Measured, not estimated
How many people still play these games
Every figure on this page was measured by JoinGG. Nothing here is scraped from another site, estimated, or extrapolated to servers we do not list. We query all 5,542 listed servers every hour, directly over the protocol each one speaks — Source and GoldSrc servers over Valve's A2S, Minecraft over its server list ping, FiveM over its info endpoint — and record what each server answers. The counts below are those answers. They describe the servers listed on JoinGG, which is not every server in the world: read them as a large sample of public community servers, not as a census of a game.
Last sweep Sep 3, 2026, 12:34 PM UTC. This page rebuilds every five minutes.
- 26,017
- Players connected
- 5,463
- Servers answering
- 5,542
- Servers listed
- 31
- Games covered
reported by the servers themselves
79 listed but silent
every one queried directly
of 147 you can list
These figures rest on 310,260 status checks recorded over the last 14 hours, of which 8 hours cover enough of the list to compare with the others. Readings are kept for 7 days, so this page gets deeper as they accumulate rather than longer.
Players across every listed server
310,260 status checks · 15 hours measured
- Busiest hour measured
- 24,126 at Thu 11:00 UTC
- Quietest hour measured
- 19,423 at Thu 05:00 UTC
- Latest sweep
- 25,673 players
- Hours plotted
- 8 of 15
7 hours are marked but not plotted. The poller reached under 90% of the list in those hours, so the totals are low because fewer servers were asked — not because fewer people were playing. Drawing them on the same line would invent a drop.
By game
Servers, how many are answering, and how many people are on them right now. Every game name links to its hub, where the individual servers behind the number are listed.
| Game | Servers | Online | Players here | Slots used | In game (Valve) | Busiest hour |
|---|---|---|---|---|---|---|
| Counter-Strike 1.6 | 1,591 | 1,563 | 10,941 | 25.1% | 7,668 | 2 of 3 days |
| Counter-Strike: Source | 597 | 593 | 4,006 | 19.4% | 1,858 | 2 of 3 days |
| Counter-Strike 2 | 854 | 838 | 3,354 | 12.1% | 772,095 | 2 of 3 days |
| Team Fortress 2 | 568 | 565 | 2,123 | 12.5% | 48,881 | 2 of 3 days |
| Garry's Mod | 352 | 350 | 2,058 | 10.6% | 15,183 | 2 of 3 days |
| Counter-Strike: Global Offensive | 272 | 266 | 758 | 11% | — | 2 of 3 days |
| Left 4 Dead 2 | 443 | 439 | 480 | 11.7% | 13,197 | 2 of 3 days |
| Day of Defeat | 80 | 79 | 389 | 22.4% | 76 | 2 of 3 days |
| Day of Defeat: Source | 56 | 56 | 358 | 24.1% | 190 | 2 of 3 days |
| Half-Life | 205 | 201 | 334 | 6.7% | 475 | 2 of 3 days |
| Counter-Strike: Condition Zero | 49 | 48 | 278 | 21.5% | 309 | 2 of 3 days |
| Team Fortress Classic | 44 | 42 | 270 | 24.3% | 29 | 2 of 3 days |
| Half-Life 2: Deathmatch | 146 | 144 | 216 | 6.6% | 69 | 2 of 3 days |
| Natural Selection | 16 | 16 | 131 | 30% | — | 1 of 3 days |
| Insurgency (2014) | 28 | 28 | 74 | 19% | 114 | 2 of 3 days |
| Minecraft | 72 | 68 | 53 | 0.3% | — | 2 of 3 days |
| Insurgency (Source) | 8 | 8 | 43 | 19.1% | — | 1 of 3 days |
| Rust | 24 | 24 | 28 | 1.1% | 49,311 | 2 of 3 days |
| Space Engineers | 12 | 12 | 28 | 7.1% | 1,963 | 1 of 3 days |
| Zombie Panic! Source | 15 | 14 | 27 | 8.6% | 20 | 2 of 3 days |
| Left 4 Dead | 58 | 58 | 24 | 5.1% | 282 | 1 of 3 days |
| Fortress Forever | 6 | 6 | 24 | 14.8% | — | 1 of 3 days |
| 7 Days to Die | 29 | 29 | 16 | 2.7% | 24,544 | 1 of 3 days |
| SA-MP | 3 | 3 | 2 | 0.9% | — | 2 of 3 days |
| Alien Swarm | 6 | 6 | 1 | 1.1% | 6 | 1 of 3 days |
| Killing Floor 2 | 1 | 1 | 1 | 16.7% | 4,028 | 1 of 3 days |
| ARK: Survival Evolved | 2 | 2 | 0 | 0% | 16,371 | 1 of 3 days |
| Contagion | 2 | 2 | 0 | 0% | 13 | 1 of 3 days |
| Conan Exiles | 1 | 1 | 0 | 0% | 4,849 | 1 of 3 days |
| FiveM | 1 | 0 | 0 | — | — | 2 of 3 days |
| Red Orchestra 2 | 1 | 1 | 0 | 0% | 80 | 1 of 3 days |
The two player-count columns measure different things
Players here is ours: people connected to the servers JoinGG lists, counted by asking each server. In game (Valve) is Steam's live count of everyone running that game anywhere — official matchmaking, single-player, community servers and idle menus included. Neither is a subset of the other that we can compute: our list is not every community server in existence, and Valve's total is not limited to multiplayer. So we print both and do not divide one by the other.
Valve's figures come from the Steam Web API and are cached for an hour, so they can lag ours by up to that much. Counter-Strike: Global Offensive has no Valve column because Steam replaced CS:GO with CS2 under the same app id — there is one number for the two of them, and it is on the CS2 row. Minecraft, FiveM and the standalone mods have no Valve column because they are not separate Steam products.
Where our count is the bigger number
On Counter-Strike 1.6 (10,941 vs 7,668), Counter-Strike: Source (4,006 vs 1,858), Day of Defeat (389 vs 76), Day of Defeat: Source (358 vs 190), Team Fortress Classic (270 vs 29), Half-Life 2: Deathmatch (216 vs 69) and Zombie Panic! Source (27 vs 20) we can see more players on the servers we list than Valve reports for the entire game. That is not a mistake in either figure. Server queries count bots as players, because the protocol reports them as players — a bot-filled rotation looks busy to us. And a large part of the surviving Counter-Strike 1.6 and Day of Defeat population plays on non-Steam builds, so those players are on the servers we query and invisible to Steam. Both effects push the same way, and on twenty-five-year-old GoldSrc games they push hard enough to cross over.
Why the busiest-hour column is empty
Naming the hour a game fills up needs at least 3 days of readings behind it. We have 2 days. With less than that, the busiest hour is just the hour we happened to be measuring in — a fact about our poller, not about the game — so the column stays empty until the answer is about the game. It fills in on its own.
What they are actually playing
The counts above are people on servers, and servers run modes. These are the game and mode combinations with at least 5 servers behind them, taken from what the servers report they are running.
Readings are kept for 7 days and pruned by the poller, so this page describes the last week rather than all time. Peak hours are given in UTC — for example 20:00 UTC — because the servers behind them sit in a few dozen countries and a local clock would be wrong for most readers.