Colordle Solver

Colordle scores each guess with a similarity percentage against a hidden named color. A higher number means closer in hue, brightness, and saturation. A score of 100 means the name matches exactly. Everything below that is a distance reading wearing a friendly face.

The solver runs that scoring in reverse. It takes your guess, computes the same color difference the game uses, Delta E CIE2000 in LAB space, against every candidate in the named-color list, and deletes anything that would not produce your exact score. Not demoted. Deleted. The arithmetic either lands inside the game's window or the color is gone.

That strictness is the feature. Eyes drift with screens, lighting, and fatigue. The formula does not drift at all. If the solver says a color cannot score 34 against your guess, it cannot, whatever your monitor suggests.

A single guess plus its percentage defines a shell. Every color sitting at that exact perceptual distance survives, and everything closer or farther drops out. One shell still leaves plenty of candidates, which surprises players who expect a single number to nearly solve the puzzle.

The second guess is where the puzzle breaks open. A new shell centered on a different color intersects the first, and the survivors are the overlap. Two well-placed guesses routinely compress thousands of named colors into a list short enough to read. A third usually ends it.

Each percentage is a distance measurement to an unknown point, and two or three measurements from different positions pin the point down. Surveyors work this way. Colordle players can too.

Open in the game with a central named color, something balanced rather than a neon edge shade. Note the percentage it returns. Type the name into the solver, enter the score exactly as shown, and filter. The list that comes back is the menu for guess two.

Pick the surviving candidate farthest from the first guess. Distance between guesses is what makes the shells intersect usefully instead of overlapping lazily. Play it, bring the new percentage back, and watch the list collapse to a handful.

Guess three confirms. Two survivors means one of them is the answer and a single guess settles it. A list that stays long has one usual cause: the first two guesses sat too close together. The fix is a third guess genuinely far from both, not more nudging.

Edge shades filter unevenly. A neon or near-black guess eliminates plenty for some answers and almost nothing for others. Central balanced colors split the space evenly, so every possible answer teaches something. Boring guesses are better instruments.

Naming matters as much as position. The solver speaks in exact names, so a guess you can name precisely beats one you can only describe. Salmon is a tool. That pinkish-orange shade is a shrug. The game draws from a named palette, so the vocabulary transfers in both directions.

Primaries make fine early probes for the same reason. Red, blue, green, yellow, purple, and orange anchor the wheel, and their scores divide the space into readable regions. After two or three anchors, the hue family is usually settled and the real narrowing starts.

Scores in the high nineties feel like victory and behave like quicksand. Dozens of named colors can sit nearly the same distance from a guess, so the candidate list stays stubbornly long right when patience runs thinnest. The escape is a guess from a different part of the palette entirely, even one known to be wrong. A distant shell intersecting a near-miss shell leaves almost nothing.

Rounding is the quieter trap. The game shows decimals and the filter honors them, so enter every digit exactly as displayed. A mistyped decimal quietly deletes the answer while keeping hundreds of impostors. When a list looks wrong, the entered score is the first thing to recheck.

Small screens compound both traps. Tired eyes at night cannot separate neighbors the formula separates easily. Trust the list over the glance, and let the arithmetic finish what perception started.

Consistency is the first win. The solver applies identical arithmetic to every row under every condition, while human color judgment wobbles with screens and hours. For steadiness alone it earns a place beside the game.

Memory is the second. Five percentages in play means five constraints held simultaneously, and players reliably track about three. The solver holds all five and never suggests a color an earlier row already killed, which is precisely the error tired players make near the end.

Use it to learn, not just to win. Each filtered list teaches how the named palette is organized: which families cluster, where the gaps sit, how far apart near-neighbors really are. That education compounds into unaided solves faster than any amount of staring.