When to Switch to Pencil Marks
Switch too early and you waste minutes marking. Switch too late and you scan the same wall forever. The right moment, by tier.
Pencil marks are powerful and slow. They unlock every technique past hidden singles, but they take real time to lay down and more time to maintain. Knowing when to switch them on is one of the small skills that separates competent solvers from fast ones.
Switch too early and you have spent four minutes marking cells that would have placed themselves. Switch too late and you are scanning a wall of empties hoping for a placement that is not coming.
What a "clean scan" means
A clean scan is a deliberate walk through every digit, by unit, looking for hidden singles. You start at 1, walk the boxes, rows and columns, place any forced cells, then move to 2, and so on. The whole pass takes one to two minutes on a typical medium puzzle.
If you complete the pass and your hand has not moved, you have confirmed that no naked or hidden singles remain. That is the switch moment.
The wrong reasons to switch early
Beginners switch to marks when they get stuck. "Stuck" usually means "I have not done a clean scan and the placements feel invisible." The right response is to redo the scan, not to mark. Most stuck moments dissolve into placements within thirty seconds of a careful walk.
Some solvers switch to marks because marking feels productive even when no progress is being made. Filling cells with little digits is its own activity, and it is easy to confuse with solving. Resist. If you have not exhausted scanning, marks are the wrong tool.
- Switch too early
- You will spend ten minutes filling marks for cells that scan would have placed in two.
- Switch on time
- Marks fuel pairs, triples and pointing patterns within minutes of being laid down.
- Switch too late
- You will scan and re-scan the same wall, getting more frustrated, for ten or fifteen minutes.
The reverse decision: when to stop marking
Once you have placed five or six digits after the initial mark sweep, candidates collapse fast. Some solvers keep updating marks until the very end. Others stop maintaining marks after the cascade picks up and revert to scanning. Both can work.
The argument for keeping marks alive: they remain accurate, you catch every pair, and you finish with confidence. The argument for dropping them: scanning is faster once the grid is mostly full, and stale marks become more likely as the solve accelerates. My preference is to keep marks until I am within ten placements of the finish, then revert to scanning.
The difficulty effect
Easy puzzles: never mark. The scan finishes them.
Medium puzzles: mark around the ten-placement mark, where the cascade slows. Most mediums finish in another fifteen minutes of marked play.
Hard puzzles: mark immediately after the obvious hidden singles run out, usually around the fifteen-placement mark. Hard puzzles do not yield singles after that without help.
Expert and evil puzzles: mark almost from the start, but mark only after a full hidden-single scan. The few early placements a clean scan reveals make the initial marking faster, since each placement reduces the candidate sets of its peers.
One small habit
When you switch to marks, do it everywhere at once. Half-marked boards are worse than fully marked ones; they lie about what is constrained. Set aside ninety seconds, fill every empty cell with its candidates, then go back to solving with confidence.
For the formal version of this idea, see the pencil marks strategy page, or the workflow walkthrough at pencil marks workflow.
Keep reading
- Sudoku for Absolute Beginners: Your First Ten PuzzlesA plan for the first ten puzzles you ever play. What to look for, what to ignore, and the small habits that keep you from quitting in frustration.
- How to Solve Sudoku Step-by-StepA working method for solving sudoku, ordered from the simplest scan to the moves that crack hard puzzles. Use it as a checklist whenever you get stuck.