• PieMePlenty@lemmy.world
      link
      fedilink
      arrow-up
      3
      ·
      8 hours ago

      Imo, this test is flawed and doesn’t take display refresh rate into account. Well, at least flawed in the sense that you can’t compare it to other people because they may have measured it differently on different hardware. Its not universal.
      I’m at work right now and using this shitty screen at 60hz, I got 230ms. I upped the refresh rate as fast as it can go, to 75hz and improved my reaction time to 200ms.
      To get accurate results, you’d need to do this test at different stages of your life on the same hardware with the same software version of the test. So take it with some salt.

      • Holytimes@sh.itjust.works
        link
        fedilink
        arrow-up
        1
        ·
        2 hours ago

        Honestly, all you actually would need to do is just go to the steam survey, check what the most common hardware is.

        Cuz all you got to do is test against that. As if your response time for example is in the top 10% of people. When your baseline is the most common hardware available then you’re in at worst the top 10% of people. Reaction speed changes depending on task just as much as it does on hardware.

        Like personally, I take a reaction test using the most common steam hardware and I’m looking at about 150 milliseconds response time by doing it on my normal hardware which is many times faster. In refresh rate it only shaves like two milliseconds off average and it fluctuates just as much. Like two milliseconds off average and it fluctuates just as much.

        I even have a little like toy thing that I picked up one day years ago that tests reaction time using a mechanical device. Testing it. It’s within about 5 to 10 milliseconds of variation off of what most digital online tests that I can find.

        And it within margin of error shows exactly the same amount of reaction speed that the online tests show.

        So they’re accurate enough at least within a handful of milliseconds.

        The toys basically just a taser and two people grip a thing. Whoever grips and clicks the thing faster doesn’t get tased its rather fun.

        I almost always win even against my younger cousins and step Brothers who are as young as 12. So my 35 year old ass still has it!

        Just don’t ask me to have any amount of reaction speed with my legs. Probably try and kick something. I will miss it. By a mile. 3 hours after you threw it.

    • daannii@lemmy.world
      link
      fedilink
      English
      arrow-up
      1
      ·
      edit-2
      9 hours ago

      So I said there is a difference of about 200ms. But humans cannot typically react faster than 200-300 ms. Even young people. Because it takes ~ 200ms for a signal to be sent. So I’m suspicious of this result.

      I used to do research on reaction time. We throw out any number under 150 because it’s considered not humanly possible and it’s an error measurement.

      I rarely saw 200.

      Usually people hovered around 300 to 400.

      Also I was doing research on 18-20 year olds

      • Holytimes@sh.itjust.works
        link
        fedilink
        arrow-up
        1
        ·
        edit-2
        1 hour ago

        Keep mine. It really does depend on tests. A lot of the online tests start with a negative offset to account for ping. So a 160 you would typically add anywhere from 30 to 60 on average. The better online tests will actually run a short ping test in jitter test to set the offset more accurately. Most though don’t and just use a fixed value. The annoying part is it’s usually in the fine print somewhere that you have to manually f****** re-add it. They don’t tell you that it has a offset. So that way the numbers look lower. It’s stupid but it’s the trend.

        My rule of thumb is most of the time the devs just slap a 50 in there surround number and it kind of covers most bases. So a 160 would be a 210. Which for a test where you can do it repeatedly and be hyper fixated on knowing what you’re doing. You can get pretty damn close to 200, not reliably but that’s going to be like you’re 1% best result which is what people f****** posting.

        Like for example, my best results in these sort of tests accounting for offsetter between 190 and 210. But that’s like one in 50 tries. Where my normals closer to like like 230.

        And usually I have like a 10 millisecond variance on my average tries.