Steam On Linux Ended 2017 With A Slight Increase In Marketshare

[phoronix]

Phoronix: Steam On Linux Ended 2017 With A Slight Increase In Marketshare

With PUBG popularity appearing to decline a bit during December, the Steam Survey results for December 2017 show an increase in the Linux percentage,..

http://www.phoronix.com/scan.php?page=news_item&px=Steam-Survey-December-2017

[nuetzel]

Yee haw, celebrating. - New Year, new luck?!

[dungeon]

32bit Windows OS share is still bigger than Linux and OSX combined - 2.28% vs 2.17% if i counted right.

It is safe to prognose how we will beat that together in 2018.

[Veerappan]

I finally got the survey in Linux today for the first time since I rebuilt my home desktop..., in April... That's after signing in at least weekly for most if the year, and being signed in for a pretty good amount of time after each login.

I don't really know if there's truth to the survey being biased, but I wouldn't be at all surprised.

[Leopard]

It has Vulkan systems section too at last.

http://store.steampowered.com/hwsurvey/directx/

[franglais125]

32bit Windows OS share is still bigger than Linux and OSX combined - 2.28% vs 2.17% if i counted right.

It is safe to prognose how we will beat that together in 2018.

Hey welcome back :-P. You seem to disappear every 6 months and then come back out of the blue.

[dungeon]

Hey welcome back :-P. You seem to disappear every 6 months and then come back out of the blue.

Yeah, so many games, so little time

[Luke_Wolf]

I finally got the survey in Linux today for the first time since I rebuilt my home desktop..., in April... That's after signing in at least weekly for most if the year, and being signed in for a pretty good amount of time after each login.

I don't really know if there's truth to the survey being biased, but I wouldn't be at all surprised.

It's not that their survey is "biased" in the sense you mean it's that their sampling methodology leaves much to be desired, especially when they could do a population survey rather than sampling and get actually accurate results. While you can guess at the causes behind the trends the variance is just too high in general for the statistics to be considered reliable, and in spite of Michael's delusions otherwise these numbers cannot be considered Truth, and NetMarketShare and similar's statistics are vastly superior because while they have problems, their sampling methodology isn't completely broken.

[LaeMing]

I'm probably too late to be in last year's stats, but I am in for 2018! :-) I tried installing Steam on my 64-bit Debian box a few years back and gave up in frustration but this time it was a doddle to get multi-arch going for it. "Oxygen not Included" is a damned addictive game!! (I got hooked just from watching YouTube play-throughs!) :-D

[oooverclocker]

I don't see what shall be seen as "slight" about an increase of nearly 60% in relative Linux market share but at least it shows how fragile the steam survey is and there are so many systems in markets where you can't make reasonable revenues as a game developer. This might explain that the real revenue from Linux users usually seems to be more about 3% - closer to the numbers that netmarketshare measures (when you combine Linux + ChromeOS and assume the majority of "unknown" systems to be Linux systems).