How to Solve an Evil Sudoku Without Guessing

You are twenty minutes into an Evil grid, nothing has moved for the last five, and the thought arrives: what if I just try a 4 here and see what happens?

You never have to. Not on a well-made puzzle, and never on one from here - the board below is a real Evil grid from our generator, and every Evil board we hand out is checked before you see it to make sure it can be finished by reasoning alone.

Why guessing is not a shortcut

A guess feels like progress because something finally happens. What actually happens is that every digit you place afterwards inherits the guess. Twenty placements later the grid contradicts itself, and now you have a second problem: you cannot tell which of those twenty were sound and which only worked because of the assumption underneath them. Most people at that point wipe the grid and start over, having lost both the guess and the half hour of honest work built on top of it.

Compare that with being stuck. Stuck costs you nothing. The board waits, the deductions you have made stay valid, and the move you have not spotted yet is still there when you come back to it.

The first move on the board above

Look at the top-left cell of the third row - the amber one. On its own it is hopeless: nothing in its row, column or box rules out enough, and four different digits still fit there. Stare at that cell and you will conclude the grid is unsolvable.

Now stop looking at the cell and look at a digit instead. Take the 5 and ask where it can go in the first column. Three of that column's cells are already filled. Of the six still empty, rows 1, 6, 7 and 9 each hold a 5 somewhere else along their length, and the bottom-left box already contains one - which also rules out row 8. Exactly one empty cell survives, and it is the amber one. Four digits fit there; only one of them has nowhere else to go.

That is a hidden single, and it is the single most valuable habit on a hard grid. Cell-first thinking asks "what can go here?" and drowns in options. Digit-first thinking asks "where must this go?" and often gets one answer.

What Evil boards actually demand

Here is the part that surprises people. We ran our own solver over 60 generated Evil boards and recorded the hardest technique each one genuinely required:

Three quarters of Evil boards need nothing harder than a hidden single. Not an X-Wing, not a chain - a technique you already know from Easy.

So when an Evil grid stops moving, the honest diagnosis is almost never "this needs a technique I have not learned". It is "there is a hidden single on this board and I have not found it yet". Evil is hard because it starts you with 23 digits instead of 36, so the same move is buried in far more places to look - not because it demands exotic patterns.

The order to work in

When you stall, go down this list rather than reaching for the hardest thing you know:

1. Finish your pencil marks. Not most cells - all of them. Every pattern below is invisible while a single cell is unmarked, and half of all stalls are really just an incomplete grid.
2. Sweep digit by digit. Take the 1 and ask where it can go in each box, then the 2, and so on. This is the hidden-single hunt, and by the table above it ends three quarters of Evil stalls on its own.
3. Look for pairs. Two cells in a unit holding the same two candidates lock those digits away from everything else in that unit - see naked pairs and hidden pairs. Cheap, and they cascade.
4. Check the box-line interactions. If a digit sits in only one row of a box, it leaves that row everywhere else - pointing pairs, and the same idea in reverse with box-line reduction.
5. Only now the bigger patterns. The X-Wing and XY-Wing earn their keep on the last few percent of boards, and they are much easier to spot once the cheap eliminations have thinned the candidates out.

The order matters more than the list. Most people stall, jump straight to hunting an X-Wing, fail to find one, and conclude the puzzle is broken - while a hidden single sits two boxes away.

When nothing moves at all

Two things worth trying before you decide the board is impossible.

Re-derive one region from scratch. Pick the box with the fewest empty cells and recompute its candidates without looking at your existing marks. A single wrong pencil mark early on poisons everything downstream, and it is far more common than an unsolvable grid.

Put it down. Fresh eyes find hidden singles that tired ones walk past - the scan is the first thing to degrade when you have been staring at the same nine boxes for half an hour. Unlike a guess, a break cannot corrupt your grid.

How we know you will not need to guess

Most sites promise their puzzles are solvable by logic. Here is what that promise usually rests on: the generator checks that the puzzle has exactly one solution. That is a weaker claim than it sounds. A grid can have a single solution that no ordinary technique will reach - unique, but a dead end for a human.

Our generator checks both. Before a digit is removed from a board, the board is handed to a solver that knows only the techniques taught in the Sudoku School - singles, pairs, pointing pairs, box-line, X-Wing, XY-Wing - and that never tries a value to see what happens. If that solver cannot finish the board, the digit goes back. So if you are stuck on one of our Evil grids, there is a move on it, made of something you can learn.

Where to try it

The board at the top of this page is a genuine Evil grid. If Evil is a step too far today, Expert asks for the same habits with two more digits to start from - and the digit-first sweep is worth practising there first, because it is the move that carries you through both.

Where to practise this

This technique starts paying off at Evil, which is where puzzles first stop yielding to anything simpler. Reading about a pattern and spotting it on a live grid are different skills - the second one only comes from playing.

Play Evil Sudoku About Evil Sudoku