[veber-alex]

I wouldn't be too exited. Statistics like this are very problematic.

For example, I have Steam installed on my Macbook pro and I occasionally play a single very simple game there. Does that make me a macOS gamer? of course not. The vast majority of games I want to play don't work on macOS.

I suspect that most of those 5% are just Linux users who have steam installed and play a small amount of games. Some probably just installed it to check what's available and don't play anything.

Everyone I know who is a "serious" gamer, as in exited about upcoming releases of AAA games is using Windows.

[dainank]

Indeed. The bigger problem is also that consistently the most played games are multiplayer competitive titles with anti-cheat software that is only written for Windows (and sometimes MacOS). I suppose this issue will solve itself, once enough people start playing on Linux. Then developers will be forced to support that too in order to not lose too much of their player base, but we are still a far cry from this threshold.

[Doxin]

I'm not convinced "most played" is the relevant criteria here. That's easily skewed by games which are time-sinks, like PvP games with draconic anti-cheat and DRM. That doesn't make those games more important than games which absorb less playtime.

[npodbielski]

That would mean that it still would be around 0,5%. If you want to split the hair probably 4,5% of this 5% is Steam Deck.