Valve has now released the latest Steam Hardware & Software Survey for April 2023, and it shows that things have calmed down again.
I reported earlier last month, that the details they had for March 2023 looked really off. It showed a massive surge in Simplified Chinese as a language on Steam with it becoming top, along with Windows 10 surging up over Windows 11 as well. It was all a bit odd. Valve never issued any corrections though, or announced any issues with it at all.
For April 2023, things have settled down again with Linux back to 1.32% (+0.48%), macOS back to 2.30% (+0.89%) and Windows at 96.38% (-1.37%). Languages changed dramatically again too with Simplified Chinese dropping back down to 25.03% (26.60%) so English being once again the top language used on Steam.
When filtering the April 2023 survey over to Linux here's the current most popular ways to game:
- SteamOS Holo 64 bit 22.77% +1.57%
- Ubuntu 22.04.2 LTS 64 bit 10.84% +1.23%
- Arch Linux 64 bit 9.79% -0.57%
- Freedesktop.org SDK 22.08 (Flatpak runtime) 64 bit 7.61% +0.51%
- Manjaro Linux 64 bit 6.25% -0.70%
- Linux Mint 21.1 64 bit 4.97% +4.97%
- Pop!_OS 22.04 LTS 64 bit 3.86% +3.86%
- Other 33.90% -10.89%
So it's clear the Steam Deck is currently beating all other ways on Steam for playing games on Linux.
You can view the data over time on the GOL Steam Tracker.
Quoting: tuubiThe mere fact that the system is that easy to mess with makes me wonder anew about its accuracy.Quoting: HighballQuoting: GuestThat's it, the Steam survey is definitely skewed against Linux users: out of the ~120 million monthly active users on Steam, 7 (seven) of them commented on gamingonlinux.com lamenting they never, or very rarely, received a survey.
Clearly, this should be a priority for Valve and fixed ASAP!
6 of them. I gave instructions for making sure a survey would occur. No need for "lamenting" when you are the one finding the solution.
I don't know if you're solving anything or just messing with the system.
Quoting: GuestThat's it, the Steam survey is definitely skewed against Linux users: out of the ~120 million monthly active users on Steam, 7 (seven) of them commented on gamingonlinux.com lamenting they never, or very rarely, received a survey.Haha, I wasn't lamenting it, just making an observation because several others also had. This made me laugh a bunch, so have a Like.
Clearly, this should be a priority for Valve and fixed ASAP!
Quoting: PenglingYeah, I wasn't lamenting either, just pointing out my experience. I used to care about it more, but over time I've mellowed out. I almost added an addendum about how humans sometimes find true randomness "less random" than a pseudo-randomness that smooths out weird things like one person getting the survey every month for a year and another not getting it for two years, but didn't have time.Quoting: GuestThat's it, the Steam survey is definitely skewed against Linux users: out of the ~120 million monthly active users on Steam, 7 (seven) of them commented on gamingonlinux.com lamenting they never, or very rarely, received a survey.Haha, I wasn't lamenting it, just making an observation because several others also had. This made me laugh a bunch, so have a Like.
Clearly, this should be a priority for Valve and fixed ASAP!
Like, you can flip a fair coin and get 10 heads in a row. It's quite a small probability to happen, but it's technically just as random an outcome as getting 5 heads and 5 tails, or 10 tails. But people tend not to see the extreme cases as random, even if they really did happen randomly. (One problem, of course, is that you can't ever really prove something did happen randomly.)
Last edited by Philadelphus on 3 May 2023 at 5:16 am UTC
Quoting: HighballIf you close steamYeah, this works -- even though I'm using steam in a container. I wish they could just do it every three months -- at least for Linux users. I think that's a reasonable time between distro-hopping and stuff, so that the Linux distro division remains fresh.
run this command
sed -i -e '/SurveyDate/ s/"[0-9].*"/"'$(date +%Y-%m-%d -d "1 year ago")'"/' ~/.steam/steam/config/config.vdf
Then reopen steam
You will get a steam survey.
I would prefer the login and playtime data as well. Hit me up Valve; I can make this happen for you.
Last edited by fenglengshun on 3 May 2023 at 5:43 am UTC
Quoting: PhiladelphusLike, you can flip a fair coin and get 10 heads in a row. It's quite a small probability to happen, but it's technically just as random an outcome as getting 5 heads and 5 tails
Nnnno.
Getting HHHHHHHHHH is as likely as getting HHTHTTTHTH, but there's many more versions of 5 heads 5 tails.
Quoting: EikeHe didn't say just as likely, only just as random.Quoting: PhiladelphusLike, you can flip a fair coin and get 10 heads in a row. It's quite a small probability to happen, but it's technically just as random an outcome as getting 5 heads and 5 tails
Nnnno.
Getting HHHHHHHHHH is as likely as getting HHTHTTTHTH, but there's many more versions of 5 heads 5 tails.
Quoting: Purple Library GuyThis. Getting 5 heads and 5 tails is the most likely result, probabilistically. But all results are equally random. But getting 10 heads doesn't feel random, even when it actually is.Quoting: EikeHe didn't say just as likely, only just as random.Quoting: PhiladelphusLike, you can flip a fair coin and get 10 heads in a row. It's quite a small probability to happen, but it's technically just as random an outcome as getting 5 heads and 5 tails
Nnnno.
Getting HHHHHHHHHH is as likely as getting HHTHTTTHTH, but there's many more versions of 5 heads 5 tails.
And if you have millions of people flipping coins, statistically a non-negligible number of them will get 10 heads…then go online to point out how non-random their results are.
See more from me