Mapping the war—what data reveals about Ukrainian losses

The war has faded from the headlines. The names keep coming

Hi, I’m Roman Mohuczy. This is The Root Access — a newsletter about data journalism and AI tools, built on real cases. Each publication presents original findings or explains the methods behind them.

This piece is the first of two. The next will pull back the curtain on how this dataset was constructed — the scraping, the cleaning, and the verification — and show how you can apply the same methods to your own research.

***

Russia’s invasion of Ukraine appears to be Europe’s deadliest conflict since 1945. Despite efforts by both sides to obscure military casualties, tens of thousands of confirmed names have been recorded on lists documenting those killed in the war.

Volunteer groups in both countries have been compiling and verifying named records of soldiers and officers who have died in the conflict. On the Russian side, this work has been carried out by BBC Russian in collaboration with the independent media outlet Mediazona, which coordinates volunteer researchers. Ukrainian losses have been documented by the anonymous project UA Losses. The group did not respond to my request for comment.

Ukraine blocked the UA Losses website in March 2024 following an order by the State Service of Special Communications and Information Protection under martial law legislation. Authorities accuse the project of spreading false information about military casualties. Kyiv’s official death toll contains significantly fewer names than those listed by the website.

Do you know anything about the UA Losses project? I would love to hear from you. You can message me securely on Signal at mohuczy.36 or send me an e-mail at mohuczy@proton.me.

Why do I consider this source credible?

I parsed all entries published on the website, cleaned the dataset, and manually verified 384 randomly selected names from a total database of approximately 186,000 records. This yields a 95% confidence level in the overall validity of the dataset. The following graphics present the main findings drawn from this data.

I

More than 185,000 KIA names? Highly unlikely

Media reports often state that the UA Losses database contains 186,000 KIA names. In practice, the number is lower.

Based on this dataset, roughly 174,000 individuals have been recorded as killed or missing since the start of the full-scale invasion.

However, the list includes fewer confirmed deaths than missing personnel: 81,734 are recorded as deceased, compared with 92,576 listed as missing. The missing category may include deserters, civilians, or individuals who were later found alive. UA Losses does not disclose how — or whether — this part of the list is verified or cleaned.

II

This war is not for the young and affluent

The median age of Ukrainian soldiers killed in action or listed as missing is almost 40, the dataset shows. Most came from small towns and villages. Although nearly 40% of Ukraine’s population lived in large cities before the full-scale war, only 23.5% of those killed or missing were born in cities, while 76.5% were born in smaller towns. It is worth noting that average poverty rates in small towns in Ukraine are higher and more stubborn than in large cities.

Nearly 40% of Ukraine’s pre-war population lived in large cities, yet 76.5% of those killed or missing were born in smaller towns.

This map shows the birthplaces of soldiers who were killed or went missing. I geocoded each city and village from the dataset — nearly 17,000 distinct locations across Ukraine.

UA Losses does not cover two regions: Crimea and the city of Sevastopol. The peninsula and its largest city — which is treated as a separate administrative region — were annexed by Russia in 2014. It is unclear whether UA Losses excludes these regions by design or whether there are simply no recorded losses from there.

According to the BBC’s database, however, at least 1,748 soldiers and officers from Crimea have been killed in action while serving in the Russian army.

III

All roads lead to Donetsk, unfortunately

As the birthplaces of these individuals are widely spread across the country, nearly all death locations in this war are concentrated in Donetsk and Kharkiv oblasts. The fatalities visualisation makes this clear: the front lines are visible from the KIA dots alone. Those who died deep within the country died in hospitals or were killed by drone and missile strikes.

IV

There’s nothing quiet on the Eastern front

The UA Losses list contains almost no names of soldiers who died in 2026. This is likely linked to processing capacity: the project team requires significant resources to add new names quickly enough, and those resources are probably not available.

Even based on the existing data, 2025 was the deadliest year for the Ukrainian army: 31% of the death toll and 43% of missing entries were added to the database that year. That share will likely increase as new names are added.

The BBC list confirms this trend on the Russian side as well. The war has not wound down; it has simply faded from the headlines. And, honestly, I don't know what's worse: the war continuing to fade from the news, or the news from this war coming back, again.

---

A quick note on transparency: I spent a brief period volunteering for the BBC/Mediazona casualty project. These findings are entirely my own.

If this kind of work is useful to you — the data, the methods, or both — subscribe. The next post covers how this dataset was built, geocoded, and visualised.

And if you want to support independent research, you can buy me a coffee, even with crypto.