i feel like this needs like 5x as many data points before giving a result, also at some point to me the only correct answer would be "neither" because the middle point is just Cyan to me, which isn't blue and definitely isn't green, just like how orange isn't yellow but definitely isn't red.
That’s cause if you know name of colors with associated color, you more easily distinguish them, there was nice article in scientific american around that stuff. Anglos don’t typically separate for example light blue/cyan as a big color, and thus spend more time choosing lighter blue hue among others. This test is largely for funsies.
And 8 bit colors don’t have lots of space to do more datapoints tbh.
i more mean that it needs to show you the same colour several times, before and after different colors each time, because right now my scores are jumping wildly between runs.
ideally it should also show red between each hue, to reset your eyes and brain a bit.
Probably more effort than worth it, due to displays, and display tech being particularly shit around true green/cyan colors (if you checkout rec2020, you can see green/cyan zone is precisely the point the displays don’t reach and don’t hope to reach)
I think that the difficulty in deciding which to pick when the truth feels to be "neither" is an intended effect of the test. It would be interesting to see the results of a test that allowed for "neither" though
i don't think that's the case, the point of the website seems to quite clearly be seeing what shades people consider either colour, and if a shade is "neither" then that's the answer you want to track rather than polluting your data with forced false answers.