Table of Contents
This week, a joint investigation by the independent news outlet Mediazona and the OSINT initiative known as the Conflict Intelligence Team (CIT) has cast doubt on the accuracy of Russian military casualty figures released by Ukraine’s military intelligence. The figures, which suggest that nearly 87,000 Russian soldiers died in the first eight months of 2025, were shared by the Ukrainian Defense Ministry’s project titled “I Want to Live.”
Mediazona highlighted the concerns expressed by both organizations regarding the reliability of these statistics.
The investigation utilized a comprehensive analysis of casualty records, employing forensic techniques similar to those used in the examination of electoral fraud.
Methodology and findings of the investigation
The Coalition Investigation Team (CIT) employed a statistical analysis to scrutinize reported military loss data, specifically looking for signs of manipulation.
Drawing from psychological studies, the team observed that when individuals are asked to generate random numbers, certain digits tend to be favored subconsciously. This leads to discernible patterns in the data. The aim of this methodology is to uncover irregularities that may indicate the figures were not produced randomly.
During the examination of casualty figures, CIT identified several anomalies that required additional scrutiny. Notably, the distribution of the final digits in the reported numbers deviated significantly from expectations associated with a genuinely random sequence. The analysis indicated an overrepresentation of the digits one and two, while the digits six and seven appeared disproportionately less frequently.
Statistical validation of anomalies
To quantify these findings, researchers utilized Pearson’s chi-squared test, a statistical method that evaluates the relationship between observed and expected distributions. The results revealed that the probability of the data being random was only 1.3 percent.
When the data was further divided into two categories—primary (raw, unit-level) and aggregated (sums, totals)—the differences became more evident. In the primary data set, the chance of observing such a pattern by random occurrence decreased to just 0.4 percent.
This led the Center for Investigative Journalism (CIT) to determine that the figures may have been compiled individually for each military unit before aggregation. This could have distorted the distribution of the last digits, particularly for zeros and fives. Such a process likely contributed to the uniformity noted in the final totals.
Mediazona’s perspective on the data
In a biweekly update on the losses suffered by Russian forces, Mediazona’s editor, Dmitry Treshchanin, aligned his comments with the concerns raised by the Conflict Intelligence Team (CIT). He suggested that the Russian military might be involved in altering the records. Treshchanin stated, “It’s hard to envision a Russian officer who doesn’t alter the records in some way.” He also pointed out significant omissions in the leaked data, particularly the lack of information regarding desertions, which further casts doubt on the reliability of the records.
Irregularities in documentation
From a military bureaucratic perspective, several irregularities arise from the leaked documents. There is a notable absence of clear recipients and an evident purpose for the records. Furthermore, details such as who compiled the data, when it was generated, or where it was filed are missing, raising additional questions regarding its authenticity.
Future assessments of casualty figures
Despite the findings and skepticism surrounding the leaked numbers, Treshchanin emphasized that Mediazona and its collaborators currently lack the means to either confirm or refute the assertion that approximately 87,000 Russian soldiers were killed in Ukraine during the first eight months of 2025. He stated, “We won’t be able to reliably assess these figures until 2026, likely by next fall.” This limitation arises from Mediazona’s methodology, which relies on published obituaries and public probate records to estimate military casualties.
The leaked data presents a staggering figure regarding Russian military losses; however, the ongoing investigation by Mediazona and CIT raises significant doubts about its reliability. As developments unfold, further analysis may clarify the true extent of casualties in the conflict.
