Announcement

Collapse
No announcement yet.

Valve's Steam Survey Numbers For October Just Point To More Inaccuracies

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Valve's Steam Survey Numbers For October Just Point To More Inaccuracies

    Phoronix: Valve's Steam Survey Numbers For October Just Point To More Inaccuracies

    With the start of a new month we are always eager to see what Valve reports via their "Steam Survey" for Linux usage (among other stats) for the month prior. The October 2019 numbers are now published but they don't indicate any Linux change in marketshare but with some odd indications...

    http://www.phoronix.com/scan.php?pag...r-October-2019

  • #2
    The numbers are like web surveys: Not supposed to be taken serious. Especially not below 1% which the given accuracy.

    Comment


    • #3
      I only play CSGO on Linux. I would play more games, but most do not allow Linux/Windows players to play against each other. 0 AD is a fun free RTS game for Linux.

      Comment


      • #4
        Phoronix should stop reporting these numbers if they are so clearly meaningless...

        Comment


        • #5
          Maybe one could argue that Linux has certainly established a strong bridgehead here. If so, I could call it a success so far. And considering all things, it has been success. There are more and more grounds for users to settle in better than ever before.

          Comment


          • #6
            Originally posted by lucasbekker View Post
            Phoronix should stop reporting these numbers if they are so clearly meaningless...
            Pointing out how obviously wrong they are is a first step to that. If you were to just stop without addressing the issue, it would not make much of a difference.

            Comment


            • #7
              People avoid surveys like fire. I never understood the point of them, really.

              Could someone explain it, please? Thanks in advance.

              Comment


              • #8
                Originally posted by lucasbekker View Post
                Phoronix should stop reporting these numbers if they are so clearly meaningless...
                lucasbekker should stop writing these comments if they are so clearly meaningless.

                Comment


                • #9
                  Valve's survey numbers usually get revised later in the month, IIRC.

                  Here's a dedicated page, at gamingonlinux.com: https://www.gamingonlinux.com/index....am_linux_share Thanks, liamdawe

                  Edit: numbers from before Nov. 2017 have been removed from that page due to a few bugs with how Valve reported marketshare... That made it over-report computers in cybercaf├ęs. This was highly correlated with the rise of PUBG in China, with a big increase in the playerbase & (stretching my knowledge a bit) steam-backed cybercaf├ęs there, which created a huge spike in chinese-speaking numbers, as reported in the steam survey, while this caused a big dip in Linux number. Hence why the extra graphs on that page, to explain that.

                  There's not much point in them now that this bug has been fixed. I wish the longer history wass still available on the site to provide soe context, though.
                  Last edited by [email protected]; 11-02-2019, 06:27 AM.

                  Comment


                  • #10
                    Originally posted by xinthose View Post
                    I only play CSGO on Linux. I would play more games, but most do not allow Linux/Windows players to play against each other. 0 AD is a fun free RTS game for Linux.
                    I tried playing CSGO. It was laggy as hell and I immediately uninstalled it. PIng seemed to be OK (~40ms), FPS was also OK (>200), but it felt like there was a 0.3 seconds delay between my actions and the reaction of the game. It felt like I was in water. I couldn't win a single duel. No, thank you.

                    Comment

                    Working...
                    X