How to Solve KenKen Without Guessing: Cage Combinations, Candidates, and Elimination

KenKen Is a Logic Puzzle, Not a Guessing Game

You can solve KenKen without guessing by turning every cage into a list of valid number combinations, writing candidates in each cell, and eliminating options that conflict with the row, column, or cage rules. Each deduction reduces uncertainty until only one legal value—or one legal arrangement—remains.

That is the real rhythm of KenKen: combine, compare, eliminate, repeat. The arithmetic provides possibilities, while the grid tells you where those possibilities can go.

Start With the Three Rules That Control Everything

Before exploring strategies, make sure the basic rules are clear:

1. An N × N grid uses the numbers 1 through N. A 6 × 6 puzzle, for example, uses 1 through 6.
2. Every number must appear exactly once in each row and column.
3. The numbers inside each outlined cage must produce its target using the shown operation.

A single-cell cage contains its displayed number automatically. In subtraction and division cages, the numbers may generally appear in either order: a two-cell 2− cage could contain 4 and 2, while a 3÷ cage could contain 6 and 2. The official KenKen introduction provides a helpful visual explanation of these rules.

One detail deserves special attention: numbers may repeat inside a cage if the repeated cells are in different rows and different columns. A cage does not behave like a Sudoku box. The row and column restrictions—not the cage border itself—control repetition.

Turn Every Cage Into Combinations

A cage clue does not immediately tell you which number belongs in each cell. It tells you which groups of numbers are possible.

Consider a two-cell 7+ cage in a 6 × 6 puzzle. Its possible combinations are:

• 1 and 6
• 2 and 5
• 3 and 4

At first, any of those pairs may work. However, if one cell shares a row with an existing 6, the 1-and-6 combination can no longer place 6 in that cell. If the other cell also sees a 6 in its column, the entire 1-and-6 combination is eliminated.

This distinction is important:

• A combination is the group of digits that satisfies the cage.
• An arrangement is the order in which those digits occupy the cage’s cells.

For a two-cell 12× cage in a 6 × 6 grid, the combinations are 2 and 6 or 3 and 4. Each combination normally has two possible arrangements until row and column information determines the order.

Learn the Personality of Each Operation

Different operations create different kinds of clues.

Addition cages often have several combinations, especially when they contain many cells. Look closely at very low or very high totals because they tend to be more restricted.

Multiplication cages reward factor recognition. For example, a two-cell 18× cage in a 6 × 6 puzzle must contain 3 and 6.

Subtraction cages are usually small and constrained. A two-cell 5− cage in a 6 × 6 grid can only contain 1 and 6.

Division cages also produce short lists. A two-cell 2÷ cage in a 6 × 6 puzzle can contain 1 and 2, 2 and 4, or 3 and 6.

The official collection of KenKen solving tips and useful cage patterns lists many common combinations for different grid sizes.

Write Candidates, Then Remove Them Systematically

Candidates are the numbers that could still legally occupy a cell. They make invisible logic visible.

Suppose a cell belongs to a 7+ cage in a 6 × 6 puzzle. From the cage alone, its candidates might be {1, 2, 3, 4, 5, 6} because it could be one half of several different pairs. If its row already contains 2 and 6, and its column contains 3, its candidates shrink to {1, 4, 5}.

Now check those candidates against the cage’s possible combinations. If its partner cannot be 6, then this cell cannot be 1. If its partner cannot be 2, this cell cannot be 5. It may therefore be forced to 4.

For larger or more difficult puzzles, recording candidates prevents you from repeatedly recalculating the same possibilities. The Puzzles Arcade guide to using a puzzle solver’s scratchpad offers useful ways to keep possibilities and confirmed facts organized.

Use Rows and Columns as Powerful Filters

Cage arithmetic creates possibilities, but rows and columns destroy the impossible ones.

Whenever you place a number, remove it as a candidate from every other cell in its row and column. Then revisit the cages affected by those removals. A cage with three combinations may suddenly have only one.

Look for Missing Numbers

If a row in a 5 × 5 grid contains 1, 2, 4, and 5, its empty cell must be 3. This simple “missing number” check is easy to overlook when attention is fixed on cage arithmetic.

Scan rows and columns regularly rather than waiting until the end.

Find Forced Positions

Sometimes a number has several candidates in a row, but only one of those cells can actually accept it.

Imagine that a 6 must appear somewhere in a row. Four empty cells remain, but three cannot contain 6 because of their columns or cage combinations. The fourth cell must be 6—even if it still has other candidates written inside it.

Use Pairs and Sets

If two cells in the same row can only be {2, 5}, those two cells must contain 2 and 5 in some order. No other cell in that row may contain either number.

The same principle extends to three cells containing only the same three candidates. These locked sets are a standard elimination tool in number-placement puzzles and are especially useful when KenKen cages line up within one row or column.

For a broader explanation of this style of reasoning, see Puzzles Arcade’s guide to the process of elimination in logic puzzles.

Pay Attention to Cage Shape

The same target can behave differently depending on the cage’s shape.

In a 4 × 4 puzzle, a three-cell 6+ cage lying entirely within one row must use 1, 2, and 3. Repeated digits are impossible because no number may repeat within that row.

An L-shaped cage may allow combinations containing a repeated digit, provided the matching digits occupy cells that do not share a row or column. Therefore, never list combinations using arithmetic alone. Apply the cage’s geometry at the same time.

Cage shape can also lock a number into part of a row or column. If every valid arrangement places a 4 in one of two cage cells that share the same row, then 4 can be removed from all other cells in that row.

Use Row Totals for Tough Addition Cages

Every row and column in an N × N KenKen has a fixed sum:

[ 1+2+\dots+N=\frac{N(N+1)}{2} ]

In a 6 × 6 grid, every row and column totals 21. In a 5 × 5 grid, each totals 15.

Suppose the known or logically grouped cells in a 6 × 6 row total 16. The remaining cells must total 5. If those cells have candidates {1, 2, 3, 4}, you can immediately narrow the possibilities to 1 and 4 or 2 and 3.

Be careful when cages cross row boundaries: use only the values or contributions belonging to the row you are analyzing, not the full target of a cage that extends elsewhere.

Follow a Reliable No-Guessing Routine

When facing any KenKen, work through this checklist:

1. Fill every single-cell cage.
2. Scan for rows or columns missing only one number.
3. List combinations for the most restricted cages.
4. Convert combinations into cell candidates.
5. Remove candidates already used in each row and column.
6. Check for forced positions, pairs, and locked sets.
7. Revisit every cage affected by a new placement.
8. Use row or column totals when ordinary elimination stalls.
9. Confirm that every completed cage reaches its target.

If you become stuck, do not stare harder at the same cage. Move to another part of the grid and search for the smallest candidate list. A deduction elsewhere may remove exactly the option that was blocking you.

Let Certainty Build the Solution

Solving KenKen without guessing does not require instant insight or extraordinary calculation. It requires careful bookkeeping and a willingness to make small, certain deductions.

Treat every cage as a combination problem, every cell as a candidate list, and every row and column as an elimination system. Progress may begin with a single crossed-out number, but that elimination can force a pair, solve a cage, complete a row, and unlock the entire grid.

The goal is not to predict the answer. It is to make every wrong answer impossible until the correct one is all that remains.