Valve amended the Steam survey for December 2025 - Linux actually hit another all-time high

https://www.gamingonlinux.com/2026/01/valve-amended-the-steam-survey-for-december-2025-linux-actually-hit-another-all-time-high/

Initially Valve's Steam Hardware & Software Survey for December 2025 showed Linux at 3.19%, but they appear to have amended it with a nice boost for Linux. There were a few issues I spotted like languages and such not adding up, but now they do so it appears there's a few corrections that were made to it this month.

Now the latest figures from Valve show for December 2025:

Windows: 94.23%
Linux: 3.58%
macOS: 2.18%

179 points · 10 comments · view on lemmy.world

10 Comments

ClassyHatter@sopuli.xyz · 42 pts · 253d (1 reply)

So, from Jan'25 to Dec'25, Linux market share grew from 2.06% to 3.58%. That is about 74% growth: (3.58% - 2.06%) / 2.06% * 100% = 73.786...%

Same numbers for 2024 gives about 17% growth for that year: (2.29% - 1.95%) / 1.95% * 100% = 17.435...%

hperrin@lemmy.ca · 24 pts · 253d

Year of the Linux desktop and all that.

Die4Ever@retrolemmy.com · 19 pts · 253d (6 replies)
Creat@discuss.tchncs.de · 7 pts · 253d (5 replies)

The heck happened in March?

Danitos@reddthat.com · 14 pts · 252d

Survey error. They oversampled the Chinese market. In fact, you can also see a huge increase in Chinese language whenever there's a huge drop un Linux marketshare. It hashappened 2 times that I can remember.

GraveyardOrbit@lemmy.zip · 11 pts · 252d
[ removed ]
Die4Ever@retrolemmy.com · 6 pts · 252d (2 replies)

could've been a flawed survey like this one but Valve never fixed it?

Creat@discuss.tchncs.de · 4 pts · 252d (1 reply)

I mean there's isn't anything to fix. Could just be an unlucky month in selecting people for the survey. As the other person said, it does seem to be a statistical anomaly or actual error. But that is still a pretty massive dip for either of those two options. In and of itself also kinda statistically unlikely, hence my question.

Die4Ever@retrolemmy.com · 3 pts · 252d

The December survey was bugged, March could've been similar but never fixed. Before this fix, December had numbers that were adding up to more than 100% so it was a bug not just bad luck, they didn't even redo the survey they just reprocessed the same data.

WillowBe@lemmy.blahaj.zone · 18 pts · 253d

Nice! thats pretty cool

RandomStickman@fedia.io · 9 pts · 253d

I remember sering it hit 1% and the big bump with the deck. Great stuff