KDA 50 with Zero Deaths: The Dota 2 Record Ordinary Players Will Never Touch — and Why It Means Nothing

I write this article for you to argue with me, not to agree with me.

When I was 14, the 2026 World Cup taught me that underdogs don't win through miracles. But today, I'm not talking about football. I'm talking about a number that anyone who has ever played Dota 2 knows is nearly impossible: a KDA of 50 with zero deaths in a professional match.

People call it delusion; I call it a hypothesis worth testing.

Context: The records exist, but no one is telling you the full story

In August 2026, at the PGL Arlington Major, bzm — OG's midlaner — finished a match with a 27/0/23 line on Ember Spirit in 44 minutes of play. KDA 50, zero deaths, at a Major in the Dota Pro Circuit system. Not a friendly. Not a pub game. The biggest stage before The International.

But bzm wasn't the first to do it.

In 2026, at the China qualifiers of SL i-League StarSeries Season 3, Shirley — CAVALRY's support — also achieved KDA 50 with zero deaths on Bounty Hunter, posting a 10/0/40 line. The match lasted only 22 minutes. While bzm built his record through 27 kills, Shirley did the same through 40 assists — two completely different paths to the same summit.

And then there's Vol. Also in 2026, at the ESL One Katowice qualifiers, Blue Pikachu's offlaner finished a match with 27 deaths on Centaur Warrunner in 51 minutes — the reverse record nobody wants to break.

The story media tells you: bzm is a genius, Shirley is a silent legend, Vol is a disaster. The story I'm about to tell: all three are misunderstood.

Core: KDA is the most deceptive metric in Dota 2

From my six years of watching professional matches, I've realized something the stat sheet never reflects: Dota 2 is not a game of kills. It's a game of decisions.

Look at bzm's numbers again: 27/0/23. Impressive. But ask yourself: in those 44 minutes, how much space did bzm create for his team? How many times did he force the enemy to react? These numbers don't appear in KDA.

What about Shirley? 10/0/40 on Bounty Hunter. A support who never died in 22 minutes. Sounds perfect. But I'd bet that in that match, there were moments Shirley should have sacrificed his life to protect the carry — but didn't, because he was chasing a pretty number. That's why I'm always suspicious of zero-death performances from supports.

What interests me most is Vol's case.

27 deaths in 51 minutes. On average, dying once every two minutes. To the casual viewer, this is the worst performance in history. To me — someone who has watched hundreds of Centaur Warrunner games — this is an offlaner doing his job: sacrificing himself to create space, drawing all the enemy's attention so the carry can farm.

One lost teamfight is worth more than a boring win.

The data supports me: Centaur Warrunner is the tankiest hero in the game when properly itemized. A Centaur player dying 27 times isn't because he's weak — it's because the enemy has to pour all their damage into him to stop him from initiating. Every time Vol died, his team gained 30-60 seconds to take objectives or recover. That's value that doesn't show on the scoreboard.

The empty arenas of 2026 were a data laboratory nobody asked permission to use. And from that laboratory, I learned: home advantage is only 33% from the crowd. Similarly, a Dota 2 player's value is only about 30-40% reflected in KDA. The rest — vision, space creation, initiation, shot-calling — is invisible to the casual viewer.

flyfly, a Chinese carry player, once achieved KDA 49 — just one point short of the record. Does anyone remember that match? I bet not. Because KDA 49 isn't a round milestone. But it proves a point: the line between record and near-miss is a single decision in a match lasting dozens of minutes.

Puppey once had a match with zero deaths and 30 assists. SumaiL has had impressive KDA stretches across multiple games. Illidan, Madara, TaiLung — all have had memorable statistical moments. But none of them are remembered for those numbers. They're remembered for titles, for decisive moments, for what they did when the match was on the line.

Contrarian: Where could I be wrong?

Of course, I could be wrong. Maybe bzm really is an absolute genius, and his KDA 50 truly reflects superior skill. Maybe Vol really is a bad player, and 27 deaths is undeniable proof of incompetence.

But if so, why has Dota 2 — a game with an extremely developed data analysis ecosystem via OpenDota, Stratz, Dotabuff — never created a unified metric to evaluate player value? Why do we still use KDA, a formula from the Warcraft 3 era, to evaluate players making decisions hundreds of times more complex?

Maybe because KDA is easy to understand. Maybe because media is lazy. Maybe because audiences don't want to learn something more complicated.

I choose to believe the third possibility.

Open Ending: What are records for?

Records by bzm, Shirley, and Vol will live in Dota 2's record books — possibly forever, because these numbers are nearly impossible to replicate. But their true value isn't in who achieved them; it's in the question they pose: What are we using to evaluate players?

A support player who dies 27 times but creates the win for his team — is he a bad player? A midlaner with a perfect KDA who doesn't pressure the map — is he a good player?

I don't have the answer. I only have data, and the data says: the answer is more complicated than you think.

Want to go further? Ask yourself: by what measure do you evaluate Dota 2 players — and why?