Compare the two Stitches Puzzle approaches by testing whether each produces one A–B stitch and one B–C stitch, with no shared cell and exact row and column endpoint counts.
That comparison is between two reasoning orders—clue-first and placement-first—not two built-in modes. In the fixed 3×3 Stitches Puzzle, three vertical regions are labeled A, B, and C. The task is to place two stitches, one for each neighboring region pair, without using the same cell as an endpoint twice. Because this board is disclosed rather than randomly generated, the useful comparison is about which method exposes errors sooner.
The two approaches at a glance
A clue-first approach reads the printed numbers before choosing a border. A placement-first approach starts with the region-pair obligations and then checks the clues. Neither order changes the rules, but each emphasizes different risks.
Clue-first solving
Begin with the row clues 2, 0, 2 and column clues 1, 2, 1. Because every stitch contributes two endpoints, the zero in the middle row excludes both middle-row candidates. A clue of 2 in the top and bottom rows allows one stitch in each of those rows. You then apply the requirement to use A–B once and B–C once.
This method is useful when you need to rule out locations immediately. Its limitation is that endpoint totals do not automatically decide which region pair belongs in the top occupied row and which belongs in the bottom one.
Placement-first solving
Begin with the two mandatory pairs: one A–B stitch and one B–C stitch. Select their borders while keeping their endpoint cells distinct, then audit the row and column totals. This order directly guards against using a pair twice or treating one cell as two holes.
Placement-first reasoning is useful when the region-pair rules feel clearer than the clue display. Its limitation is that satisfying both pair obligations does not automatically produce the correct endpoint totals, so a final clue check is still required.
Compare their order of deduction
The approaches can be judged by where they begin, what they check first, and where they tend to expose a mistake.
| Criterion | Clue-first approach | Placement-first approach |
|---|---|---|
| Starting point | Row and column endpoint budgets | A–B and B–C pair requirements |
| First useful exclusion | No stitch can occupy the middle row | A region pair cannot receive a second stitch |
| Checks most directly | Exact endpoint totals | Pair completion and separate cells |
| Common blind spot | May leave the top-versus-bottom pair assignment open | May satisfy both pairs while missing a clue |
| Required finish | One A–B stitch, one B–C stitch, no shared cell, and exact totals | The same four checks |
The strongest comparison combines the methods. Use clue-first reasoning to eliminate the middle row and establish the two occupied rows, then use placement-first reasoning to assign one A–B stitch and one B–C stitch without sharing a cell. Finish with a separate row and column audit. This hybrid approach catches both location errors and region-pair errors without treating any single check as sufficient.
Turn the clues into endpoint budgets
A row or column clue counts endpoints, not whole stitches. The Puzzle Stitches FAQ and rules confirms that each stitch contributes one endpoint to each of its two cells. The region, hole, and clue framework is also cross-checked by Puzzle Genius.
| Constraint | Printed value | What it requires |
|---|---|---|
| Rows, top to bottom | 2, 0, 2 | One stitch in the top row, none in the middle row, and one in the bottom row |
| Columns, left to right | 1, 2, 1 | One endpoint in A, two in B, and one in C |
| Region pairs | A–B once and B–C once | Two stitches in total |
| Cell use | One use per cell | Two stitches cannot share the same endpoint cell |
The row clues force the two stitches into different rows: one above the middle row and one below it. The region-pair rule then means they must use different boundaries, so one is A–B and the other is B–C. The column clues require region B to receive both endpoints while A and C each receive one.
The disclosed solution places A–B across the top row and B–C across the bottom row. The rule set also permits the mirrored pair assignment: B–C across the top and A–B across the bottom. It produces the same 2, 0, 2 row totals and 1, 2, 1 column totals while using four different cells. The disclosed route is therefore a verified solution, not proof that the arrangement is unique.
Using region B in both rows is not a shared-hole violation. The top cell in column B and the bottom cell in column B are different cells. The rule prohibits reusing one exact cell, not placing two endpoints in the same column.
Solve the fixed Stitches board
- Open the puzzle and identify the three vertical regions A, B, and C from left to right. Read the row clues as 2, 0, 2 and the column clues as 1, 2, 1 before selecting any border.
- Exclude the entire middle row because its clue is zero. Plan for one stitch in the top row and one in the bottom row.
- Use the arrow keys to move through the six candidate borders. Up and Down change rows, while Left and Right change the selected region boundary. Confirm the button’s announced row and pair before placing a stitch.
- For the disclosed route, select the top-row A–B border and press Space or Enter. Space or Enter places the selected stitch, and pressing it on an already placed stitch removes that stitch.
- Read the validation result after the placement. The top A–B stitch contributes two endpoints to the top row, one to column A, and one to column B. Column A already matches its clue at 1 of 1; columns B and C are underfilled at 1 of 2 and 0 of 1, so the puzzle remains playable because another legal stitch may complete them.
- Select the bottom-row B–C border and press Space or Enter. Its endpoints are in the bottom A? No: the bottom B and bottom C cells. Those cells are separate from the top-row endpoints, so the two stitches share no cell.
- Confirm that A–B and B–C each appear once, the row totals are 2, 0, 2, and the column totals are 1, 2, 1. To test the other reasoning order, restart and try the mirrored placement of B–C on top and A–B on the bottom.
Read validation as a rule audit
Validation runs after every placement. It immediately reports contradictions such as assigning the same region pair more than once, using the same cell as an endpoint for two stitches, or pushing a row or column above its printed clue.
An underfilled clue is different from a contradiction. The first A–B stitch leaves some column totals incomplete, but the puzzle remains playable because the B–C stitch can still fill those totals. This distinction is important when comparing approaches: a method that checks early should not reject a legal intermediate position simply because it is not yet complete.
Restart returns the board to an empty state and restores the first candidate, score, and contradiction state. That makes it possible to test clue-first and placement-first orders on exactly the same fixed fixture rather than comparing two different boards.
Choose the approach that exposes errors sooner
Choose clue-first solving when you are unsure which rows can contain stitches. Choose placement-first solving when you need to concentrate on completing A–B and B–C without duplicating a pair or a cell. For most players, the hybrid sequence is clearest: eliminate the middle row, assign the two different pairs, avoid a shared cell, and then verify every endpoint total.
The score calculation is max(0, 1000 − 50 × placement actions). For the disclosed clean two-stitch solve, 1,000 − 50 × 2 = 900. A mirrored clean route also uses two placements, so the methods tie on score when both are solved without extra actions. No approach has an inherent scoring advantage.
There is also a clear limit to the comparison: this product uses one disclosed three-by-three fixture rather than random boards, so the result does not establish which method would perform better on larger or generated Stitches puzzles. The approaches are manual ways to order deductions, not selectable game modes. Compare them by legal completion and action count, then use whichever ordering helps you catch a rule conflict before the final audit.
For a deeper look, see Can I Paint Cells on the Source Half in Symmetry Puzzle?.
For a deeper look, see How Do I Choose My Next Step in Tents Puzzle?.