Most MMOs Are Ghost Towns. Here’s How to Spot One Before You Install It

Take every online game we track — nearly 4,900 of them — and line them up by how many people log in on an average day. The shape you get is not a gentle slope. It's a cliff.

Four games sit above 100,000 estimated daily players. Sixty-two more land somewhere between 10,000 and 100,000. Below that the floor drops out: 250 games in the low thousands, 456 in the hundreds, and 1,124 games with fewer than 100 people playing on a given day. Another 1,335 games in the catalogue produce no measurable activity signal at all — not a small number, an absent one.

The median game that we can measure has 47 daily players. That is the typical online game: not the one you've heard of, but a world with a few dozen people in it.

This is worth understanding before you spend a weekend downloading something, because the number that matters when you pick an MMO isn't how good the game is. It's how many other people are in it.

Half the genre's activity lives in 25 games

Concentration at the top is even steeper than the tail is long. The 25 busiest games account for roughly half of all the daily activity we can account for across the entire catalogue. The top 100 account for about 79%. By the time you reach the top 200 you've covered 89% of everyone playing anything.

The thresholds are humbling once you see them as ranks:

• Rank 50 is about 14,700 daily players
• Rank 100 is about 5,700
• Rank 200 is about 2,300
• Rank 500 is about 420

A game with 400 people playing every day is in the top 10% of everything we track. It will still feel empty when you're standing in its capital city at 2am.

Worth noting what's actually at the very top: Minecraft, Palworld, and Rust lead the list, with Escape from Tarkov just over the 100,000 line. Not one of them is a traditional MMORPG — which is a pattern we've written about before. The most-populated persistent worlds in 2026 mostly aren't the ones that call themselves MMOs.

Why the tail is so long

Three things built that cliff, and none of them are going away.

Persistent multiplayer got cheap to ship and stayed expensive to run. A small team can put a survival sandbox or a browser MMO into early access with a handful of people. Nearly 2,000 of the games in our catalogue released in the last seven years. Launching is now the easy part; funding a live team, a content cadence, and server infrastructure for a decade is the part almost nobody clears.

Population is the product. In a single-player game, a small audience is a business problem. In a persistent world it's a design problem: empty markets don't price goods, empty zones don't form groups, empty PvP has no one to fight. A shooter with 300 players is niche. An MMO with 300 players is broken in ways the developers never intended. That makes the failure mode self-reinforcing — thin population degrades the experience, which thins the population.

Players know this and pool accordingly. Choosing an MMO is choosing where to spend hundreds of hours, and the rational move is to join the game that will still be busy next year. Everyone reasoning that way at once produces exactly the distribution above. It isn't a market failure; it's the network effect working as designed.

Small is not the same as dead

Here's the important caveat: a low number is not automatically a verdict.

Plenty of long-running games operate happily at a scale that would read as catastrophic for a new release. A 25-year-old world with a stable few thousand regulars, a functioning economy, and a subscription base that covers its costs isn't dying — it's sized. Some of these games have been at roughly the same population for a decade. That's not decline, that's equilibrium, and equilibrium is more than most launches ever achieve.

What actually kills a game's playability isn't a low number. It's three other things:

• A falling number. Direction beats level. A game holding 2,000 daily players for three years is in far better shape than one that fell from 40,000 to 6,000 in eight months.
• A number spread too thin. 5,000 players across four regions and thirty shards is five separate ghost towns. The same 5,000 on one megaserver is a busy town square. Ask how the population is distributed, not just how big it is.
• A number the developer has stopped investing in. Population can plateau for years and still be fine. Population plus a dead patch notes page is a countdown.

The five-minute check before you install

You don't need our data to do this — though the popular and trending standings will save you most of the work. Here's the sequence:

1. Look at the trend, not the snapshot. Open the game's page and look at the last six to twelve months of activity. One number tells you almost nothing. A slope tells you what the next year looks like. Every game page on this site carries the full history of its social signals for exactly this reason.

2. Check the patch cadence. Find the official update log and look at the dates, not the contents. Monthly-ish for a year is a live game. "Server maintenance" three times in eighteen months is a game in maintenance mode, whatever the announcements say.

3. Read the marketplace, not the forums. If the game has a player economy, an auction house with thin listings and stale prices is the single most reliable tell there is. Economies need volume. They can't be faked by a marketing team.

4. Compare Discord online against Discord total. A server with 400,000 members and 900 people online is a community that already left. A server with 20,000 members and 2,000 online is alive. The ratio is far more informative than either number alone.

5. Ask how long it takes to group. In any game built around group content, the honest question in the community isn't "is it dead" — it's "how long is the queue for a dungeon at 9pm on a Tuesday." That answer, from a current player, is worth more than every "IS THIS GAME DEAD IN 2026" video combined.

What the distribution should actually change

If you take one thing from that cliff, make it this: the number of available MMOs and the number of viable ones differ by two orders of magnitude. There are thousands of persistent worlds you could install tonight. There are a few hundred where you'll reliably find someone to play with, and a few dozen where the world will feel full.

That's not a reason to only play the top ten. Some of the best worlds ever built run at four figures, and a small game with a stable, invested community will treat you better than a big one in freefall. But it is a reason to check the numbers before you commit the hours — and to know which numbers are the ones worth checking.

If you want the details of how the estimates behind all of this are produced, and where they're wrong, we've documented the method.