Why neither Russia nor Ukraine can turn the tide of the war — and how they still pull off successful offensives like Kyiv’s Operation Vivaldi

Ukrainian soldiers who took part in Operation Vivaldi say a new tactic made their successful offensive near the Siverskyi Donets River north of Sloviansk possible. They are withholding the details so they can exploit the Russian army’s vulnerabilities again. Judging solely by the available evidence — videos published by both sides — the offensive’s pace and methods were no different from those of other operations in the full-scale war: Ukrainian forces spent months “chipping away” at the enemy’s defenses. They used similar methods in the Zaporizhzhia region this summer, as did Russian forces near Dobropillia over the summer and fall. Small assault groups advanced mainly by infiltration. There was not even a hint of maneuver warfare; the advance was 10–15 kilometers deep, typical for this stage of the war. What is preventing the two armies from breaking the positional stalemate? What does fighting look like today? How do soldiers counter drones on the front lines? And is there any reason to expect the dynamics of the war to change anytime soon?Operation Vivaldi: How did Ukrainian forces succeed — and what does their success mean?The Ukrainian military has not disclosed how long the operation lasted, what stages it went through, or what methods it used. Former 3rd Army Corps commander Andriy Biletsky and his subordinate brigade commanders said preparations for the offensive began back in May. In the fall, after a lengthy period of “shaping the battlefield,” Ukrainian forces quickly liberated a substantial area. On a podcast featuring prominent military analyst Rob Lee, an American volunteer fighting in the 3rd Army Corps compared these preparations to an avalanche in the mountains: Changes in the snowpack build up slowly, but then the mass sweeps away everything in its path. The soldier declined to explain his metaphor.Metaphors aside, the few published videos of Operation Vivaldi — most released several months after they were filmed — show that the advance began slowly back in the summer. Ukrainian forces managed to choke off supplies to support units of Russia’s 20th Combined Arms Army, which was defending positions along a stretch of the front north of Lyman. These units — drone operators, electronic warfare crews, and others — form the backbone of military formations’ defenses in this war.After this prolonged preparation, Ukrainian assault troops advancing from the west, north, and south had only to suppress infantry positions in several settlements: Novoselivka, Shandryholove, Kolodiazi, and Nove. These were the strongpoints anchoring the Russian defenses. After that, having lost its reconnaissance capabilities and most of its strike assets, the 20th Army’s forward force had to abandon almost all the positions west of the Chornyi Zherebets River that it had captured over the previous year. Videos suggest that artillery fire and drone strikes from the eastern bank, where the main forces of the same 20th Army are deployed, stopped the Ukrainian advance at the river.Over the months-long offensive, Ukrainian forces advanced 10–20 kilometers. In tactical terms, the success looks modest when measured by the area liberated or the depth of the advance. Its operational significance, however, is considerable.That said, Ukraine’s command is even trying to overstate the success. For several months, Ukrainian military sources have said they uncovered a Russian strategic plan to encircle the entire Ukrainian force in the Kramatorsk–Sloviansk urban area and capture all of the Donbas. Officials in Kyiv say the plan called for Russia’s 20th and 25th Armies to capture the area around Sviatohirsk on the Siverskyi Donets, cross the river — more than 100 meters wide — and reach the Sloviansk–Izium road to cut Kramatorsk and Sloviansk off from the Kharkiv region. At the same time, according to Ukrainian sources, another Russian force, the Center grouping, was to capture Dobropillia southwest of Kramatorsk and advance toward Barvinkove in the Kharkiv region to cut the roads leading west out of the Donbas. Such a plan would require Russia to assemble strike forces and advance 30–50 kilometers.Russian troop deployments and attacks, however, suggest that Russia’s command is pursuing a more modest plan. The 20th Army — two motorized rifle divisions, about eight or nine motorized rifle regiments, or 24–27 rifle battalions — clearly did not look like a main strike force capable of crossing a major river and encircling Ukrainian forces in Kramatorsk and Sloviansk. Its actions suggest its task was to reach the Siverskyi Donets near Sviatohirsk and secure the northern flank of the 25th Army, which was assaulting Lyman. Both armies were eventually to reach the Siverskyi Donets north of Sloviansk and help Russia’s South grouping advance directly toward Sloviansk along the southern bank.Operation Vivaldi forced the 20th Army to retreat beyond the Chornyi Zherebets, exposing the northern flank of the 25th Army, which still holds positions near Lyman. The 25th Army, in turn, protects the northern flank of Russia’s main strike force, the South grouping, which had been advancing toward Sloviansk along the southern bank of the Siverskyi Donets. Biletsky has just been promoted and now commands Ukraine’s newly formed Center grouping, making him responsible for operations on the southern bank as well. Ukrainian forces there also appear to have begun “shaping the battlefield” for an offensive. Preparations are still at an early stage, and the enemy they need to defeat is clearly stronger than the 20th Army’s forward detachments north of Lyman.Meanwhile, Russian forces conducted a similar operation southwest of Kramatorsk. After spending several months “shaping the battlefield,” they captured Dobropillia and reached the Dobropillia–Kramatorsk road. Disrupting Ukrainian logistics and driving out support units took months. Russian forces have advanced 20–25 kilometers since January 2026. They clearly intend to move directly toward the western outskirts of Kramatorsk, rather than toward distant Barvinkove to the north.Russia’s and Ukraine’s strategies for victory are clear. The Kremlin is trying to capture Kramatorsk and Sloviansk and show President Volodymyr Zelensky — and the Ukrainian people — that his resistance to handing over territory in the Donbas was futile. Kyiv wants to convince Vladimir Putin that Russia’s offensive toward Kramatorsk has no chance of success.Both armies have learned to advance under favorable conditions, but these successes take months of preparation. Advances of a few dozen kilometers a year require hundreds of thousands of soldiers on both sides and come at the cost of enormous casualties. What are all these people doing at the front?Infiltration instead of maneuver: What does the war actually look like — and why isn’t either side attacking in force?People fighting in the war can have very different experiences, as soldiers’ blogs and interviews show. Their experiences depend on the stretch of the front where they fight, their tasks and place in the military hierarchy, the stage of the war, the season, and other factors. Yet the experiences soldiers describe most often are similar in both armies. (Meduza published one of the best analyses of the personal experiences of Ukrainian soldiers who faced Russian drones in 2024 and 2025; Russian soldiers’ accounts of similar situations were much the same.)Alongside the individual accounts from the front mentioned above, this article draws on recent studies of combat experience by researchers who specialize in tactics. These include a series of podcasts by military analyst Rob Lee of the Foreign Policy Research Institute for the Ukraine Military History Institute channel, featuring Ukrainian soldiers and American volunteers fighting for Ukraine. On the Russian side, we draw on four volumes of Andrey Markin’s series “Obobshcheniye boyevogo opyta SVO” (“Lessons from Combat in the Special Military Operation”). The author wrote books and manuals on Russian and Soviet army tactics long before the current war. After the 2022 invasion, he took a job with the occupation administration in the Zaporizhzhia region and began interviewing Russian soldiers for a new series on the tactics of the war against Ukraine. We also use findings from field studies by several Western military think tanks, based on material collected during visits to the front and combat units’ headquarters.Over nearly five years of war, the nature of fighting has changed radically many times at the tactical, or lowest, level. But all these changes have followed an overall trend: Both armies’ reconnaissance and strike capabilities have grown as precision weapons and sensors that collect information about the battlefield and rear areas have become more widespread. Units’ ability to maneuver, meanwhile, has steadily declined. At each stage over the past two and a half years, the armies have become better at hitting the enemy with firepower but worse at outmaneuvering it. Yet throughout the 20th century, military theorists and practitioners developed both approaches to warfare, arguing that decisive victory was impossible without superiority in both firepower and maneuver.The loss of the ability to maneuver is a direct result of the proliferation of reconnaissance sensors and precision weapons at the front, with drones filling both roles. Sensors have made the battlefield and the immediate rear “transparent” — and, over the past year, areas farther behind the lines as well. This has made it impossible to concentrate the forces and equipment needed for maneuver near the front: first armored vehicles and mobile artillery, and eventually infantry too. Precision weapons and the rapid transmission of information from reconnaissance sensors to weapons systems have made it difficult for any troops or equipment to move in the open, both at the front and for several dozen kilometers behind it. Often, a single drone is both the sensor platform and the weapon.All these changes happened gradually, and their effects accumulated over years. Neither side willingly gave up maneuver. Both tried to devise tactics to counter the ubiquitous cameras and precision weapons.Both armies first tried to destroy the sensors and the positions of strike systems. In 2023 and 2024, both sides made extensive use of artillery, including precision-guided shells, to strike 10–20 kilometers behind the front line. They also used counterbattery radars capable of locating enemy artillery positions and engineering vehicles for remotely laying mines. But they soon had to abandon the mass deployment of equipment that was difficult to hide. Artillery is still used, but many of its tasks are now handled by less vulnerable drone operators, who also lay mines remotely, and by aircraft with glide bombs when a target, usually a fortified one, requires heavy munitions.Driving drone operators away from the front proved much harder: They are easier to hide. So the mix of weapons and personnel delivering strikes changed, but the problem remained. Concentrating resources near the front and maneuvering are still difficult, although both armies continue trying, as we discuss below. The alternative is to disperse infantry and have them move covertly — tactics that have become widespread on both sides.Until the end of 2024, attacks by relatively small assault groups of 10–15 soldiers still worked, as did transporting assault troops to enemy positions in armored vehicles. This was already a major departure from prewar ideas about maneuver, which envisioned entire mechanized units — companies and battalions in designated combat formations — going on the offensive. But relatively small groups could still maneuver: In the second half of 2024, Russian forces used assault groups and armored vehicles to advance rapidly through the southern Donetsk region, capturing a large area and reaching the borders of the Zaporizhzhia and Dnipropetrovsk regions. Ukrainian forces, meanwhile, tried to use rapid maneuvers by mechanized units to advance deep into the Russian army’s rear in the Kursk region.Starting in early 2025, even these options faded away, at least on the most active stretches of the front.Almost simultaneously, both sides learned to use precision weapons on a large scale against the areas immediately behind the enemy’s combat units. Russian forces cut off supplies to the Ukrainian force in the Kursk region, leading to its disorderly retreat. The Ukrainian army, meanwhile, tested a “drone wall” for the first time near Pokrovsk, disrupting Russian rear operations in that sector. The assault on the city was consequently postponed for almost a year.In 2025, logistical problems caused both sides’ forward assault units to “shrink” substantially: Supplying large assault forces and their support units near the front had become virtually impossible. Even mobile groups on motorcycles and ATVs, used in an attempt to support assault troops, struggled to reach the front. As a result, troops became very thinly spread along the line of contact, on both the attacking and defending sides.These conditions gave rise to infiltration by tiny groups, the dominant tactic throughout 2025 and the first half of 2026. Assault groups shrank first to three to five soldiers, then to pairs, and even lone fighters. Pairs move toward enemy positions under cover of bad weather or at dusk — less often at night, using special capes that shield them from enemy drones’ thermal imagers. Once several pairs have assembled, they try to storm a position: a trench system or a house in a settlement, likewise defended by a handful of soldiers. Some pairs never reach the positions because they fail to move undetected. But they draw enemy drones away, allowing the others to advance. A company commander, sometimes a battalion commander, directs the pairs by radio and tracks their movements with a drone. This is usually a Mavic with good optics, allowing it to remain far from the assault troops without giving away their movements.Previously, a group tasked with holding the captured position would move up immediately behind the assault troops. In 2026, the assault troops have to defend it themselves. Defense generally means simply trying to find reliable shelter from enemy strikes. Often, the captured position is no longer suitable because it was destroyed during the assault — for instance, by anti-tank mines thrown into trenches and dugouts. New shelters have to be built.Assault troops often have to remain in a new position for several days or even weeks. Supplies arrive by air or on radio-controlled wheeled or tracked ground vehicles.A countermeasure soon emerged. Ukrainian forces, more often on the receiving end of infiltration by Russian assault troops, tried to hold individual positions even when neighboring strongpoints had been captured and enemy pairs had moved into their rear. They also began infiltrating with their own small groups in response. In settlements, opposing positions can be on the same street without any exchange of small-arms fire.These tactics leave the two sides’ positions intermingled along the line of contact, inevitably leading to protracted fighting between concealed assault troops — most often a mutual hunt in which each side uses drones to find and kill the other. Only after every position in a sector has been cleared can drone operators move up to the new front line and the offensive continue.In the summer of 2026, the Russian command decided that small groups could use the sparsely held front to infiltrate covertly into the enemy’s rear, to depths of 10 kilometers or more. With luck, several pairs can even find and destroy poorly guarded positions of enemy drone operators or disrupt the enemy’s logistics. These infiltrations force Ukrainian troops to devote resources to searching their rear areas.Attacks on motorcycles and the use of armored vehicles to deliver assault troops to enemy trenches are becoming less common: For tiny infantry groups, stealth and concealment matter more than maneuver and speed.As a result, only a small proportion of soldiers even in dedicated assault units take part in assaults. A company, with an authorized strength of up to 100, assigns seven to 10 pairs to an assault. The rest try to organize supplies, install sensors such as cameras, build shelters in the immediate rear, search rear areas, and guard the approaches to drone operators’ positions. Formations such as brigades and divisions, as well as large units such as regiments, are deployed in depth. Troops are thinly spread along the line of contact, where only a small proportion of assault troops and short-range drone operators are based; on active stretches of the front, their positions are often intermingled with the enemy’s. The main forces are 10–15 kilometers behind the front: most armored vehicles, artillery, logistics, longer-range strike and reconnaissance drone operators, and support and command units. Those killed in combat are mainly assault troops taking part in infiltrations and members of the larger group of soldiers trying to keep supplies moving. The latter have to spend long periods outside shelter near the front line, in an area full of enemy drones.In recent months, both armies, frustrated with infiltration tactics that no longer produce the same results, have resumed experimenting with ways to neutralize the enemy’s weapons and reconnaissance systems. The goal is to reach enemy formations’ rear areas, 10–15 kilometers behind the front, and force entire formations to retreat.Counterproductive incentives: How do rewards for success harm both armies — and how do drones hinder their own infantry?Offensives under these conditions require a high level of organization and coordination between units. Such coordination often conflicts with the incentives built into the two armies’ structures.Most units in both armies are geared toward earning rewards, monetary and virtual, for hitting targets. They can also face punishment for poor performance if they do not hit targets regularly. This incentive system has a particularly strong influence on drone operators in tactical formations and units — brigades and regiments. An offensive, however, sometimes requires accepting lower performance on paper to allow other branches of the military, neighboring units, or both to operate.One common example is targeting enemy drone operators. To hunt down their counterparts on the other side, drone operators need time to search for concealed positions, which may require many sorties that produce no hits. Once a position is found, the incentives favor striking it immediately with their own weapons, even if passing the target to artillery or aviation would be more effective. But an immediate strike may work against the formation’s offensive as a whole: It is often better to synchronize strikes with the infantry’s advance. If struck immediately, enemy operators will move to backup positions and receive reinforcements if they suffer casualties. When the formation goes on the offensive, its forward troops will then come under drone attack from new positions that have not been located.Quickly locating drone operators’ positions remains difficult. Electronic intelligence systems, which detect devices emitting radio waves, are not accurate enough. Experienced drone operators also use remote antennas connected to their controllers by cables tens or hundreds of meters long. Sometimes a position can be found by intercepting a drone’s analog video feed: The images can reveal where the drone lands, if it is reusable, or, less often, where it takes off. But experienced operators set up launch sites away from their own positions and leave the job of placing drones on launch tables to less “valuable” soldiers. Prolonged aerial observation is therefore an effective way to locate the operators’ actual shelters. It can reveal supply routes that can be traced to the crews’ positions. This takes considerable time and many drone sorties.Another incentive that hinders coordination is units’ desire to protect themselves, sometimes at the expense of the overall offensive plan. For example, neighboring units may refuse to turn off their powerful electronic warfare systems to allow the units launching an offensive to use drones and other weapons without interference. They fear the enemy will exploit their temporary vulnerability while the jamming systems are off.Both armies make extensive use of sophisticated decoy equipment, dummy positions, and other deceptive devices and techniques that are difficult to distinguish from the real thing. There is often no incentive to work out what is a decoy and what is real equipment. Drone operators and other weapons crews receive credit for hitting realistic decoys. These strikes do nothing to help other units fight.Finally, coordination problems arise when the infantry attacks. Providing drone support for an attack — both to conduct reconnaissance and to strike enemy positions — is generally difficult and can sometimes even be harmful. Drone activity can reveal assault groups’ routes and the objectives of an offensive. Once the enemy detects the threat, it can concentrate its own reconnaissance drones and weapons in the threatened sector. Attacks often have to proceed without preparatory fire, with strikes on the targeted positions delayed until the last moment before the assault or abandoned altogether.Blinding the enemy: How do armies shape the battlefield for a successful offensive — and where should we expect fierce new battles?In recent months, both armies have been developing new tactical and operational methods to overcome defenses saturated with reconnaissance equipment and strike systems.In late 2025, Britain’s Royal United Services Institute (RUSI), one of the leading think tanks studying the Russia–Ukraine war, laid out how to organize effective offensives based on Ukrainian officers’ accounts. In 2026, Ukrainian forces put the theory into practice near Lyman.The method centers on the same preliminary “shaping of the battlefield”: Drone operators are concentrated to help locate enemy drone crews, electronic warfare systems, and artillery. Strikes on them must be synchronized with the infantry’s advance. Preparations begin with strikes on supply routes and the destruction of enemy reconnaissance drones in the air to “blind” the enemy temporarily. Ukrainian soldiers say that during Operation Vivaldi near Lyman, they did not even assemble a strike force, to avoid attracting the attention of Russian reconnaissance units. The offensive consisted of many small attacks and infiltrations.At the same time, the Russian army was using almost identical methods near Dobropillia. After several months of “shaping the battlefield,” one drone operator in that sector wrote on his Telegram channel in October: “Our little mill grinds slowly, but it grinds fine. We spent nine months choking off Dobropillia, and now it’s choked off. Vehicles don’t get in there anymore. They burn on the roadsides. Now it’s up to the infantry.” In the following weeks, Russian infantry captured this important city southwest of Kramatorsk.The new tactical method’s main shortcomings remain the slow pace of advance and the lack of maneuver that could break through enemy defenses to a significant depth. Near Lyman, Ukrainian forces ran up against a new defensive line along the Chornyi Zherebets River. Beyond the river, Russian drone operators and artillery await them again, with the artillery able to fire on targets as the front draws closer.Russian forces, meanwhile, have yet to advance north from Dobropillia. They are probably engaged in another round of “shaping the battlefield.”At Meduza, we are committed to transparency about our use of artificial intelligence in the newsroom. The story you’re reading was written by one of our living, breathing journalists and translated from Russian using an AI model configured to follow our strict editorial standards. This translation process is the result of extensive testing and refinements to ensure our English-language coverage is timely and accurate. A Meduza editor reviews every draft before publication.If you find any errors in this translation, please contact us at [email protected].To read Meduza’s exclusive content in English, please subscribe to our newsletter.