THE SHORT VERSION
Hitori: a grid pre-filled with digits. Shade some cells so that no digit appears twice in any row or column, shaded cells don't touch each other, and all unshaded cells stay connected. Three rules — and they interact.
Takuzu (also called Binairo): an empty grid. Fill every cell with a 0 or 1. Each row and column must have equal counts of 0s and 1s, and no three consecutive cells in a row or column can hold the same digit. Two rules — simpler, but still deductive.
Neither puzzle uses arithmetic. Both are won by following the constraints to their logical conclusion.
SIDE BY SIDE
Starting state
Hitori: grid filled with multi-digit values — you shade
Takuzu: empty grid — you fill with 0s and 1s
Number of rules
Hitori: three (no-repeat, no-touch, connectivity)
Takuzu: two (balance, no-three-in-a-row)
Rule interaction
Hitori: rules cascade — each shade triggers the next deduction
Takuzu: rules are mostly independent constraints
Arithmetic required
Hitori: none
Takuzu: none
Grid size
Hitori: typically 5×5 to 8×8
Takuzu: even-sized grids (6×6, 8×8, 10×10)
Given information
Hitori: all digits pre-filled — you choose which to shade
Takuzu: some 0s/1s pre-placed as seeds — you fill the rest
THE RULE INTERACTION DIFFERENCE
Takuzu's two rules operate somewhat independently. The balance rule (equal 0s and 1s per row/column) constrains totals; the no-three-in-a-row rule constrains local runs. Solving Takuzu means alternating between these two lenses — finding where one rule forces a cell, then checking if the other confirms it.
Hitori's three rules create a cascade. Shade one cell (to fix a no-repeat violation), and its four neighbours are immediately confirmed unshaded (no-touch rule). Each confirmed-unshaded cell can then force another duplicate to resolve elsewhere (no-repeat again). The connectivity rule adds a third lens: a shade that would cut the unshaded region in two is always wrong. One correct shade can unlock a chain of five or six forced moves.
This cascade is what makes Hitori feel elegant to experienced solvers and opaque to beginners. Takuzu's logic is more local — each deduction is usually visible within a single row or column.
WHICH IS HARDER?
Takuzu is the faster entry puzzle. The rules are fewer and more transparent; a beginner can solve a 6×6 grid by scanning rows for near-complete runs without needing to think about the whole grid at once.
Hitori is harder at equivalent difficulty tiers. The three-rule cascade requires holding more state in your head simultaneously, and the connectivity check demands a global view of the grid that Takuzu's local-scan approach never requires. On a hard Hitori, you may need to reason about how a shade on one side of the grid would affect connectivity on the other side.
The payoff is proportional: a Hitori chain reaction, once you see it, feels noticeably more satisfying than Takuzu's steady row-by-row fill.
WHAT STAYS THE SAME
Both puzzles are solved purely by elimination — no arithmetic, no guessing. You follow the constraints until only one valid state remains. If you find yourself guessing in either puzzle, you've missed a deduction.
Both reward finding the most-constrained cell first. In Takuzu, that means rows with only one or two empty cells. In Hitori, that means digits that appear three times in a row (the middle instance is almost always forced).
Experience in either puzzle sharpens general grid-scanning instincts that transfer to the other — and to Shikaku, which shares the same "no arithmetic, all constraint" flavour.
WHEN TO MAKE THE SWITCH
If you play Takuzu: try Hitori when you want a puzzle where the rules create chain reactions rather than independent checks. Hitori's cascade logic is a genuine step up in complexity, and the connectivity rule adds a spatial dimension Takuzu doesn't have.
If you play Hitori: try Takuzu when you want a faster-paced constraint puzzle that you can pick up and put down mid- row. Takuzu's binary simplicity makes it easier to pause and resume without losing track of your reasoning.