The Steam Machine is destined for the console colosseum.

Before you start making plans to evict your current rig, I want to explain how the Steam Hardware Survey actually works. Rather than representing every player that uses the storefront, the monthly census instead uses a small percentage of participants who’ve opted in as a sample. It’s anything but representative of all 154 million active users, meaning the statistic above isn’t an absolute.

Does the author know how *survey sample sizes work?

Do I?

Is it because it’s opt in or opt out?

  • njm1314@lemmy.world
    71·
    1 day ago

    Technically all surveys are opt-in. Gallup can’t force you to answer questions when they call you on the phone. You could just hang up. It’s rather silly to argue that opting in makes it an invalid survey.

    • SirActionSack@aussie.zoneEnglish
      1·
      9 hours ago

      This is a silly take. People are less likely to say no on the phone than they are to click no on a survey opt-in popup. And I am not even sure that Steam does a pop-up, users may have to go find the option to join the survey.

      So there absolutely is a sampling bias. How significant it is and how/if Valve attempt to control for it doesn’t seem to be public information but it is definitely there.

      • njm1314@lemmy.world
        1·
        9 hours ago

        All right first of all no, you don’t get to just claim that people are less likely to do something based on nothing. I’m not going to let that stand in this conversation. You got to prove that. If anything I’m much less likely to talk on the phone than I am to click on a pop up. Avoiding talking to people is one of my passions. So not only have you not proven it but it goes directly against my own inclinations and nature. I deny that 100%.

        Also since you came to this conversation with no actual information apparently let me share something with you. It is a pop up. You probably should have checked on that before you based your argument on it. You don’t have to go looking for it. Just pops up on your screen.

        However I do agree with one point you raised here. There absolutely is a sampling bias. Just as I said in my previous post. And I appreciate you backing me up on that point. The bias is toward people who choose to answer the survey. As is true for all surveys.