Breaking Down Russian Missile Salvos: What Drives Neutralization?

Remote Visualization

For almost three years, Ukraine has been subject to daily bombardments of Russian missiles and drones. These salvos extend beyond the frontlines to target critical infrastructure and wage a campaign of terror designed to compel Kyiv to surrender. Over this period, Ukraine has made significant advances in neutralizing these attacks by employing an improvised mix of equipment alongside advances in detection, electronic warfare, and integrated Western interceptors. Yet, what determines the rate at which Ukrainian defenders neutralize Russian missiles?

As part of the Russian Firepower Strike Tracker, this commentary uses data science to analyze the correlates of neutralization. Specifically, researchers in the CSIS Futures Lab show that it is not simply the mass of missiles fired that matters but rather their make and model. While there is a small but significant correlation between the total number of missiles launched on any given day and the neutralization rate, the mix of missile models—particularly high-speed ballistic missiles and hypersonics—better predicts how effectively Ukraine’s defenders safeguard their skies. In fact, the study reveals that short-range, ballistic missiles remain the deadliest weapon in the Russian arsenal, creating a need to rethink how to help Ukraine protect its sovereignty as it heads into tense negotiations where attacks are only likely to increase. Specifically, U.S. and European policymakers should bolster Ukraine’s ballistic missile defenses by integrating rapid detection, high-speed interceptors, and improved resource allocation. Ukraine, in turn, should expand its arsenal of advanced Western interceptors (e.g., Patriot and SAMP/T), invest in real-time intelligence sharing, and ensure sufficient interceptor stockpiles to sustain high-intensity salvos.

Russian Firepower Patterns and Trends

On average, Russia has fired 24.3 missiles and drones at Ukraine a day since the start of the war. Over time, Russia has significantly increased the diversity of missile models deployed to attack Ukraine. In certain months, over 25 unique missile and drone models were launched, with a daily maximum of 14 unique models (see Table 1 of the appendix). The monthly average of unique missiles launched is 16.08. This diversification includes ballistic missiles, cruise missiles, kamikaze drones, and reconnaissance unmanned aerial vehicles (UAVs).

Alongside this growth, there have been notable changes in the mix of weapons. Compared to the October 2022–September 2023 period, the overall intensity of launches has doubled, largely driven by the increased deployment of Shahed-131/136 drones. In fact, these drones, alongside Kh-101 and X-555 air-launched cruise missiles, account for 75 percent of all launched missiles. In contrast, Kalibr missile use has fallen dramatically, shrinking to just one-fifth of its previous level in the 2022–2023 period.

Neutralization Rate Patterns and Trends

Across the entire study period from September 28, 2022, to October 6, 2024, the overall intercept rate stands at 84.1 percent. This high intercept rate indicates that approximately 4 out of every 5 missiles or UAVs launched by Russian forces were intercepted by Ukrainian defense systems or went missing, likely due to electronic warfare capabilities.

Image
Yasir Atalan
Data Fellow, Futures Lab, Defense and Security Department
Image
Benjamin Jensen
Director, Futures Lab, and Senior Fellow, Defense and Security Department
Remote Visualization

Despite maintaining an overall high intercept rate, specific periods within the study timeline exhibited certain declines in daily interception effectiveness. Notably, between December 2023 and February 2024, intercept rates fell below the average. Given the increasing intensity of the missile salvos by Russians, these decreasing intercept rates tend to become more problematic in terms of aerial defense. Particularly as peace talks start to take place, Russia’s salvos with varied missiles to exploit gaps in Ukraine’s air defense will put extra pressure on the Ukrainian side at the negotiating table.

According to our model-level analysis in the appendix Table 2, Russia relies on a mixed group of expensive, high-precision missiles, as well as legacy systems and cheaper models that allow for constant daily pressure on Ukraine. While the neutralization data show that these legacy systems and cheaper drone models are not successful in hitting targets, they allow Russia to increase the intensity of the firepower during the offensives. In contrast, operationally, salvos that include high-speed ballistic (Iskander-M/KN-23) and certain cruise missiles (X-47 Kinzhal) prove most effective at striking targets. Strategically, this implies that as Russia escalates its reliance on fast, high-precision ballistic and cruise missiles, Ukrainian defenses face a growing challenge in preventing strikes. Our statistical analysis confirms that the technical characteristics of projectiles (such as speed and type) matter most in explaining the neutralization rate (see Table 3 in the appendix); simply put, faster missiles are harder to shoot down. At the same time, projectile type—whether a model is cruise, ballistic, or drone projectile type is important as ballistic missiles appear harder to intercept based on the data, as seen in Table 3 of the appendix.

Although attack volume helps explain the likelihood of a successful hit, its statistically small relationship with neutralization suggests that modest increases in volume matter less than extremely large salvos (e.g., over 150 missiles). Meanwhile, the diversity of missile models per day (unique models) is not a significant factor in predicting neutralization rates. In other words, operational measures like daily model variety are less important for explaining interception success than the individual technical features of the missiles themselves, and large salvos matter primarily when they reach exceptionally high numbers. For a more detailed discussion of the challenges in making inferences, please see A3 in the appendix.

From Data to Policy

Russia is almost certain to continue attacking Ukraine during forthcoming peace negotiations. These attacks, in turn, will see Moscow defaulting to a dark, old political art: bombing to win. As a result, Ukraine will need continued air defense support as it navigates negotiations. Specifically, the U.S. and European policymakers will need to increase Kyiv’s capacity to intercept ballistic missiles. This can take the form of expanding the delivery of advanced interceptors like Patriot, SAMP/T, and NASAMs, as well as supporting better intelligence required to track fast missiles in real time. These efforts should also continue deeper integration with Ukraine in terms of production, similar to the FrankenSAM and more recent Gravehawk adaption.

Even if politically, the decision to wind down its larger security commitments in Europe bolstering the arsenal of Ukraine becomes a key to buying down risk on the continent as the need to deter future wars in Europe and the Black Sea will persist. As a result, security cooperation activities that support both strengthening Ukraine’s negotiation position and setting conditions for a new peace should be defense policy priorities for the Trump team.

Yasir Atalan is a data fellow for Futures Lab in the International Security Program at the Center for Strategic and International Studies (CSIS) in Washington, D.C. Benjamin Jensen is a senior fellow in the Futures Lab at CSIS.

Appendix

A1: Missiles by Time

Remote Visualization

The analysis of Russia’s missile deployment strategy underscores a significant escalation in the diversity of missile models utilized against Ukraine over the period from September 28, 2022, to October 6, 2024. Daily, the number of unique missile models launched ranged from 1 to 14, with an average of approximately 4.08 unique models per day. In contrast, the monthly analysis reveals a much broader spectrum, with 5–26 unique models launched each month, with a mean of 16.08 unique models. This substantial increase in monthly diversity suggests a strategic intent to deploy a wide array of missile types consistently over extended periods, thereby enhancing the unpredictability of the offensive operations.

Missiles By Model

The intercept rates exhibit significant variability across different missile models, reflecting both the technological sophistication of the missiles and the corresponding defensive countermeasures. The table below summarizes the intercept rates for selected missile models by share launched.

Remote Visualization

Dominating this offensive are the Shahed-136/131 models, accounting for over 65 percent of all launches with 8,621 missiles deployed. This heavy reliance on a single missile type underscores Russia’s preference for proven and widely available weaponry to sustain continuous pressure on Ukrainian defenses. Additionally, models such as X-101/X-555 and Kalibr contribute significantly, together representing a substantial portion of the missile arsenal. The sheer volume of launches not only highlights the intensity of Russia’s military operations but also reflects its logistical capabilities to maintain a prolonged assault. The sustained high number of launches indicates a strategic intent to overwhelm defenses, degrade infrastructure, and assert territorial control, thereby shaping the broader dynamics of the conflict.

The analysis of intercept rates for Russian missile models deployed in Ukraine highlights a pronounced disparity in defensive effectiveness. A notable number of missile types, including Lancet, ZALA, Orlan-10, Supercam, Merlin-VR, Orion, and various UAVs, were intercepted entirely, demonstrating the robustness of Ukrainian defense systems against these specific reconnaissance UAVs. Conversely, models such as the X-22 and certain iterations of the Iskander series exhibited markedly lower intercept rates, indicating significant vulnerabilities against ballistic missiles. This variability suggests that while Ukrainian defenses are highly effective against certain drone technologies, they face challenges with others that may possess enhanced capabilities with increased speed.

A2: Correlates of Neutralization Rate

The literature on modern warfare highlights multiple factors that can shape the effectiveness of both offensive missile capabilities and defensive systems. These factors include technical characteristics of the projectiles—such as the type of munition (for instance, ballistic or cruise missiles, UAVs), speed, flight altitude, range, warhead type, and the presence of datalink or precision guidance—as well as the operational strategy behind their launch. The latter involves the volume of missiles deployed, the variety of types launched simultaneously, and the extent to which they form part of coordinated salvos that combine multiple missile types. Timing and conditions also play a role, as night versus daytime launches and weather can affect both the accuracy of incoming projectiles and the responsiveness of defense systems. Geographical factors, including the specific locations from which missiles are launched and the types of targets they aim for, can significantly influence interception success rates. Finally, defensive capabilities, including the availability and sophistication of air defense systems, ammunition stockpiles, electronic warfare resources, and personnel training, further shape overall neutralization effectiveness.

Our dataset to test this correlation is structured around missile launches, capturing key attributes that influence neutralization rates. It includes technical characteristics of the projectiles, such as speed, range, battle network link, and launch platform, which provide insight into how missile specifications impact interception effectiveness. Additionally, the dataset accounts for operational factors by measuring the volume of missiles launched and the number of unique missile models deployed within a given period, reflecting the complexity of an attack and the potential strain on defensive systems. The dataset also incorporates a temporal dimension by categorizing launches into quarterly intervals (using quarter-fixed effects), allowing for an analysis of seasonal changes in defensive capabilities and weather conditions that may affect interception rates. However, some critical factors remain unobserved, including precise launch and target locations, which could influence defensive response times and interception probabilities. Furthermore, the dataset does not explicitly capture whether missiles were launched simultaneously as part of a coordinated salvo, a factor that could significantly impact defense saturation. To analyze these dynamics, we employ regression models that assess the relationship between missile characteristics, operational strategies, and neutralization rates. Of note, these results do not imply any “effect” per se, instead, they aim to capture how associations between variables change with different model specifications.

Remote Visualization

A3: Challenges in Making Inferences

Before drawing conclusions with the available data, several threats to the analysis must be addressed.

  1. We cannot assume that Ukraine’s defensive capabilities remained constant throughout the study period. Any significant changes in defense systems, technical capacity, strategic locations, or personnel numbers could confound the relationship between missile launches and interception rates. Thus, to infer the impact of Russian firepower on the Russia-Ukraine war (as well as the impact of airstrikes on modern warfare), it is essential to account for potential variations in Ukraine’s defensive capabilities. While we aimed to capture this using time (quarter) fixed effects, this would still not capture these capabilities in a strong sense.
  2. Although the data reported by Ukrainian accounts is granular at a daily level, we do not have the exact timing of these launches in most cases. This means we lack detailed information on whether missiles were launched strategically as coordinated attacks (e.g., combining Shahed-131, X-22, and reconnaissance UAVs in short time frames) or independently by different battalions. Such strategic combinations could have a more substantial impact on defense systems compared to standalone launches. Therefore, when testing our hypotheses, we must ensure that launches with lower intercept rates are not systematically different in terms of Russian strategic approaches. Since the data does not allow us to test this, we should consider this.
  3. Weather conditions may influence intercept rates. Previous studies indicate that factors like cloud cover can limit the precision of drone strikes, potentially affecting interception effectiveness. It is important to verify that days with lower intercept rates do not systematically differ in weather conditions from other days, as otherwise, weather effects could be erroneously attributed to missile launch intensity and diversity. However, since we do not know the exact timing of the launches, we will not be able to match the weather data.
  4. Missile launch locations likely play a significant role in determining interception success. While our data includes disaggregated information on each missile launch and interception, it lacks precise details about launch locations. This omission could affect our understanding of how geographical factors influence intercept rates.
  5. The present data relies on reports from Ukrainian official military accounts. In an ongoing conflict, each side may have incentives to exaggerate or conceal information regarding their defensive capabilities. Therefore, any inferences made are contingent upon the accuracy and reliability of these reports.