The most consequential question about Russia’s failures in Ukraine is not why they happened. That part is comparatively well covered, and the broad shape of it is established in the open record: the opening campaign of the 2022 full-scale invasion faltered on supply, on command, on intelligence, and on the coordination of arms that modern land warfare demands, and the force that had been advertised as reformed and professionalized did not perform as advertised. The harder and more decision-relevant question is the one that follows: which of those failures are still true. A defense planner does not benefit from a catalogue of an adversary’s past embarrassments. A planner benefits from knowing which of those embarrassments describe a permanent property of the force and which describe a moment that has since passed.

That distinction is the whole analytical problem, and it is routinely skipped. The failures of the opening campaign became, within weeks, a genre. They were narrated, mocked, and turned into a settled picture of an army that could not do the basic things. The picture was accurate about the campaign it described. It became progressively less accurate about the force as the war continued, because a force under existential pressure does not stay still. Some of what went wrong was fixed, at cost and imperfectly. Some of what went wrong was never fixable, because it grew out of how the institution is built rather than out of what it did on a given morning. And some of what went wrong belonged only to that specific campaign, with its specific assumptions, and would not recur in a different one.

This article is an adversary read, not a transfer exercise. It is about what the failures reveal regarding the force itself, held deliberately short of the question of what any of it means for a specific defender’s risk. That larger read-across is owned elsewhere in the series, and the discipline of keeping the two apart matters: an assessment that leaps from “Russia bungled the drive on Kyiv” straight to “therefore a NATO frontier state faces a manageable problem” has skipped the step where the failure is dated, weighed, and checked against what the force did afterward. The step that gets skipped is the step that decides whether the conclusion is worth anything.

The organizing device here is a ledger, not a list. Each failure gets sorted into one of three bins: structural, meaning it grows from durable institutional properties and is likely to persist; corrected, meaning the force demonstrably changed its behavior after paying for the lesson; and situational, meaning it belonged to the particular conditions of that campaign and would not necessarily reappear under different ones. The bins are not tidy, several failures sit on a boundary, and the ledger has to be read with its uncertainties visible. That is the point. A ledger that pretends to certainty about an adversary’s internal condition is worse than useless, because it invites confidence where the open record supports only judgment.

The Question a Capability Read Has to Answer

Capability analysis of an adversary tends to fail in one of two directions. The first is the accountant’s error: counting things. Tanks, brigades, tubes, launchers, aircraft, all tallied and compared, as though warfare resolved into arithmetic. The count matters, but the opening campaign in Ukraine is the strongest available demonstration that a favorable count does not produce a favorable outcome, because the counted force could not be delivered, supplied, coordinated, or commanded to the place where the count would have mattered. The second is the pundit’s error: character judgment. The force is declared incompetent, or hollow, or a paper tiger, and the judgment is treated as a property of the institution rather than as a description of an observed performance under specific conditions. That error is more comfortable than the accountant’s error, because it feels like insight, and it is more dangerous, because it produces complacency that survives the evidence that should overturn it.

The alternative is unglamorous. It requires taking each observed failure, asking what caused it, asking whether the cause is a fixture or a circumstance, and then asking what the force did about it afterward. That is slow work and it produces hedged answers. It is also the only method that yields a picture with a shelf life, because it separates the parts of the assessment that will still be true in some years from the parts that were already stale by the second year of the war.

What is the most decision-relevant question about Russia’s failures?

Not why the failures happened, but which ones still hold. A failure that has been corrected is history. A failure that grows from how the institution recruits, promotes, and reports is a property that persists. Sorting the two is what turns a catalogue of past errors into an assessment a planner can use.

The reason this matters is that failures have a half-life and assessments do not automatically respect it. An assessment written from the evidence of the first months of the full-scale invasion described a force that had been handed a plan built on assumptions that collapsed on contact, that had been denied the time and the information to prepare properly, and that was attempting a coup de main with a force posture better suited to a demonstration. The observed performance was genuinely poor. But the observation was of a specific interaction between a plan, a force, and a defender, and every one of those three variables changed afterward. Treating the observation as a permanent property of the middle variable is a category error, and it is the single most common mistake in the popular reading of this war.

The Difference Between a Failure and a Weakness

A failure is an event. A weakness is a property. The confusion between them runs through most of the commentary on this war, and untangling it is the first analytical move.

When a column of vehicles stalls on a road because the fuel trucks are not where the plan said they would be, that is a failure. Whether it is also a weakness depends on why the fuel trucks were not there. If they were not there because the plan concealed its own timing from the logisticians until the last hours, that is a failure caused by a deliberate choice about secrecy, and a different choice would produce a different result. If they were not there because the force does not maintain enough wheeled transport to sustain a mechanized advance beyond a certain distance from a railhead, that is a weakness, because it describes a structural property of how the force is built and supplied, and it will reappear whenever the same distance is attempted. The observable event is identical. The analytical meaning is opposite.

This is why the popular narration of the war misleads even when its facts are right. It records events with reasonable accuracy and then converts each event into a property without doing the intermediate work. The stalled column becomes “Russia cannot do logistics,” which is a property claim, drawn from an event, without the causal step that would license it. Some version of that property claim is defensible, as the ledger below argues, but it is defensible for a narrower and more specific reason than the event alone supplies, and the narrower version is the one that survives contact with what happened later.

The same test applies in reverse, to corrections. When a force that could not coordinate its fires with its reconnaissance begins doing so with tolerable reliability, that is an event, and it is evidence against the property claim that the force is incapable of it. Whether it is decisive evidence depends on whether the improvement came from a durable change to how the force trains and organizes, or from an ad hoc arrangement that works in a static positional fight and would not survive a return to maneuver. Again the event is the same. Again the meaning depends on the cause.

What Russia’s Failures in Ukraine Actually Were

Working from the durable open record rather than from the churn of daily reporting, the failures of the opening campaign cluster into four families that interacted with each other. Each deserves treatment on its own terms before any of them can be sorted.

The Plan That Assumed Its Own Success

The opening campaign was not a conventional invasion plan executed poorly. It was a regime-collapse plan executed exactly as written, whose central assumption failed. That distinction is load-bearing for everything that follows.

The design contemplated rapid movement along multiple axes toward political objectives, with the expectation that resistance would be limited, that command and control on the defending side would fragment quickly, and that the political structure would not hold. Under that assumption, the force posture makes sense: light columns moving fast on roads, minimal artillery preparation, air defense and electronic warfare assets held back rather than pushed forward, sustainment sized for a short operation rather than a campaign. Under that assumption, the absence of a serious rear-area security plan makes sense, because there would not be a hostile rear area for long. Under that assumption, the compartmented planning makes sense, because operational security against a short-notice move matters more than the readiness of subordinate units that will not have to fight much anyway.

The assumption was wrong. And when the central assumption of a plan is wrong, the plan does not degrade gracefully. It fails in a way that looks like general incompetence, because every choice that the assumption justified becomes a liability at once. The columns that were fast because they were light became vulnerable because they were light. The sustainment that was thin because the operation was short became a crisis when the operation was not short. The compartmentation that protected the surprise meant that units met the war without the preparation that would have let them adapt to it.

An analyst reading that sequence has a choice. One reading says the plan’s failure proves the force is incompetent. The other says the plan’s failure proves that this plan rested on a political assessment that was catastrophically wrong, and that the force’s performance under a plan built on a false premise is weak evidence about its performance under a plan built on a sound one. The second reading is more defensible and less satisfying. It is also the one that survives the observation that the same force, later, fighting a different kind of war under different assumptions, performed differently.

That does not exonerate the force. A serious army is supposed to have the institutional capacity to tell its political leadership that a plan’s premise is wrong, and the failure to do that is itself an institutional property, one that lands squarely in the structural bin. But the distinction between “this force cannot fight” and “this force cannot correct its leadership’s assumptions” is enormous, and only one of those two claims is well supported.

Logistics as the First Constraint

The supply failures of the opening weeks are the most photographed and the least well understood part of the campaign. The images of abandoned vehicles on roads and the arithmetic of fuel consumption produced a simple story: the force ran out of gas. The durable analytical point underneath is more specific and more useful.

The Russian ground force is built around rail. That is not an accident or an oversight; it is a deliberate design inheritance, appropriate to the geography it was built to fight across, where distances are long and road networks are thin and a railhead can be pushed forward behind an advance. A force built that way carries a characteristic signature: it can sustain very heavy consumption close to a functioning railhead, and its capacity falls off sharply as the distance from that railhead grows, because the wheeled transport that bridges the gap exists in quantities calibrated to a shorter bridge than a deep road-bound advance requires.

The opening campaign attempted advances that put substantial distance between the fighting and any railhead the force controlled, in a country whose rail network the defender was actively denying, along road axes that the defender could interdict. The result was not a surprise to anyone who understood the design. It was the design meeting a mission it was not built for.

That framing changes what the failure means. It is not “Russia cannot do logistics.” The force sustained enormous artillery expenditure for extended periods in later phases, which is a logistics achievement of real scale, and it did so because those phases were fought close to rail. The defensible claim is narrower: the force’s sustainment capacity is a function of distance from rail, and it degrades steeply past a threshold that is shorter than Western mechanized forces would assume from their own design. That claim is durable. It sits in the structural bin, because it follows from how the force is built, and rebuilding a logistics architecture around road transport at scale is a decade-long institutional and industrial project rather than a lesson learned.

It also has a corollary that cuts the other way, and the corollary is the part the popular narration drops. If the constraint is distance from rail, then a contingency fought close to rail does not trigger it. An assessment that banks on Russian logistical incapacity without asking where the railheads are has converted a conditional weakness into an unconditional one, and has thereby made itself wrong in exactly the cases where it matters most.

Command, Control, and the Cost of Compartmentation

The command failures of the opening campaign had two distinct sources that are worth separating, because they sort into different bins.

The first was the compartmentation of the plan. Subordinate commanders received their missions extremely late, in some cases without the context that would let them understand the operation they were part of. Units crossed the border without having rehearsed the movement, without accurate knowledge of the resistance they would meet, and in some cases without having been told that they were going to war rather than to an exercise. Command performance under those conditions is not a measurement of command capability. It is a measurement of what happens when a force is denied the preparation cycle that command capability depends on. That failure is substantially situational: it flowed from a specific choice about secrecy driven by a specific political calculation, and a different operation with a normal preparation cycle would not reproduce it.

The second source is different and does not go away. The force’s command culture concentrates decision authority high and distributes it reluctantly. Initiative at the level where contact actually happens is not systematically cultivated, and the professional layer that in Western armies carries small-unit initiative, the senior noncommissioned officer corps, does not exist in the same form or with the same authority. This is a design choice with a long institutional history and a coherent internal logic, and it has real advantages: it makes a force controllable, it reduces the training burden at the bottom, and it suits a mass army built around conscription rather than around a professional career core. It also produces a predictable failure mode. When the plan breaks, a force built that way waits for direction rather than improvising toward the intent, because improvising toward the intent is not what it was built to do and is not what its people are rewarded for.

That second source is structural. It is not a mistake that can be corrected by a lessons-learned process, because it is not a mistake at all. It is a coherent design with a cost, and the cost shows up precisely when circumstances depart from the plan. An adversary read that expects this property to disappear because the force noticed the cost misunderstands what it is looking at.

A related structural property compounds it. In an institution where reporting upward is coupled to career consequences, and where the tolerance for unwelcome information at the top is low, the information that reaches the decision-maker is filtered by the interests of the people passing it along. That is not unique to any one country and no military is immune. But the degree matters, and the open record of the opening campaign supports the judgment that the degree here was severe enough to be causal: the assessment of the defender that the plan rested on had passed through a chain with strong incentives to tell the top what it wanted to hear. A force can fix its drone doctrine in a year. It cannot fix that in a year, because fixing it requires changing what happens to the person who delivers bad news, and that is a property of the political system the force sits inside rather than of the force.

Intelligence That Told the Plan What It Wanted

The intelligence failure preceding the invasion was not primarily a collection failure. The relevant facts about the defender were substantially available. It was an assessment failure, and specifically the kind that occurs when the analytical product is shaped by the political preference of its consumer.

The picture that informed the plan held that the defending state’s political cohesion was brittle, that resistance would be localized and short, and that a substantial portion of the population would be indifferent or welcoming. Each of those was an assessment, not a fact, and each was wrong in the direction that the political leadership preferred. That directionality is the tell. Random error scatters. Error that all points the same way indicates a systematic bias in the process producing it.

This failure lands hard in the structural bin, and it is the most consequential entry in the whole ledger, because it is the one that produced all the others. The plan that assumed its own success assumed it because the assessment said it could. And the mechanism that produced the assessment, an analytical apparatus reporting to a decision-maker with known preferences and low tolerance for contradiction, is intact. It is intact because it is not a malfunction of the system; it is the system working as designed for other purposes, purposes like regime security that it serves quite well.

The corollary matters as much as the finding. If the mechanism persists, then the next plan is also vulnerable to a premise error, and the direction of the error is predictable: it will overestimate the fragility of an adversary’s cohesion and underestimate the adversary’s will. That is an assessment about an assessment process, and it is one of the few genuinely forward-looking things the failure record supports. It does not say the next plan will fail. It says the next plan carries a specific and identifiable risk of resting on a false premise, and that the false premise will run in a known direction. For an analyst tracking the question of how an adversary reads intentions, that is a usable finding rather than a rhetorical one.

Combined Arms in Name Only

The fourth family of failure is the one most often cited and most often flattened. The force did not integrate its arms. Armor moved without adequate infantry screening it. Infantry moved without adequate fires supporting it. Fires operated without adequate reconnaissance feeding them. Aviation operated on a schedule of its own rather than in support of the ground scheme. Engineers were not where the obstacles were. The result, repeatedly, was a force that presented itself to a defender one arm at a time, which is the condition every defender wants and none expects to be handed.

The immediate cause was training and time. Combined-arms integration at the tactical level is the single most training-intensive thing a land force does, and it decays fast without practice. The units committed at the opening had not practiced the integration they were asked to perform, partly because of the compartmentation already discussed and partly because the training system in peacetime had not been generating that competence at the scale the force’s paper structure implied.

The deeper cause is more interesting and cuts toward the structural bin. Combined-arms integration under contact requires exactly the thing the command culture does not cultivate: junior leaders making rapid coordination decisions without waiting for authority. A tank commander and an infantry squad leader adjusting to each other under fire are not executing a plan; they are improvising a solution inside an intent. A force that concentrates decision authority high can produce combined arms when the plan holds and the timeline is the planned timeline. It struggles to produce combined arms when the plan breaks, because the improvisation required has no institutional home.

That is why this family straddles two bins, and the straddle is honest rather than evasive. The specific incompetence of the opening weeks was substantially situational, a product of unprepared units executing an unrehearsed plan. The underlying difficulty in generating combined arms under conditions of disruption is structural, because it flows from the command culture, which is a design property. The force has demonstrated that it can improve the first and has not demonstrated that it can fix the second.

What the Failure Record Can and Cannot Support

Before sorting, the honest limits of the exercise have to be stated, because a ledger that overstates its own confidence produces worse decisions than no ledger at all.

What the Open Record Actually Supports

The open record supports the broad shape of the failures with high confidence. That the opening campaign faltered on supply, command, intelligence, and integration, and that the war then shifted toward an attrition character, is established across multiple independent classes of source: open defense reporting from the theater, official force-structure statements from several governments, the published output of research institutions with sustained analytical programs on the war, and the physical record of documented equipment losses maintained through open imagery verification. These sources disagree at the margins and converge on the shape. Convergence across sources with different methods and different interests is the strongest thing the open record offers, and it is strong enough to reason from.

The open record supports the direction of adaptation with moderate confidence. That the force changed its practices in identifiable ways after the opening phase, notably around unmanned systems, electronic warfare, artillery employment, and prepared defense, is well attested and visible in the pattern of the fighting itself. What is less well attested is the depth of those changes: whether they represent doctrine, training, and institutional learning, or whether they represent local adaptation by units that will not transmit it. That distinction is not resolvable from open sources with confidence, and any assessment that claims it is has exceeded its evidence.

The open record does not support precise claims about readiness rates, the true condition of stored equipment, the internal quality of reconstituted formations, or the depth of specific stockpiles. Numbers in these areas circulate widely, they vary substantially between sources, and they change. They should be treated as ranged and directional rather than as measurements, and any figure encountered in this domain should be confirmed against current reporting before it is used in a decision. The honest position is that the shape is knowable and the precision is not.

Why Failure Analysis Ages Badly

Every failure assessment has a date stamped on it, whether or not the assessment says so. The trouble is that the date wears off in transmission. A finding generated from the evidence of one phase gets cited in a later phase without its date, then cited again from that citation, until it circulates as a standing fact about the force rather than as a description of a moment.

The mechanism is not mysterious. Findings that are vivid travel further than findings that are hedged. “The force cannot supply an advance beyond a hundred kilometers from rail” is a hedged finding with a condition attached. “Russian logistics collapsed” is vivid. The vivid version travels, sheds its conditions, and arrives at the decision-maker as a property. By the time it arrives, the force may have addressed the specific problem the original finding described, and the decision-maker is now reasoning from a fact that has quietly become false.

The defense against this is procedural rather than intellectual. Every failure finding should be recorded with the phase it was observed in, the cause attributed to it, and an explicit statement of what would count as evidence that it had been corrected. A finding recorded that way carries its own expiration logic. A finding recorded as a bare claim does not, and will outlive its accuracy. Readers who track this material seriously often keep a private, versioned note for exactly this purpose, and that is a natural place to save and annotate this assessment privately in VaultBook, where each finding can carry its date, its cause attribution, and its falsification condition without any of it leaving the reader’s own device.

What a Failure Cannot Prove

A failure is evidence about a performance under conditions. It is not evidence about a ceiling.

This is worth stating plainly because the inference from observed failure to capability ceiling is made constantly and is almost never licensed. A force that performs badly under conditions of surprise, compartmentation, and a false premise has told you what it does under those conditions. It has told you very little about what it does under conditions of long preparation, honest assessment, and a plan whose premise holds. Those conditions may never obtain. Assessing that they are unlikely to obtain is a reasonable judgment with real support, given the structural findings about the assessment process. But that judgment is doing the work, not the failure record, and the two should not be confused.

Symmetrically, an observed correction is evidence about a performance under conditions too. A force that learns to mass fires against a static defense has demonstrated something real. It has not thereby demonstrated that it can conduct a fluid combined-arms breakthrough against a prepared and mobile defender, because that is a different problem requiring different competencies, and the demonstration does not extend to it.

The reader who takes only one methodological point from this article should take this one: the failure record constrains the assessment, it does not determine it. It rules out certain confident claims in both directions. It leaves a range. The job is to state the range honestly and name what would narrow it.

The Failure-Durability Ledger

The ledger is the artifact this article exists to produce. It takes the failure families above, breaks them into specific entries, sorts each into structural, corrected, or situational, states the confidence, and names the evidence that would move the entry to a different bin. That last column is the one that makes it a ledger rather than an opinion, because it commits the assessment to a falsification condition in advance.

Which Russian failures in Ukraine are structural, and which were corrected?

Structural entries grow from institutional design: assessment integrity, command centralization, the absent professional junior leader layer, and rail dependence. Corrected entries are behaviors the force demonstrably changed under pressure: unmanned systems, electronic warfare, artillery practice, and prepared defense. Situational entries belonged to the opening plan’s specific and unrepeatable conditions.

Failure entry	Bin	Confidence	Why it sits there	What would move it
Assessment shaped by leadership preference	Structural	High	The error ran one direction, indicating systematic bias rather than noise; the reporting incentives that produced it are properties of the political system, not the force	Evidence that unwelcome assessments now reach and change decisions at the top, sustained over more than one decision cycle
Centralized command, thin junior initiative	Structural	High	A coherent design choice with a long institutional history, suited to a mass conscript force; the cost appears only when plans break	A rebuilt professional junior leader corps with real delegated authority, visible in training structure rather than in rhetoric
Sustainment degrades steeply with distance from rail	Structural	High	Follows from force design and industrial base; rebuilding around road transport at scale is a decade-class project	Sustained road-based sustainment of a deep advance, or a visible expansion of wheeled transport formations
Combined arms under disruption	Structural	Moderate	Requires the junior initiative the command culture does not cultivate; improvement is visible in set-piece conditions, not in fluid ones	Demonstrated fluid integration against a mobile defender under conditions the plan did not anticipate
Officer casualty exposure from forward direction	Structural	Moderate	A direct consequence of centralization; senior officers go forward because subordinates will not act without them	Evidence that units execute intent when the commanding officer is absent
Unmanned systems at scale	Corrected	High	Went from marginal to central across the force in identifiable phases; visible in the character of the fighting itself	Not applicable; the correction is established. The open question is depth, not existence
Electronic warfare integration with fires	Corrected	Moderate	Demonstrated repeatedly at the tactical level after the opening phase	Evidence the integration is ad hoc and does not transmit to newly formed units
Artillery employment and counter-battery practice	Corrected	Moderate	Improved markedly from the opening phase, though from a low base and in a fight that favored it	Evidence that the improvement depends on positional conditions and lapses under maneuver
Prepared defense and fortification	Corrected	High	The force demonstrated it can build and hold prepared defensive belts at scale, which is a genuine institutional competence	Not applicable; the correction is established for defense, and does not extend to offense
Force generation without general mobilization	Corrected	Moderate	The force found methods to sustain manpower inflow while avoiding the political cost of full mobilization	Evidence the methods are exhausting their pool or degrading quality below usable thresholds
Rear-area security absent from the plan	Situational	High	Followed from the regime-collapse premise; a plan without that premise includes rear-area security as a matter of routine	Nothing; it is an artifact of one plan, though its cause sits in the structural bin
Units committed without rehearsal or context	Situational	High	Followed from the compartmentation choice driven by a specific secrecy calculation	Evidence that compartmentation of this severity is standard rather than exceptional
Multi-axis dispersion beyond supportable frontage	Situational	High	Followed from the assumption of light resistance; a force expecting a fight concentrates	Evidence the force cannot concentrate even when it intends to
Air campaign not sequenced with the ground scheme	Situational	Moderate	Reflected the short-war premise and a defended air environment; a different premise changes the sequencing	Evidence that the force cannot sequence air and ground even when it plans to
Communications discipline and improvised networks	Situational	Moderate	Aggravated by the unrehearsed commitment; partially addressed afterward	Evidence the problem recurs in units with normal preparation cycles

Fifteen entries, three bins, and the distribution is the finding. Four entries in the structural bin with high or moderate confidence, and one of them, the assessment-integrity entry, is causally upstream of most of the situational bin. Five corrected entries, all of them real, none of them trivial, and every one of them concerning either defense or the positional character of the fight rather than offensive maneuver. Five situational entries, all of them downstream of the same false premise, all of them likely to be absent from a plan that does not share that premise.

Reading the Ledger

The ledger is not a scorecard and the bins are not weighted equally. The right way to read it is by asking what each bin licenses.

The structural bin licenses forward-looking claims with a long shelf life. If the assessment apparatus still filters bad news, then the next plan carries premise risk. If command remains centralized and the junior leader layer remains thin, then the force will remain brittle at the point where a plan meets an unexpected defender. If sustainment remains rail-coupled, then reach remains a function of geography rather than of will. These claims will be approximately as true in some years as they are now, because the things producing them move on institutional timescales measured in decades, not on campaign timescales measured in months.

The corrected bin licenses the opposite kind of claim: a warning against banking on a weakness that no longer exists. Anyone planning against a force that cannot use unmanned systems, cannot integrate electronic warfare, or cannot mass artillery effectively is planning against a force that stopped existing early in this war. Those corrections are the clearest evidence in the record that the institution can learn under pressure, and they are the reason the permanent-incompetence reading collapses on contact with the second half of the war.

The situational bin licenses the least and is the most cited. Almost every vivid image from the opening campaign lands here: the traffic jams, the abandoned vehicles, the units that did not know they were at war, the columns strung out along roads without flank security. These are the memorable failures, and they are the least predictive, because they were produced by a plan built on a premise that a different operation would not share. They are worth studying for what they reveal about the premise and about the institution that generated it. They are close to worthless as predictions of how the same force behaves under a plan that expects a fight.

That inversion, the most vivid failures being the least predictive and the least vivid being the most, is the core of why the popular reading of this war misleads. It is not that the popular reading has its facts wrong. It is that it weights them exactly backward.

What the Three Bins Do Not Capture

Two limits belong on the record. First, the bins describe failure durability, not net capability. A force can carry four structural weaknesses and still be dangerous, because danger is a product of weaknesses, strengths, mass, geography, and the defender’s own condition. The ledger is one input to a capability judgment and not the judgment itself. The net verdict on whether the force is overrated or underrated is worked through in the series’ assessment of whether Russia’s army is overrated or underrated, which weighs the failure record against the strengths this article does not attempt to inventory.

Second, the bins are a snapshot of a moving object. The corrected bin was empty at the start of the war and filled over time. There is no principled reason to expect it to stop filling. Entries currently in the structural bin are there because of a judgment about institutional timescales, not because of a law of nature, and a force that survives a long war and rebuilds under new leadership can change things that looked fixed. The confidence column reflects that: high confidence on the entries tied to the political system, moderate on the entries tied to the force alone, because the force alone is the part more capable of changing.

The Structural Bin: Failures Likely to Persist

Each structural entry deserves its own treatment, because the reason each sits in that bin is different and the evidence that would move it is different.

Assessment Integrity and the Premise Problem

The most important entry in the ledger is also the least military. The plan for the opening campaign rested on a political assessment of the defender that was wrong in a specific direction, and the mechanism that produced that assessment is not a military mechanism.

The general shape of the problem is well documented across many states and many eras, and it has a name in the tradecraft literature: politicization of intelligence. It occurs when the analytical product is shaped, consciously or not, by the known preferences of the person receiving it. It does not require anyone to lie. It requires only that the career consequences of delivering an unwelcome assessment be worse than the career consequences of delivering a welcome one that later proves wrong. Under that incentive structure, the analytical product drifts toward the preference, and it drifts most on exactly the questions where the preference is strongest and the evidence is most ambiguous. Adversary political cohesion is the paradigm case: the evidence is genuinely ambiguous, and the preference is genuinely strong.

What makes this entry structural rather than correctable is that the fix is not available to the institution that suffers from it. A military can reform its training. It can restructure its logistics. It can rewrite its doctrine. It cannot unilaterally change what happens to the person who tells the head of state that a cherished premise is false, because that is determined by the head of state and the system around them. The failure sits above the force in the hierarchy, which means the force cannot correct it from below.

The forward-looking implication is narrow and specific, and it should be stated carefully to avoid overreach. It does not say the force will always plan badly. It says a specific class of risk persists: plans that depend on an assessment of adversary cohesion or will are vulnerable to premise error running in the direction of underestimating that cohesion. The class of risk is identifiable. What it will produce in any particular case is not.

This is also the entry most vulnerable to a symmetric error on the analyst’s own side. It is comfortable to conclude that an adversary’s assessment process is corrupted and one’s own is not. The honest version of this finding includes the observation that the mechanism is general, that it operates in every system where career incentives and analytical judgment intersect, and that the difference is one of degree and of the institutional checks that exist to counteract it. The judgment that the degree is severe in this case is supportable from the observed directionality of the error. It is not a claim of categorical difference, and stating it as one would be exactly the flattery that produces the same failure in reverse.

Command Centralization and the Missing Layer

The second structural entry is the one Western analysts find hardest to read fairly, because it is a design so different from their own that it looks like an error rather than a choice.

Western land forces are built around a professional noncommissioned officer corps that carries small-unit leadership, holds real authority, and is expected to act inside the commander’s intent without waiting for direction. That structure is expensive. It requires long careers, deep training investment, and a promotion system that keeps talented people in enlisted ranks for decades. It also requires a political willingness to distribute decision authority downward, which has its own costs in control.

The Russian design chose differently, and the choice is coherent. Authority concentrates in the officer corps. Junior enlisted service is short and largely conscript. The professional layer between the private and the lieutenant is thin, and what exists does not hold the authority its Western counterpart does. This produces a force that is controllable, that can be generated quickly from a conscript base, and that does not require the sustained investment a professional core demands. Those are real advantages, and a state that expects to fight a mass war with a mass army has reasons to value them.

The cost appears at the moment the plan stops matching reality. A force organized this way executes a plan well and improvises poorly, because improvisation requires exactly the delegated authority the design withholds. When the plan holds, this looks like discipline. When the plan breaks, it looks like paralysis, and the paralysis is not a failure of the individuals involved. It is the design producing its designed behavior under conditions the design was not optimized for.

There is a second-order effect worth naming, because it shows up in the ledger separately. In a force where subordinate units will not act without direction, senior officers must go forward to provide it. Going forward exposes them. The pattern of senior officer casualties in this war is consistent with that mechanism, and the mechanism is structural, which means the pattern should be expected to recur rather than treated as a one-time anomaly.

Could this be corrected? In principle. Building a professional junior leader corps is a known problem with known solutions, and the force is not incapable of institutional change, as the corrected bin demonstrates. But it is a generational project. It requires changing recruitment, pay, career structure, training pipelines, and the officer corps’ willingness to give up authority, and the last of those is the hardest because it asks the people making the decision to reduce their own prerogatives. The evidence that would move this entry is not a statement of intent or a reorganization on paper. It is a training structure that visibly invests in the layer, sustained across years, and units that execute intent when the commander is absent. Until that is visible, the entry stays where it is.

Sustainment and the Tyranny of the Railhead

The logistics entry has already been introduced, and the point to develop is what it implies rather than what it is.

The claim is conditional: sustainment capacity degrades steeply past a threshold distance from a controlled railhead. The conditionality is the whole value of the claim, because it tells you exactly when the weakness bites and exactly when it does not. A contingency fought within the threshold does not trigger it. A contingency requiring a deep advance beyond it does, and the deeper the advance the harder it bites.

Two things follow. First, an assessment of what the force can do somewhere has to begin with a map of rail rather than a count of brigades, because the count is not the constraint. Second, the force’s own planners know this, which means their planning will tend to prefer objectives inside the threshold and will attempt to extend the threshold before attempting anything past it. That preference is itself an indicator, and it is one of the more legible ones the structural bin yields, because rail extension and repair effort is visible and slow.

It is worth being precise about the strength of this claim, because it can be overstated. The force is not incapable of road-based sustainment; it does it constantly at shorter distances. The claim is about the shape of the curve, not about a cliff, and about a design imbalance rather than an absence. Anyone converting it into “the force cannot move beyond its railheads” has taken a useful conditional and made it into a false absolute, which is the same error the popular reading makes everywhere else.

Combined Arms Under Disruption

The fourth structural entry is deliberately narrower than the claim usually made, and the narrowing is what makes it survive the evidence.

The broad claim, that the force cannot do combined arms, is false. It was arguably false even in the opening campaign, where some formations performed the integration adequately, and it is clearly false in later phases where the force conducted coordinated operations combining fires, unmanned reconnaissance, electronic warfare, and ground maneuver with tolerable effectiveness. That is combined arms. It happened.

The narrower claim, that the force struggles to generate combined arms under conditions of disruption, is well supported and follows directly from the command entry. Integration when the plan is holding and the timeline is the planned timeline is an execution problem, and the force executes. Integration when the plan has broken and the timeline has collapsed is an improvisation problem, and improvisation is what the design does not supply. The distinction predicts the observed pattern: adequate integration in set-piece operations the force prepared for, degraded integration in encounters the force did not anticipate.

The confidence here is moderate rather than high because the evidence is harder to read. Distinguishing “could not integrate” from “integrated adequately but was defeated by a competent defender” requires more granularity than open sources reliably provide, and the temptation to attribute every reverse to the adversary’s incapacity rather than to the defender’s skill is strong and should be resisted. The entry stands on the causal link to the command structure rather than on the raw observation, which is the right basis for it but a less certain one.

The Corrected Bin: What the Force Actually Fixed

The corrected bin is where the permanent-incompetence reading dies, and it deserves as much attention as the structural bin gets, because ignoring it is the specific error this article exists to prevent.

Adaptation Under Pressure Is the Norm, Not the Exception

Militaries at war learn. This is one of the most robust findings in the study of conflict, and it holds across states, eras, and levels of initial competence. Forces that begin a war badly and are not destroyed early tend to improve, because war supplies the two things peacetime training cannot: real feedback and existential motivation. The historical record of major wars is substantially a record of forces that opened poorly, absorbed the lesson, and fought the later phases as different organizations.

There is no reason to expect this force to be an exception, and the evidence says it is not. The question was never whether it would adapt. The question was what it would adapt, how fast, and how deeply, and those are the questions the corrected bin tries to answer.

Speed first. The adaptations that arrived fastest were the ones requiring the least institutional change: employing more unmanned systems, adjusting artillery practice, digging in. These are behaviors a force can change without changing what it is. The adaptations that have not arrived are the ones requiring institutional change: delegated authority, honest reporting, a professional junior layer. That pattern is exactly what the structural and corrected bins predict, and its consistency is a point in favor of the ledger’s construction. A force adapts where adaptation is cheap and does not where adaptation costs identity.

Unmanned Systems and the Reconnaissance-Fires Loop

The most complete correction in the record concerns unmanned systems, and it is complete enough that any assessment still describing this force as behind on drones is describing a force that stopped existing early in the war.

The trajectory is well attested in open reporting: from a marginal capability employed by specialist units, to a pervasive one employed at nearly every echelon down to the small unit, integrated into the targeting cycle so that reconnaissance and fires close on each other in timescales that would have been implausible at the war’s start. Both sides drove this, each responding to the other, and the resulting competition compressed adaptation cycles to a pace neither side’s peacetime institutions had ever operated at.

The finding this correction supports is not just about drones. It is about the reconnaissance-fires loop, which is the actual military object. A force that can find, decide, and strike inside a short window has changed what its artillery mass means, because mass that arrives on an accurate location quickly is a different weapon from mass that arrives late on a stale one. That transformation is the single most consequential thing the corrected bin contains, and it argues directly against reasoning from the opening campaign’s fires performance to any current judgment about what the force’s artillery represents.

The honest caveat is depth. Whether this is doctrine or practice, institutionalized or personality-driven, transmitted to new formations or resident in the units that learned it, is not resolvable from open sources. The correction is real. Its permanence is an open question, and a serious analyst holds it as such.

Electronic Warfare, Fires Practice, and Prepared Defense

Three further corrections belong on the record, each with its own confidence.

Electronic warfare integration improved markedly, and it improved in the specific direction that matters against a drone-saturated battlefield: forward-deployed systems tied to tactical units rather than held at high echelon. That is an organizational change as much as a technical one, and organizational changes transmit better than technical ones, which is why the confidence here is moderate rather than low.

Artillery employment improved from a low base, and the caveat matters as much as the improvement. It improved in a fight whose character rewarded exactly what the force is built for: positional, attritional, close to rail, with the reconnaissance loop feeding it. Improvement under favorable conditions is real evidence but weaker evidence than improvement under unfavorable ones, and the entry’s moderate confidence reflects that.

Prepared defense is the correction with the highest confidence and the clearest evidence. The force demonstrated that it can construct and hold defensive belts at operational scale, with obstacle systems, engineering effort, and fires integration, and that it can do so well enough to blunt a determined offensive. That is a genuine institutional competence and it should be credited plainly. It should also be bounded plainly: it is a defensive competence. It does not extend to offensive maneuver, and reasoning from it to an offensive judgment is the same category error as reasoning from the opening campaign to a defensive one.

Force Generation Without General Mobilization

The last corrected entry is the least military and among the most consequential for any assessment of what a defender would face.

The opening force was sized for a short operation and could not sustain a long one. The obvious fix, general mobilization, carried a political cost the leadership was unwilling to pay. The force and the state instead developed methods to sustain manpower inflow without that step, drawing on financial inducement, regional recruitment, and various forms of pressure short of a general call-up. Those methods worked well enough to keep the war going for years, which is a real finding regardless of what one thinks of the methods.

The entry sits in the corrected bin with moderate confidence, and the moderation carries the important qualification. The methods have costs that accumulate rather than resolve: they draw on a finite pool, they are expensive, and they trade quality for quantity in ways that show up in the fighting. Whether they are sustainable indefinitely is genuinely contested among serious analysts, and this article does not resolve it. The current condition of the force that these methods have produced is worked through in the series’ assessment of Russia’s army after Ukraine, which owns that question, and readers who want the condition rather than the failure record should go there.

The Situational Bin: Artifacts of One Campaign

The situational bin is the largest source of the vivid material and the smallest source of predictive value, and understanding why is worth the space.

The Assumptions Peculiar to the Opening

Every situational entry traces to the same root: a plan that assumed the defending state would fold. Rear-area security was absent because there would not be a hostile rear. Frontage was dispersed across multiple axes because the axes were not expected to be contested. Units were committed without rehearsal because secrecy against a short-notice move mattered more than preparation for a fight that was not expected. The air campaign was not sequenced with the ground scheme because a sequenced air campaign is what you do before a real war.

Every one of those choices is defensible given the premise. Every one is indefensible without it. And that is exactly what makes them situational: they are not properties of the force, they are consequences of a premise, and a plan without that premise does not produce them.

The analytical discipline this demands is uncomfortable, because it requires setting aside the most memorable evidence. The traffic jam on the road north is the image everyone carries. It is also close to meaningless as a predictor, because it was produced by a light column moving fast on a road under an assumption of no resistance, and a force expecting resistance does not move that way. Reasoning from that image to a judgment about how the force conducts a prepared offensive is reasoning from a photograph of a specific mistake to a claim about a general capability.

Why Situational Does Not Mean Irrelevant

The bin’s contents are weak predictors of force behavior and strong evidence about something else: the institution that generated the premise.

This is the connection that makes the ledger cohere rather than fragment. The situational failures are downstream of the assessment failure, which is structural. So the situational bin does not predict what the force will do, but it does illustrate, with unusual clarity, what happens when the structural assessment failure runs to completion. It is a worked example of premise risk. Read that way, the traffic jam is not evidence about logistics or about tactical competence. It is evidence about what a false premise costs when it is embedded in a plan and executed faithfully by a force built to execute faithfully.

That reading is more useful than the one it replaces, and it is the reason the situational entries stay in the ledger rather than being discarded. They are not predictions. They are illustrations of the structural entries, which are predictions.

The Air Campaign Entry and Its Caveat

One situational entry deserves a note because it is contested. The absence of a serious sequenced air campaign in the opening is often read as evidence that the force cannot conduct one, which would make it a structural entry rather than a situational one.

The evenhanded position acknowledges the strength of that reading. The force did not achieve air superiority over a defended airspace, and the reasons offered by different analysts vary: the short-war premise made the effort seem unnecessary; the defender’s air defenses were more resilient than assessed, which is the premise problem again; the force’s capacity for large-scale suppression operations may genuinely be more limited than its inventory suggests. The third of those, if true, is structural. The first two are situational. The open record does not cleanly separate them, and honest analysts disagree.

The entry sits in the situational bin at moderate confidence because the premise explanation accounts for the observed behavior without requiring the capability claim, and the simpler explanation gets provisional preference. That is a judgment, it is contestable, and a reader who weighs the capability explanation more heavily than this article does is not making an error. The disagreement is real and the evidence underdetermines it, which is precisely the kind of thing an assessment should say out loud rather than resolve by assertion.

Four Recurring Mistakes in Reading This War

The ledger is a device for avoiding specific errors, and naming those errors directly is worth doing, because each of them appears constantly in serious commentary and each has a distinct signature that makes it catchable.

Mistake One: Treating an Event as a Property

The first and most common has already been named, and it is the parent of the other three. An event gets observed, the event gets converted into a property of the institution, and the causal step that would license the conversion never happens.

The signature is a claim with no condition attached. “The force cannot supply an advance more than a certain distance from a railhead” has a condition and is a property claim that has earned itself. “Russian logistics failed” has no condition and is an event claim wearing a property claim’s clothes. Whenever an assertion about an adversary’s capability contains no condition, no cause, and no scope, it is almost certainly this mistake, and the useful response is to ask what the causal mechanism is supposed to be. If the answer is unavailable, the claim is a description of something that happened, not an assessment of something that is.

Mistake Two: Ignoring the Correction

The second mistake is failing to update, and it is not primarily a failure of intelligence. It is a failure of attention economics.

Failures are news and corrections are not. A stalled column is a story. A gradual improvement in the integration of reconnaissance with fires across a year is not a story, and it will be covered thinly if at all, because it lacks an event to hang on. The result is a systematic asymmetry in the evidence a non-specialist reader accumulates: the failures arrive with force and the corrections arrive faintly or never. An assessment built from that reading is not lazy. It is faithfully reflecting a biased sample.

The signature of this mistake is a finding cited without a date. When someone says the force cannot do something, the immediate question is when that was last true, and the frequency with which that question cannot be answered is the measure of how widespread the mistake is. The defense is procedural: a finding recorded with its date and its falsification condition cannot be cited past its expiry without someone noticing.

Mistake Three: Preparing for the Last Failure

The third mistake is the operational consequence of the second, and it is the one that costs money and lives rather than merely credibility.

Preparing for the last war is an old accusation and usually an unfair one. The sharper and more common version is preparing for the last failure: building a defense optimized against the specific mistakes an adversary made in an observed campaign, on the assumption that the adversary will make them again. It is seductive because it feels evidence-based. It is dangerous because the adversary observed the same mistakes, paid for them directly, and has stronger incentives to correct them than the observer has to notice the correction.

The asymmetry there deserves emphasis. The force that made the mistakes experienced them as casualties and lost objectives. The observer experienced them as reporting. Motivation to correct is not evenly distributed between those two positions, and the observer’s confidence that the mistake will recur is therefore built on a sample that the other party is actively working to invalidate. Any planning assumption that depends on an adversary repeating an error they have already paid for should be treated as a wasting asset with a short and unknown remaining life.

Mistake Four: The Symmetric Overcorrection

The fourth mistake is the reaction to the third, and it is less discussed because it feels like sophistication.

Having learned that the failures were overread, an analyst overcorrects into treating the whole failure record as noise. The force is declared to have learned, the corrections are generalized past their evidence, and the structural bin is quietly discarded because it seems to belong to the discredited earlier reading. This is the mirror of the permanent-incompetence error and it is equally unsupported: the corrections are real, they are also bounded, and every one of them stops precisely at the institutional line.

The signature of the fourth mistake is the phrase that treats adaptation as general rather than specific. A force did not simply “adapt.” It adapted particular things, at particular speeds, and conspicuously not others, and the pattern of what it did not adapt is at least as informative as the pattern of what it did. An assessment that says the force learned, without specifying what it learned and what it did not, has told you nothing that constrains any decision.

The Common Root

All four mistakes share a root: the absence of a causal step between observation and conclusion. Event to property with no mechanism. Failure cited with no date. Planning assumption with no expiry. Adaptation claimed with no scope. In each case the analyst has taken something true and stretched it past what the evidence covers, and in each case the fix is the same discipline of asking what caused this, what would change it, and how would I know.

That discipline is not sophisticated. It is bookkeeping. But it is the difference between a failure record that informs a decision and one that flatters whatever the reader already believed, and the volume of commentary on this war that fails the test suggests the bookkeeping is rarer than it should be.

The Correlation of Forces and How to Read It After a Failure

Capability analysis eventually has to say something about the balance, and the failure record changes how the balance should be read rather than changing the balance itself.

Mass, Quality, and the Substitution Problem

The oldest question in force analysis is whether quantity substitutes for quality, and this war has produced the best modern evidence on it, though not the clean answer either camp wanted.

The force that entered the war carried a quality claim: professionalized, contract-heavy, modernized, with the reforms of the preceding decade having produced something meaningfully better than the mass conscript army it replaced. The opening campaign falsified the claim as stated. What the force then did was revert, deliberately, toward mass: more people, more shells, more attrition, less maneuver, and a theory of victory built on outlasting rather than outfighting. That reversion was not a failure to adapt. It was an adaptation, and specifically an adaptation toward what the institution is actually good at.

The finding this supports is important and often missed. The force did not fail and then recover its quality. It failed at a quality-dependent operation and then substituted mass, and the substitution worked well enough to keep the war going for years. That is a real capability finding, and it is not flattering to the quality claim or dismissive of the force. It says the institution’s comparative advantage is where it has always been, and the reforms did not move it as far as advertised.

For a defender reading the balance, the implication is that the relevant question is not whether the force is good. It is whether the contingency in question is one where mass substitutes. Mass substitutes well in positional attrition close to rail against a defender who must hold ground. It substitutes poorly in fluid maneuver at distance against a defender with depth and mobility. The same force is a different problem in those two cases, and a balance assessment that does not specify which case it is describing is not saying anything.

The Adaptation Clock

If corrections are real, then any assessment banking on a weakness needs to know how fast that weakness gets fixed. That is the adaptation clock, and the war supports rough judgments about its speed.

Adaptations requiring only behavior change ran fast: months, sometimes weeks, driven by the brutal feedback of a contested battlefield and the competition with an adversary adapting in parallel. Adaptations requiring equipment change ran at the speed of industrial output and inventory, which is slower and more variable. Adaptations requiring institutional change have not visibly run at all, and the ledger’s structural bin is essentially the list of things the clock has not moved.

The usable rule is that behavior corrects in months, materiel corrects in years, and institutions correct in decades or under new leadership. That rule is a rough heuristic rather than a measurement, and it should be held loosely. But it does yield a practical test for any claim about the force built on an observed weakness: identify which of the three kinds of change the correction would require, and set the expected shelf life of the claim accordingly. A claim resting on a behavior weakness is a wasting asset. A claim resting on an institutional property is durable.

This is the mechanism behind the whole argument for dating findings. It is not merely good hygiene. It is that different findings decay at rates that differ by orders of magnitude, and an assessment that treats them as a single undifferentiated pile of facts will be systematically wrong about the fast-decaying half.

How fast can Russia correct a battlefield failure?

Behavior corrections ran in months during this war, materiel corrections in years bounded by industrial output, and institutional corrections have not visibly run at all. That spread is the practical test: identify which kind of change a weakness would require, and set the shelf life of any claim resting on it accordingly.

Readiness, Sustainment, and the Limits of Open Source

The parts of the picture that decision-makers most want, exact readiness, real equipment condition, true stockpile depth, are the parts the open record supports least. Saying so is not a hedge; it is the finding.

What Is Knowable

The shape of the force’s sustainment problem is knowable. Rail dependence is a design fact, visible in the force structure and in the pattern of its operations. The character of its industrial output, expansion under wartime conditions, constrained by specific inputs, weighted toward what existing lines can produce, is well attested across multiple classes of source and consistent with the observed pattern of the fighting. The general direction of equipment inventories, drawing down modern items and reaching further back into older stocks, is visible in the documented record of what appears in the field.

The pattern of adaptation is knowable, because adaptation manifests as behavior and behavior is observable. The force’s shift toward unmanned systems, toward prepared defense, toward attritional method, and toward mass substitution is legible in the fighting itself and does not depend on any privileged source.

The institutional properties are knowable, and this is the underappreciated part. Command structure, the shape of the junior leader layer, the incentive structure around reporting, and the relationship between the analytical apparatus and the political leadership are all documented in the open literature, in the force’s own published structure, and in the historical record of the institution. They do not require current intelligence, because they do not change on current timescales. That is exactly why the structural bin carries the highest confidence in the whole ledger: it rests on the most durable and most openly available evidence.

What Is Not

Precise readiness rates are not knowable from open sources, and the figures that circulate should be treated as estimates with wide error bars. The true condition of stored equipment is not knowable; what is visible is what emerges from storage and how it performs, which is a lagging and partial indicator. Stockpile depth is not knowable with precision; consumption is partly observable and production is partly observable, but the starting inventory is not, and any depletion estimate inherits that uncertainty and compounds it.

The internal quality of reconstituted formations is not knowable in the way that matters. The number of formations is partly observable. Whether a formation with a name and a number is a coherent fighting organization or an administrative shell with people in it is exactly the question, and it is exactly the question open sources answer worst. This uncertainty is not incidental. It is the central uncertainty in any judgment about what the force could do next, and any assessment that does not carry it forward has hidden its most important error term.

Depth of the corrections is not knowable, as the corrected bin’s confidence column reflects. Whether adaptations are institutionalized or resident in the units that learned them determines whether a newly generated formation arrives with them or without them, and that determines a great deal. The evidence does not settle it.

The disciplined response to this list is not paralysis. It is a stated confidence for every claim, a range instead of a point wherever the evidence supports only a range, and an explicit note that changeable figures should be confirmed against current reporting before they carry weight in any decision. Readers who work with this material operationally often find it useful to track indicators and build a risk checklist on ReportMedic, where the failure-durability entries can be carried as a standing checklist with their falsification conditions attached, so that a correction registers as a checklist state change rather than passing unnoticed.

Why the Uncertainty Runs Both Directions

There is a persistent asymmetry in how open-source uncertainty gets used, and it is worth naming because it corrupts assessments quietly.

Uncertainty about an adversary’s condition is usually deployed in one direction only, according to the deployer’s prior. Those who think the force is hollow cite uncertainty to dismiss evidence of recovery. Those who think it is dangerous cite uncertainty to dismiss evidence of degradation. Both are using the same epistemic fact to protect opposite conclusions, which is a reliable sign that the fact is not doing analytical work.

Handled honestly, uncertainty widens the range on both ends. It means the force could be in worse condition than the visible evidence suggests and it could be in better condition, and a decision that would be regretted at either end of the range is a decision that has not accounted for the range. That is a harder standard than either camp’s version, and it is the only one that does not amount to using ignorance as a rhetorical resource.

The Two Schools: Structural Weakness Versus Adaptive Learner

The debate this article sits inside has two serious camps, and both deserve their strongest statement before any verdict.

Is Russia permanently incompetent after Ukraine?

No, and the corrected bin is the evidence. The force demonstrably changed practice under pressure on unmanned systems, electronic warfare, artillery employment, and prepared defense. What persisted is narrower and more specific: institutional properties around assessment integrity, command centralization, and rail-coupled sustainment that no lessons-learned process reaches.

The Strongest Case for Structural Weakness

The structural school argues that the failures of this war are not accidents but expressions of durable institutional properties, and that those properties are not going to change.

The case is strong. The assessment failure that produced the plan is a function of a political system that has not changed and shows no sign of changing, and a system that punishes unwelcome truth will keep producing plans built on comfortable premises. The command centralization that produced tactical brittleness is a design with a century of institutional history behind it, defended by the people who benefit from it, and correcting it would require the officer corps to surrender authority voluntarily. The rail dependence is baked into force structure and industrial base. The absent professional junior layer would take a generation to build.

The school’s strongest point is the pattern in the corrected bin itself. Look at what the force fixed: it fixed the things that could be fixed without changing what the institution is. Every correction is a behavior or a technique. Not one is an institutional property. That is not a coincidence and it is not a matter of not having gotten around to it yet. It is evidence that the institution’s adaptation capacity has a ceiling, and the ceiling sits exactly where change would require the institution to become something else. A force that has fought a long, existential war and has still not delegated authority downward is telling you it will not.

The school’s honest weakness is that it can slide toward the permanent-incompetence reading it formally rejects. Structural weaknesses are weaknesses, not disabilities. They describe what the force will find hard, not what it cannot do, and a force with four structural weaknesses and enough mass, enough time, and a favorable geometry is still a serious problem. The school’s rigorous version knows this. Its popular version does not, and the popular version is the one that reaches decision-makers.

The Strongest Case for the Adaptive Learner

The adaptive school argues that this force has demonstrated a learning capacity that its critics keep underestimating, and that the underestimation is a repeating historical pattern rather than a novel insight.

The case is also strong. The force absorbed a catastrophic opening, was not destroyed, reorganized around what it does well, and fought for years afterward against a determined defender receiving substantial external support. That is not a hollow army. Every one of the corrections in the ledger was denied or dismissed at some point by observers who had concluded from the opening that the force was incapable of it. The drone transformation in particular was not predicted by the analysts most confident about structural incapacity, and it happened anyway, and it happened fast.

The school’s strongest point is historical rather than contemporary. The institution’s own past contains the sharpest available demonstration: a force that opened a war catastrophically, absorbed losses that would have destroyed most states, rebuilt its command practice and its operational method under fire, and finished that war as one of the most capable land forces of its era. This is not a recommendation of that force or that state. It is an observation about the institution’s demonstrated capacity to reconstruct itself under existential pressure, and it is the single most inconvenient fact for the structural school. The historical pattern of this specific institution’s recovery is examined in its own right elsewhere in the series, and the adaptive school leans on it heavily and legitimately.

The school’s honest weakness is the depth question. The corrections are real and the corrections are shallow, in the specific sense that all of them stop at the institutional boundary. The adaptive school can point to a force that learned to fight the war it ended up in. It cannot yet point to a force that changed what it is. And the historical precedent it relies on involved changes of exactly that depth, made under a pressure and over a timescale that are not obviously present here. Invoking the precedent while eliding that difference is the school’s characteristic overreach.

What the Disagreement Actually Turns On

Strip away the rhetoric and the two schools disagree on one thing: whether the institutional properties in the structural bin are load-bearing.

The structural school says they are, and therefore that the corrections, however real, are decorations on a building whose foundations decide what it can hold. The adaptive school says they are not, or not decisively, and therefore that a force which learns fast enough at the level of practice can generate combat power sufficient to its purposes regardless of what its internal culture looks like.

Both propositions are testable in principle and neither is settled by the evidence available. That is the honest state of the question. What can be said is where each school is on firmer ground. The structural school is on firm ground about the properties themselves: those are documented, durable, and unchanged. The adaptive school is on firm ground about the corrections: those are documented, real, and repeatedly underestimated. The disagreement is about the weighting, and weighting is a judgment, not a finding.

A reader who wants a defensible position rather than a camp can hold this: the structural entries are the better basis for long-horizon claims because they decay slowest, and the corrected entries are the better basis for near-horizon claims because they describe what the force does now. That is not a compromise for its own sake. It is the position that follows from taking the different decay rates seriously, which is what the whole ledger is for.

What the Failures Reveal About the Force a Defender Would Face

The brief for this article stops short of the transfer, and the stopping point is deliberate. But the adversary read has to arrive somewhere, and where it arrives is a characterization of the force rather than a risk estimate for any particular defender.

The Composite Picture

Put the bins together and the force that emerges is coherent, and it is neither of the caricatures.

It is an institution that plans on premises supplied by a process with a known and directional bias, which means its plans carry a specific and recurring risk of resting on an underestimate of an adversary’s will. It is a force that executes a holding plan competently and improvises poorly when the plan breaks, because authority sits high and the layer that would improvise does not exist in strength. It is a force whose reach is governed by rail rather than by intent, which makes geography a harder constraint on it than on forces designed around road-based sustainment. It is a force that has demonstrated real and rapid learning at the level of technique, notably in closing the reconnaissance-fires loop, and real institutional competence in prepared defense. It is a force whose comparative advantage is positional attrition supported by mass, and which reverted to that advantage when the quality-dependent approach failed.

That composite is not reassuring and it is not alarming. It is specific. And specificity is what a planner can use, because it says what the force is good at, what it is bad at, and under what conditions each applies, which is more than either “they are incompetent” or “they are ten feet tall” will ever say.

The Last-War Trap

The trap this article is built to prevent has a precise shape, and it is worth stating explicitly because it is the operative risk in the whole subject.

The trap is not being wrong about the force. It is being right about the force as it was and carrying that rightness forward past its expiration date. A defender who plans against the force of the opening campaign is planning against an adversary that does not exist: one that cannot use drones, cannot mass fires accurately, cannot dig in, and will present itself one arm at a time on a road. Every one of those is false now. Planning against them is not merely wasted; it is actively harmful, because it allocates attention and resources against a threat model that has been superseded while the actual threat model goes unaddressed.

The trap has a mirror that is equally dangerous. A defender who assumes the corrections were total, that the force has fixed everything and become a Western-style adaptive institution, is also planning against something that does not exist. The structural bin is real. The force did not fix its assessment process, its command culture, or its sustainment architecture, and there is little sign it can.

Escaping both requires the same thing: the ledger, dated and maintained. Not a verdict on the force but a running record of which findings are current, which have been overtaken, and what would tell you the difference. The defender’s own side of that discipline, what a frontline state should take from this war and apply to its own force, is worked through in the series’ account of what Poland learned from Ukraine’s fight, which owns the applied lesson set and approaches the same evidence from the other direction.

Where This Sits in the Read-Across

The lessons here are one input into a larger question, and the larger question is not this article’s to answer. Whether and how this war’s evidence transfers to a NATO frontier state’s risk is a separate analytical operation with its own master variable, alliance membership, which this article has deliberately not touched. That read-across, and the discipline required to do it without stamping one war onto another, is worked through in the series’ assessment of what Ukraine tells us about Poland’s risk, which owns the top-line bellwether judgment and to which the failure record here is one contribution among several.

The relationship between the two is worth stating precisely because it is where most public reasoning goes wrong. This article supplies a characterization of the adversary. The read-across supplies the transfer rules. Running them together, jumping from a failure observation directly to a risk conclusion, is the move that produces both the alarmist reading and the dismissive one, because a characterization without transfer rules can be pointed at any conclusion the writer already held.

What the Corrections Cost

A correction is not free, and the price of each one is part of the finding. A ledger that records what changed without recording what the change cost has measured half the transaction.

The reversion to mass cost the force its quality claim and its reform narrative, and it cost it the operational method that quality was supposed to buy. An army that substitutes attrition for maneuver has chosen a theory of victory that works, slowly, at a price in people and materiel that only certain political systems can pay and only for certain objectives. That constraint is real and it bounds what the substitution can be used for. It is a viable approach against an adversary who must hold ground and cannot trade space. It is a poor approach against an adversary with depth, and it is a terrible approach against a defender whose external support scales with the length of the war.

The drone and electronic warfare corrections cost less and bought more, which is why they arrived fastest. But they also embedded the force in a competition it cannot exit: every adaptation invites a counter-adaptation, and the loop now runs at a tempo that punishes any institution slow to transmit a lesson from the unit that learned it to the unit that has not. That tempo is precisely where the structural weaknesses bite, because transmission across a force is an institutional function and institutional functions are what did not improve. A force that adapts brilliantly at the unit level and transmits poorly across the institution will keep relearning the same lesson at each new formation’s expense.

The force-generation correction carried the steepest hidden price. Sustaining inflow without a general call-up preserved political room at the cost of quality, and quality loss at the individual level interacts badly with a command structure that already places heavy demands on the officer layer. A force with thinner individual quality needs more junior leadership to compensate, and junior leadership is exactly the thing the structure does not produce. The correction and the structural weakness therefore compound rather than offset, which is a specific finding worth carrying: the fix for one problem made another problem heavier.

The pattern across all four is consistent and it is the most useful thing this section produces. The force corrected in the directions that its structure permitted and paid for those corrections in the currencies its structure had available: people, ammunition, time, and the quality of its individual soldiers. It did not pay in the currency it does not spend, which is delegated authority. That is what the corrected bin looks like from the inside, and it explains both why the corrections were real and why they stopped where they did.

Verdict: Date the Failure Before You Bank It

The rule this article advances is short enough to carry into a meeting: an adversary’s failure is an asset to your planning only for as long as it remains uncorrected, so date it before you bank it.

That rule has three operational parts. First, attribute the cause, because a failure without a cause cannot be sorted and an unsorted failure cannot be dated. Second, sort by durability, because institutional properties and behaviors decay at rates that differ by orders of magnitude and treating them alike guarantees error on the fast half. Third, state the falsification condition, because a finding that cannot be overturned by evidence is not a finding, and a finding whose overturning condition is written down will announce its own expiry instead of quietly rotting in a briefing slide.

Applied to this war, the rule yields a verdict that neither camp will find satisfying. The force failed badly at the opening, and the most vivid of those failures are the least informative, because they were artifacts of a premise rather than properties of the institution. The force corrected real things at real speed, and anyone still planning against the opening-campaign force is planning against a ghost. And the force did not correct, has not shown it can correct, and likely cannot correct within any planning horizon the things that sit deepest: an assessment process that tells the top what it wants, a command culture that will not delegate, a junior leader layer that does not exist, and a sustainment architecture chained to rail.

The last of those lists is where the durable analytical value sits. It is not the exciting part. It contains no burning columns and no abandoned vehicles. It is a set of institutional properties that will be approximately as true in a decade as they are now, which is precisely why it is worth more than everything the popular reading of this war has extracted from it.

An adversary’s failures are only lessons if you know when they were true. The failure-durability ledger is the discipline that keeps them dated, and the discipline is the deliverable.

Frequently Asked Questions

Q: Why did Russia fail in the opening Ukraine campaign?

Because the plan rested on a political premise that collapsed on contact. The design assumed limited resistance and rapid political fragmentation, and every choice that assumption justified became a liability at once when it proved false: light columns moving fast on roads, sustainment sized for a short operation, compartmented planning that denied subordinate units preparation, and no rear-area security. The force then executed a regime-collapse plan faithfully against a defender who did not collapse. That sequence produces something that looks like general incompetence but is better described as a force performing under a plan built on a false premise. The distinction matters because the two readings predict very different behavior under a plan whose premise holds, and only one of them is well supported by what the same force did in later phases.

Q: Which Russian failures in Ukraine are structural and durable?

Four entries carry high or moderate confidence in the structural bin. First, an assessment process shaped by leadership preference, which produced the false premise and which the force cannot fix from below because it sits above the force in the hierarchy. Second, command centralization paired with a thin professional junior leader layer, a coherent design with a century of institutional history whose cost appears whenever a plan breaks. Third, sustainment capacity that degrades steeply with distance from a controlled railhead, which follows from force design and industrial base. Fourth, difficulty generating combined arms under conditions of disruption, which follows causally from the command entry. These persist because the things producing them move on institutional timescales measured in decades rather than on campaign timescales measured in months.

Q: Which Russian Ukraine failures have already been corrected?

Five entries sit in the corrected bin. Unmanned systems went from marginal to pervasive across the force, closing the reconnaissance-fires loop to timescales that would have been implausible at the war’s start. Electronic warfare integration moved forward to tactical units rather than remaining at high echelon. Artillery employment and counter-battery practice improved markedly, though from a low base and in a fight whose character rewarded it. Prepared defense at operational scale became a demonstrated institutional competence. Force generation found methods to sustain manpower inflow without the political cost of general mobilization. Every one of these is real, and anyone still planning against a force that cannot use drones, cannot mass fires accurately, or cannot dig in is planning against an adversary that stopped existing early in the war.

Q: What do Russia’s Ukraine failures reveal about the threat?

They reveal a force that is neither of the caricatures. It plans on premises from a process with a known directional bias, so its plans carry recurring risk of underestimating an adversary’s will. It executes a holding plan competently and improvises poorly, because authority sits high. Its reach is governed by rail rather than by intent, making geography a harder constraint on it than on road-designed forces. It learns fast at the level of technique and not at all at the level of institution. Its comparative advantage is positional attrition supported by mass, and it reverted to that advantage when the quality-dependent approach failed. That composite is specific rather than reassuring or alarming, and specificity is what a planner can use, because it names the conditions under which each property applies.

Q: Did the Ukraine war prove Russia permanently incompetent?

No, and the corrected bin is the evidence against it. The force absorbed a catastrophic opening, was not destroyed, reorganized around what it does well, and fought for years afterward against a determined defender receiving substantial external support. Every correction in the record was denied or dismissed at some point by observers who had concluded from the opening that the force was incapable of it. Militaries at war learn, and this is one of the most robust findings in the study of conflict. What is defensible is a narrower claim: certain institutional properties did not change and show little sign of changing. That is a claim about durable weaknesses, not about permanent incompetence, and the difference between the two is the difference between an assessment and an insult.

Q: Why must Russia’s Ukraine failures be dated, not assumed?

Because different findings decay at rates that differ by orders of magnitude, and an assessment treating them as one undifferentiated pile will be systematically wrong about the fast-decaying half. Behavior corrections ran in months. Materiel corrections run in years, bounded by industrial output. Institutional corrections have not visibly run at all. A finding resting on a behavior weakness is a wasting asset; a finding resting on an institutional property is durable. Findings also lose their dates in transmission, because vivid claims travel further than hedged ones and shed their conditions along the way. A finding recorded with its observation phase, its attributed cause, and its falsification condition announces its own expiry. A bare claim does not, and will outlive its accuracy in briefing slides.

Q: What supply lessons come from Russia’s Ukraine failures?

The durable lesson is conditional, and the condition is the whole value of it. The force is built around rail, which is a deliberate design inheritance suited to the geography it was made to fight across. It can sustain very heavy consumption close to a functioning railhead, and its capacity falls off sharply as the distance grows, because the wheeled transport bridging that gap exists in quantities calibrated to a shorter bridge. The opening campaign attempted advances far beyond that threshold, in a country whose rail the defender was denying. So the lesson is not that the force cannot do supply; it sustained enormous artillery expenditure in later phases fought close to rail. It is that reach is a function of railhead geography, which means assessing what the force can do somewhere starts with a map of rail rather than a count of brigades.

Q: How did command problems shape Russia’s Ukraine setbacks?

Through two distinct mechanisms that sort into different bins. The compartmentation of the plan denied subordinate commanders context, rehearsal, and in some cases the knowledge that they were going to war rather than to an exercise, and that was situational, flowing from one secrecy calculation. The deeper mechanism is structural: authority concentrates high, initiative at the point of contact is not systematically cultivated, and the professional layer that carries small-unit improvisation in Western armies does not exist in the same form. A force built that way executes well and improvises poorly, because improvising toward an intent has no institutional home. A second-order effect follows: senior officers must go forward to supply direction subordinates will not generate, which exposes them, and the casualty pattern is consistent with that mechanism.

Q: Is Russia an adaptive learner despite its Ukraine failures?

Partly, and the qualification carries the weight. The force demonstrably learned at the level of technique, fast and under pressure, and the analysts most confident about structural incapacity did not predict the drone transformation that happened anyway. The institution’s own history contains a sharper demonstration still: a force that opened a war catastrophically and finished it as one of the most capable land forces of its era. But look at the pattern of what it fixed. Every correction is a behavior or a technique. Not one is an institutional property. That is not a matter of not having gotten around to it. It suggests the adaptation capacity has a ceiling, and the ceiling sits exactly where change would require the institution to become something else.

Q: Why is preparing for the last Ukraine failure a trap?

Because the adversary observed the same failures, paid for them directly in casualties and lost objectives, and has stronger incentives to correct them than an outside observer has to notice the correction. That asymmetry is the trap’s engine. A planning assumption built on an adversary repeating an error they already paid for is a wasting asset with a short and unknown remaining life. The trap also has a mirror that is equally costly: assuming the corrections were total, that the force fixed everything and became a Western-style adaptive institution. The structural bin is real and remains unaddressed. Escaping both requires the same discipline, a dated ledger that records which findings are current, which have been overtaken, and what evidence would tell you the difference.

Q: How fast does Russia correct a failure seen in Ukraine?

At three very different speeds, and knowing which one applies is the practical test. Behavior corrections, meaning changes a force can make without changing what it is, ran in months and sometimes weeks, driven by battlefield feedback and by competition with an adversary adapting in parallel. Materiel corrections run at the speed of industrial output and inventory, which is slower and more variable. Institutional corrections, meaning delegated authority, honest reporting, and a professional junior layer, have not visibly run at all across years of existential war. The rough rule is months, years, decades, and it should be held loosely. But it yields a usable move: identify which kind of change a weakness would require before banking any claim that rests on it.

Q: Which Ukraine failures were situational rather than permanent?

The vivid ones, mostly, which is the inversion at the heart of the whole problem. Absent rear-area security, units committed without rehearsal or context, dispersion across more axes than the frontage could support, an air campaign not sequenced with the ground scheme, and improvised communications all trace to the same root: a plan that assumed the defending state would fold. Each choice is defensible given that premise and indefensible without it, which makes them consequences of a premise rather than properties of a force. A plan expecting a fight does not produce them. They remain worth studying, not as predictors of force behavior but as a worked illustration of what the structural assessment failure costs when it runs to completion.

Q: What did Russia’s flawed pre-invasion assessment of Ukraine get wrong?

It held that the defending state’s political cohesion was brittle, that resistance would be localized and short, and that a substantial portion of the population would be indifferent or welcoming. Each of those was an assessment rather than a fact, and each was wrong in the direction the political leadership preferred. That directionality is the analytical tell: random error scatters, while error that all points one way indicates systematic bias in the process producing it. The mechanism is documented across many states and eras and has a name in the tradecraft literature. It does not require anyone to lie. It requires only that delivering an unwelcome assessment carry worse career consequences than delivering a welcome one that later proves wrong.

Q: How well does the open record support Ukraine failure claims?

Unevenly, and stating where is part of the assessment. The broad shape of the failures is supported with high confidence, because independent classes of source with different methods and interests converge on it: open defense reporting, official force-structure statements, research institutions with sustained programs on the war, and the documented equipment-loss record built through imagery verification. The direction of adaptation is supported with moderate confidence, since behavior is observable. What is not supported is precision: readiness rates, the true condition of stored equipment, stockpile depth, and above all whether a reconstituted formation is a coherent fighting organization or an administrative shell. Figures in those areas vary widely between sources and change, and should be confirmed against current reporting before carrying weight.

Q: Did Russia’s drone use in Ukraine improve after the opening?

Dramatically, and it is the most complete correction in the record. The trajectory runs from a marginal capability held by specialist units to a pervasive one employed at nearly every echelon down to the small unit, integrated into targeting so that reconnaissance and fires close on each other in short windows. Both sides drove the competition, and it compressed adaptation cycles below anything either side’s peacetime institutions had operated at. The finding is not really about drones; it is about the reconnaissance-fires loop, because mass that arrives quickly on an accurate location is a different weapon from mass that arrives late on a stale one. The honest caveat is depth: whether this is institutionalized doctrine or resident practice is not resolvable from open sources.

Q: What does Russia’s Ukraine record show about combined arms?

The broad claim that the force cannot do combined arms is false, and later phases disprove it: coordinated operations combining fires, unmanned reconnaissance, electronic warfare, and ground maneuver happened with tolerable effectiveness. The narrower and better-supported claim is that the force struggles to generate integration under conditions of disruption. Integration when the plan holds is an execution problem and the force executes; integration when the plan has broken is an improvisation problem, and improvisation is what a centralized command design does not supply. That distinction predicts the observed pattern of adequate performance in prepared set-piece operations and degraded performance in unanticipated encounters. Confidence is moderate rather than high, because open sources rarely separate a failure to integrate from a competent defender’s success.