Killer Sudoku Tips

One: read every cage before you write anything

The temptation is to start solving at the first cage that resolves itself. Resist it for one full pass. Reading all of them first tells you where the anchors are, and anchors are not spread evenly: a board can carry four forced cages clustered in one corner and none at all in the opposite one. Knowing that before you commit tells you which end of the grid to work from, and that choice is worth more than any single technique.

Two: place what is forced, then stop placing

Fill the cages that admit only one combination, and then deliberately switch modes. The second phase is not about writing digits, it is about writing possibilities, and mixing the two is where errors are made. A digit written down as a guess looks exactly like a digit written down as a certainty when you come back to it twenty minutes later, and there is nothing on the grid to tell them apart.

Three: run the 45 sweep on a schedule, not on a hunch

The two moves themselves – the cell an inch inside a region and the cell an inch outside it – are set out on the hard page. What belongs here is when to look for them, because that is the part people get wrong. The sweep is mechanical: twenty-seven units, each one checked for a single cell sticking into or out of the cages that cover it, no candidate analysis anywhere in it. Which is exactly why it gets skipped. It does not feel like solving.

So do not save it for when you are stuck. Run it once as soon as the forced cages are in, and again every time the board has gone quiet for a minute. It is quick and it terminates: it either produces a placement outright or proves there is none to find, and both of those answers are worth more than the ten minutes you were about to spend staring at one cage.

The trap: big cages are not hard, they are empty

This one was measured while the generator behind this site was being built, and it is the opposite of what almost everyone assumes. Growing cages larger does not make a puzzle harder. It makes it ambiguous. Layouts built out of five-cell cages almost never had a single answer at all, and the few that did could not be finished by reasoning; layouts built from two and three-cell cages produced solvable boards across every grade. A large cage with a total near the middle of its range genuinely does not constrain anything, so the board stops being determined and has to be thrown away.

The consequence at the board is simple enough. When you meet a big cage, note only the digits it cannot contain and move on. Come back once its neighbours have narrowed it, at which point it usually collapses in a single step.

What to do when you are properly stuck

Count. Take a row that is half full and ask what it is missing; those missing digits have to fit into the cells that remain, and the cage totals crossing that row will usually rule out most of the arrangements. If that fails, go looking for a digit rather than a cell: pick a number, mark every cell in one box that could still take it, and see whether they all fall in the same row or the same column. If they do, that digit can be struck from the rest of that line. It is the elimination that unblocks most stuck boards, and it fills in nothing at the time.

The habit that matters most

Keep the pencil marks honest. Every technique above the beginner level is a statement about where a digit can still go, so a stale candidate list does not merely slow you down, it actively misleads you into wrong placements. When you write a digit, immediately strike it from its row, its column, its box and its cage. Killer sudoku strategy sounds as though it ought to be a catalogue of named techniques, and at the very top of the game it partly is. Everywhere below that it is bookkeeping. Most killer sudoku tips are this one habit wearing different clothes.