Welcome to the “Killing Floor” franchise state of the game page. One of the main reasons it exists is the mining & trading bots infestation of KF2, which makes the Steam “in-game” concurrent players number inflated and unreliable, as was reported here: Dead game: Tripwire Interactive’s Killing Floor 2 “In-Game” player base count is inflated by bots. For the first time in many years we can see how many gamers actually play KF2 concurrently, the true “in-game” concurrent players number. This page is being currently worked on and new graphs measuring the game in various ways are being added. Y axis is all the concurrent KF2 players number in the world, X is time, not much different from the ones you can find on websites like “steam charts” or “steam DB”. The only difference is their graphs are fake as they are inflated by the bots, while this one counts the real online players only.
The method used here does NOT count: players who launched the game and sit in the main menu or players who play “offline”, i.e. in the locally generated world, not on any online servers. The number of such players, however, is quite small and quantified in the link above as being about 5% of the online concurrent players. It also does not count any console players, although their number by now is extremely small, it’s very hard to assess quantitatively, my own impression is both PS and Xbox console players combined constitute less than 10% of the online concurrent PC players. The method does count both PC platforms, which are Steam and Epic Games Store, however, the amount of EGS players is also quite small.
An alternative result of applying this methodology is available from KF2 player and modder known as “baz” on his KF2 global stats page.
As it typically happens, a number of retards suggested that these graphs are fake and Carl is full of shit. Well, they are not and Carl, as usual, conveys truth, full truth and nothing but truth. Still, I would like to take time and address some of the “arguments” so to speak.
- “Hurr-durr, these graphs are too jerky and don’t look like the ones on SteamDB or SteamCharts!” Retard, I take measurements every 11 minutes, which is the most frequent and accurate measurements anyone in the world does. I could go and take it even more frequently, if there was benefit to it, but I don’t see any. If you check the charts on SteamDB and SteamCharts, they take the measurements only once an hour, which is ~5 times less accurate. This is retarded as it pertains to KF2, as “short” 4 waves games are over in <20 mins, “long” games last for ~45 or so mins, so if one takes measurements only once an hour, they surely miss whole games pass by and players joining the fray and leaving for good. Moreover, when they build their graphs, they rely on the graph smoothing, which is pleasant to an eye, but results in inaccurate data visualization and through that only such graphs are deceptive, as they aren’t literally edgy enough. If SteamDB or SteamCharts didn’t use smoothing and were taking the measurements more frequently, their graphs also would look edgy and jerky, but they prefer appearance, while I prefer essence, therefore, I am necessarily edgy as my graphs are;
- “There are drops to 0, no way there were 0 concurrent players!” Retard, this is what Steam reports, which collects the server advertisements. Sometimes either there’s Steam or EOS or TWI backend servers downtime which results in nobody able to join any online servers, which happened in December 2025 and which is a result of the retarded TWI employees not being able to maintain high enough availability for their services. These were literal downtimes when nobody could play the game. Other times, e.g. in January 2026, there were short Steam API outages which resulted in Steam giving away all zeroes. I could have dropped these records from the database to make the graphs smooth and pleasing to an eye. But that would be against the policy of conveying truth, full truth and nothing but truth, so I keep these and will do so in the future;
- “It’s apples to oranges, because Steam counts all the instances of concurrent games running, while Carl counts only players on online servers!” Well, I agree that the comparisons to SteamCharts and SteamDB or to KF1 or KF3 numbers aren’t exactly 1:1 in terms of methodology, but it’s close enough. It’s more like comparing apples of different cultivars. This argument has some merit, but just a little. Yes, I count only online players that joined any online servers and which are nigh guaranteed to not be bots. This misses players who play “offline” or sit in the main menu (as well it misses the “Idle Master” or “ASF” bots). But I make this disclaimer and assess the number of such players independently to be rather small and not changing the overall picture much. Of course, we wouldn’t have the problem of different methodologies at all, if TWI did something about the bots. They don’t care though and even if they did, they wouldn’t know what to do about it, as they are all mediocrities.
