I think there's an anchoring effect in play here. If you select blue -> blue -> green -> blue -> green -> blue -> green…, you land at the population median.
(The point being that, once you get to a somewhat ambiguous point (after two blue selections), you can say "oh, well, compared to the last one this is {opposite color}!", and it seems most people do that.)
I wouldn't assume most people do that. For me the last few looked basically the same so I selected the same color for all of them.
You can't assume most people do that, but you also can't assume most people do not do that.
My boundary was hue 188, bluer than 98% of the population, for me turquoise is green and then it shows an overall chart which I have to agree with so no, I don't agree with your assessment. I often get into blue/green arguments with my children and that's when I started to suspect that it was personal opinion.
That's if you're matching about 40% of the population.
For some, it might be blue -> blue -> blue -> blue -> green -> blue -> green -> blue.
That doesn’t explain why I landed 92% off the population median.
it's a binary search, not too surprising. search over a unidimensional ordered space
this made no sense btw, this would give 4/3