Forcing Nets vs Forcing Chains in Sudoku Explained

Shape: 1D Chain vs 2D Net

The most fundamental difference lies in structure. A Forcing Chain is strictly linear. You choose a starting candidate, assume it is true (ON) or false (OFF), and follow a single, unbroken sequence of logical consequences. If this path leads to a contradiction, the starting assumption is false.

A Forcing Net is two-dimensional. It starts the same way but can branch. Different branches of the net can impact the same cell from different directions. This allows for a more complex result: two candidates in a cell might be eliminated via separate branches, proving the third candidate must be ON, which then kicks off a new chain fragment within the net. This branching and reconverging is what makes nets so powerful.

How They Start: Following Every Consequence

Both strategies begin identically. You select a candidate digit in a cell. You then set that candidate to a state—either ON (placed as the solution) or OFF (removed as a possibility). The core mechanic is to follow *every* direct consequence of that assumption. This means flipping the state of all other candidates that 'see' the target cell (are in the same row, column, or box) based on Sudoku's basic rules. A linear chain follows one consequence at a time. A net follows all consequences simultaneously, allowing the logic to branch.

The Four Contradictions That End a Net

The goal of exploring a net is to find an impossible situation, or contradiction, which disproves the initial assumption. There are four recognized contradiction types: 1) All candidates in a cell are turned OFF (leaving the cell empty). 2) Two candidates in the same cell are turned ON. 3) All candidates for a digit within a unit (row, column, or box) are turned OFF. 4) Two candidates for the same digit within a unit are turned ON.

If the exploration does not find a contradiction but instead results in a perfect, consistent cover of the entire board, this is known as Bowman's Bingo. This is treated as a separate, absolute last-resort strategy, as it essentially proves the starting assumption was correct, solving the entire puzzle.

Three Complexity Levels of Forcing Nets

Forcing Nets are categorized by complexity, which dictates the types of logical links they use. Level 1 nets are purely linear and use only bi-value and bi-location links. This level closely mirrors the logic of 3D Medusa coloring. Level 2 nets introduce the use of grouped cells, where a link can represent a group of candidates. Level 3 nets are the most complex, using grouped cells plus internal hidden sub-chains to establish links.

Advanced solvers search for these nets systematically. They typically test OFF candidates first, then ON candidates, seeking the net with the shortest combined chain length rather than just the first one found. In testing, this strategy solved about 97% of SudokuWiki's weekly 'unsolvable' puzzles and 100% of Ruud's top 50,000 benchmark puzzles, showcasing its extreme power for puzzles that require guessing.

When Should a Human Player Use This?

Forcing Nets, and even complex Forcing Chains, are last-resort strategies. You should only consider them after exhausting all other expert sudoku techniques like Alternating Inference Chains (AIC), Unique Rectangles, and fish patterns. They are computationally intensive and not intended for casual solving.

For human solvers, manually tracing a large net is extremely difficult and error-prone. The value in understanding them is to recognize when a puzzle has entered a territory that requires such deep, almost brute-force logic. If you are stuck and all visible chains have stalled, you know you're facing a puzzle that necessitates this advanced, branching approach. It's the final tool for the most extreme challenges.

• Practice identifying bi-value and bi-location links, as these are the building blocks of Level 1 nets.
• Use software with a hint feature that implements forcing nets to see them in action without manual tracing.