Why one person's "same color" is another person's "obviously different"

Show two nearly identical greens to ten people and you'll get a small argument. A few see the difference instantly, most squint, and someone insists you're pranking them. That argument is what this test measures — and the reasons behind it are genuinely interesting.

Your eye has three color sensors, and they're not identical across people

Human color vision starts with three types of cone cells in the retina, each tuned to a different band of light: long (reddish), medium (greenish), and short (bluish) wavelengths. Every color you've ever seen is your brain comparing the signals from these three sensors. The genes for the long- and medium-wavelength cones vary between individuals — sometimes the two pigments sit unusually close together, sometimes one is missing entirely. That genetic variation, plus differences in the eye's optics, age (the lens yellows over the years), and plain old attention, is why color discrimination genuinely differs from person to person.

What "discrimination" means, precisely

Vision scientists don't ask "can you see green?" — they ask "how small a difference can you reliably detect?" That threshold is called a just-noticeable difference. Since the 1940s, researchers have measured it with hue-arrangement tasks, where subjects order dozens of subtly different color chips; the pattern of mistakes reveals both overall discrimination ability and specific weak zones. Color-difference itself is quantified with formulas the lighting industry standardized (the delta-E family): a delta-E around 1 is roughly the edge of human perception, while 5 or more is plain to most people.

How this test is built

Color Nerve borrows the spirit of those laboratory tasks and compresses it into a game. Each round shows you a grid where one tile differs from the rest. The difference shrinks by about 13% every round, with no floor below the edge of perception, so the game must eventually beat everyone — that's what makes the final round number meaningful. The changing dimension rotates every round between lightness, hue, and saturation, because those are genuinely different visual skills; the lightness rounds also mean players with color-vision deficiency can progress on brightness alone rather than being locked out at round one. Three lives absorb misclicks, and the fifteen-second timer keeps you honest.

Your percentile is currently an estimate based on expected performance, clearly labeled as such. As more people play, we recalibrate it against real, anonymous round-reached statistics — the honest way to run a percentile.

The screen problem (and why this isn't a medical test)

Here's the limit we can't engineer away: we don't control your display. An OLED phone at full brightness, a dim office monitor, and a laptop with night-shift enabled render "the same" colors differently. Clinical color-vision testing uses printed plates under controlled lighting for exactly this reason. So treat your score as a fun, roughly-right measure of your eye on your screen today. If you consistently struggle with colors in daily life — traffic lights, ripeness of fruit, colleagues' color-coded charts — that's worth a proper exam with an eye doctor, not a browser game.