantry by using drones. That has worked partly because Russia has repeated its mistakes and has not adapted well at the tactical level. If Russia starts getting better tactically, the infantry shortage becomes a real problem.
The mechanised assault in the Azov direction was, in his description, pretty sophisticated. And the timing was deliberate: Russia did not attack in perfect weather, when Ukrainian surveillance could see everything. Run the same thing against a different corps in October or November, on a day with poor weather, with those innovations, and Lee thinks it might be enough to achieve a breach.
The part that is scarier than the troops
Francis Farrell adds a reading that is political rather than military, and it is the sharper observation.
The immediate threat from mobilisation is more intensity where Russia already attacks, or a new front. But the scarier thing is what mobilisation would prove: the political conviction to move past the clear reluctance and hesitation and commit to a fight that could run for years more.
Lee then notes the second-order problem. Once Russia mobilises again, it will be prepared in a way it was not in 2022. Border control is set up, e-registration is set up, and the process will be far more efficient. Samuel Cook calls it crossing the Rubicon, with Putin effectively saying he is in this for another three years.
Against that, Lee reports the countervailing pressure inside Russia. Senior officials have told Putin the war's objectives cannot be achieved and the costs are very heavy, and some foreign policy commentators have argued Russia should declare it has won and move on.
What to watch
- Whether mobilisation is announced this autumn. It is the one variable that changes the manpower equation, and Lee treats everything else as downstream of it.
- North Korean troops crossing the border for assaults. Their earlier deployment stayed in Kursk. Use of them inside Ukraine, particularly toward Sumy, is the observable change.
- A repeat of the Azov-direction assault in poor weather. Lee named the conditions precisely, which makes it falsifiable. If a similar assault lands against another corps in bad weather, the tactical-learning thesis is confirmed.