You open the server browser, filter to vanilla because you want the actual game without eighty mods, and you get two thousand results. Great. You pick one with a decent name, load in, run to Cherno, and there's nobody there. Not quiet. Empty. So you try another one. Also empty. By the fourth you start wondering whether anyone plays vanilla any more.
They do. About two thousand three hundred people right now. The problem is where those people are, and it took me a while to work out how badly the server list hides it.
What I actually measured
Every few minutes our poller asks every DayZ server it knows about a simple question: who's on you right now. We keep the answers. That gives us two things a normal server browser can't show you.
The first is history. A browser shows a server's population at the instant you look. We can show you what that server did over the last week, which matters enormously, because a server sitting at 0/60 on a Tuesday afternoon might be packed every evening, and a server sitting at 0/60 might also have been empty since March. In the browser those two look identical.
The second is a mod score. We grade every server 0 to 100 on how far its mod list has drifted from the base game, and call it the Vanilla Scale. A completely unmodded server scores 0. A server running a total conversion scores in the 90s. Here's roughly what each grade means in practice:
| Grade | Score | Typical mod count |
|---|---|---|
| Vanilla | 0 | 0 |
| Vanilla+ | 3 to 20 | 4 |
| Lightly Modded | 20 to 45 | 7 |
| Modded | 45 to 70 | 17 |
| Heavily Modded | 70 to 90 | 42 |
| Extreme | 90 to 98 | 81 |
And here's how the live server population actually splits across those grades.
All of the numbers here come from real servers only. Roughly 86% of the servers the DayZ master list reports are fake population farms, and if you leave those in every figure becomes fiction. The distortion is not subtle:
Those farms report almost twelve million players between them. The real number is nineteen thousand. That's its own article.
The part I didn't expect
I started with the simple question: what share of servers is vanilla, and what share of players?
Vanilla is 17.7% of live servers but holds 11.8% of players. A real gap, and it explains why the two sides of this argument never agree. If you judge by scrolling a list, vanilla looks fine, roughly one entry in five. If you judge by finding a game, it feels thinner, because each of those servers holds fewer people.
That's mildly interesting and I nearly stopped there. Then I looked at the week instead of the moment.
For each server I took its highest player count at any point in the last seven days. Not the average, not right now. Its single best moment. A server that filled up once on Saturday night scores well. A server that never got anybody scores zero.
More than half of vanilla servers went a full week without one person logging in. Not a slow night. Seven days, nobody.
The median vanilla server's best moment of the entire week was zero players. That's what the median means here: line all 2,029 vanilla servers up by their best week, and the one in the middle had nobody on it at any point.
| Grade | Servers | Never had a player in 7 days | Median best moment |
|---|---|---|---|
| Vanilla | 2029 | 54.5% | 0 players |
| Vanilla+ | 183 | 41.5% | 1 player |
| Lightly Modded | 1485 | 37.8% | 1 player |
| Modded | 3859 | 34.0% | 1 player |
| Heavily Modded | 3254 | 23.6% | 2 players |
| Extreme | 632 | 17.9% | 2 players |
Read that third column from top to bottom and it's nearly a straight line. The more mods a server runs, the more likely somebody actually played on it.
Where the vanilla players actually are
Here's the number that made me rewrite this whole article.
There are 2,286 people on vanilla servers right now, spread across 2,029 servers. That averages out to about one player each, which would be miserable if it were evenly spread. It isn't.
197 vanilla servers hit 10 or more players at some point this week. That's 9.7% of them. Those 197 servers are holding 2,021 of the 2,286 vanilla players right now.
So 88% of everyone playing vanilla is on fewer than 10% of the vanilla servers. The remaining 1,832 servers are splitting 265 players between them, which works out to one player for every seven servers.
And you cannot tell which is which from a listing. These four vanilla servers all look much the same in a browser right now. Two of them filled to 60 players this week. The other two have not had a single person on them in seven days.
That's why your first four attempts failed. Pick a vanilla server at random and you have roughly a nine in ten chance of landing on one that nobody has meaningfully played on this week. You weren't unlucky. You were playing the odds and the odds are terrible.
Why vanilla specifically
The tempting conclusion is that people prefer mods and vanilla is dying. I don't think that's the main thing going on, and the mod counts in that first table are the reason.
Vanilla is what you get when you do nothing. Rent a box, it boots, and you have a vanilla server without making a single decision. Reaching Extreme means installing something like eighty mods, resolving the conflicts between them, retuning the loot economy, and repairing all of it every time Bohemia ships an update. Nobody arrives at eighty-one mods by accident.
So mod count is partly measuring how much the owner cares. A heavily modded server nearly always comes with a Discord, a wipe schedule, and somebody answering questions, because a person who did all that configuration work is a person who's invested. A vanilla server might have all of that too. Or it might be a box somebody spun up in March, played on twice, and forgot to cancel.
Both kinds sit in the same list looking exactly the same. Roughly half the vanilla list is the second kind, and they never disappear, because an abandoned server keeps reporting itself as online forever. It costs the owner a monthly fee and costs you the ten minutes you spend loading into it.
That's the real finding. Vanilla as a way to play is completely fine, and 72 vanilla servers hit 30 or more players this week, which are proper communities by any measure. Vanilla as a category in a server browser is where dead servers accumulate, and there's no signal in a normal listing to tell them apart from live ones.
How to actually find a good vanilla server
Stop judging by the number of players online when you look. That number tells you about this minute, and this minute might be 4am in the server's home region.
Judge by whether the server has a pulse over time. On dayzmetrics every server has a seven-day population graph. A server showing 0/60 with a graph that spikes to 40 every evening is a good server you happened to check at a bad hour. A server showing 0/60 with a flat line is not going to surprise you on day nine.
The practical version: set the Vanilla Scale filter to Vanilla, sort by rank rather than by current players, and skip anything with a flat graph. That takes you from roughly 2,000 candidates to about 200 real ones, and you'll find a game on the first try instead of the fifth.
Figures pulled 23 August 2026 across 11,444 live servers after fake population farms were excluded. "Best moment" is each server's highest recorded player count in the preceding seven days.