Tap two tiles to swap them. Sort the shades into a smooth gradient. The locked ends can't move.

What Huesort trains

Huesort works your gradient ordering: taking a scramble of closely related shades and ranking them into one even climb. The row runs from a dark end to a light end, and because the two endpoints also sit at different points on the color wheel, every step along the row shifts hue as well as lightness. Ordering it means comparing each tile against both neighbours at once and deciding which of two very similar shades comes first, a fine-ranking judgement rather than a spot-the-difference one.

How to play

Tap one tile to select it, then tap a second to swap the two. Keep swapping until the row climbs smoothly from the dark end to the light end. The tiles at each end carry a padlock and cannot be moved: they are your anchors, and everything between them has to bridge them cleanly. A counter shows your swaps and the par for the puzzle sits next to it. There is no timer and no limit on swaps, and the row ends the moment it is correct, so you never have to declare that you are finished. Selection and every swap are announced to screen readers.

How scoring works

Only one thing is scored: how many swaps you needed. Par is the number of movable tiles, so a clean run means solving the row without wasted moves. Finish at or under par and the result reads "Perfect"; finish above it and it reads how many swaps you took. Time does not count and neither does how long you deliberated, so thinking before you tap is free and tapping to see what happens is expensive. The shared result shows three squares: all three filled at or under par, two up to two swaps over, one beyond that.

What your score means

Read the result as a rough band rather than a precise threshold:

Outcome	Rough interpretation
At or under par	You ranked the shades before moving anything
One or two swaps over par	Solid; one pair was genuinely hard to separate
Well over par	Mostly trial and error; slow down and rank first

One day is noisy. A single ambiguous pair can cost two swaps, so the trend over many days is the honest signal.

Common mistakes

The most expensive habit is swapping to see what happens. Every exploratory move costs the same as a correct one, so decide the whole order first and then execute it. Second, players compare only adjacent tiles; when two shades are close, hold one against a tile two or three places away, where the difference is large enough to settle the direction. Third, the hue shift misleads: the row is a climb in lightness with a hue drift riding along, so sorting by "which looks bluer" fights the ordering. Squinting lets lightness dominate. Fourth, a tile's neighbours change how it looks, which is why a shade can seem right in one slot and wrong after a swap elsewhere; see simultaneous contrast.

Practice progression

(5 min, beginner) Play the daily row, but before touching anything, point at the tiles in the order you think they belong. (10 min, intermediate) Use the practice tab, keep the tile count low, and try to hit par every time rather than solving quickly. (weekly) Raise the tile count and move to the hard tier, where the lightness span is compressed and the steps are much closer together, then read value and saturation in composition so the ranking has a vocabulary behind it.

The science behind it

Arranging colored samples into an ordered series is one of the oldest ways of measuring color discrimination: the classic hundred-hue arrangement test asks you to put a tray of caps into a smooth sequence, and the errors reveal where in color space your discrimination is weakest. Huesort is a small, informal cousin of that idea. It also leans on relative rather than absolute judgement: nobody can name the lightness of a tile, but almost everybody can say which of two is lighter, and a sort is a long chain of those comparisons. More in hue, saturation and lightness.

Limitations

Huesort is a game and an informal self-check, not a clinical or diagnostic tool. Screen calibration, brightness, viewing angle and the theme you play in all change how separable the steps look, so results from different devices are not comparable. Swap count also rewards planning as much as perception. If you suspect a color vision difference, a proper test is the right route; see color vision differences.

Where it shows up

Ordering shades into an even run is the invisible half of a lot of visual work. A data designer building a sequential color scale has to make each step feel the same size as the one before, or a chart will invent structure that is not in the data, such as a band that looks like a boundary because two neighbouring steps happen to be further apart. A decorator laying paint samples on a table sorts them light to dark before deciding which three belong together. A photographer editing a set for a gallery arranges the frames so the tonal weight moves smoothly instead of jumping, which is a sort by lightness before it is a sort by subject. A printer checks a step wedge, a strip that should march evenly from paper white to solid, and reads any uneven jump as a press problem. In all of these the task is the same, and so is the failure mode: a single item in the wrong place makes the whole run look uneven, even when nobody can say which item is at fault. Huesort trains the noticing and the ranking, not the taste.

Ready to sort today's gradient? Seven shades, two locked ends, no timer, and a new row tomorrow.

Play today's Huesort ↑

Frequently asked questions

Why can't I move the end tiles?

They are locked on purpose. They give you a fixed dark end and a fixed light end to sort against, which turns an open-ended arrangement into a solvable ranking.

What is par?

The number of movable tiles in the row. Finishing at or under par counts as a perfect solve; going over just means you spent moves finding the order.

Is the row sorted by hue or by lightness?

Both move together. The gradient runs from a dark endpoint to a light one while the hue drifts between two points on the wheel, so each step differs in lightness and in hue at the same time.

Is there a time limit or a move limit?

Neither. You can take as long as you like and make as many swaps as you need; only the swap count is recorded.

Can I practise without waiting for tomorrow?

Yes. The practice tab lets you set difficulty and the number of tiles, and adds a Check button that tells you how many tiles are already in place. Practice rounds do not affect your daily streak.

Huesort

About Huesort — order shades into a smooth gradient