Picture-grid puzzles look like maths and play like detective work. Here is the handful of habits that turns a blank sheet of squares into a toy, one certain cell at a time.
What the numbers along the edge are telling you
Every row and column in a picture grid comes with a short list of numbers. Each number is a run of shaded cells, in order, with at least one empty cell between runs. So a clue of 2 4 means: somewhere in this line there are two shaded cells together, then a gap of one or more, then four shaded cells together. Nothing else in that line is shaded.
The clue never tells you where the runs start, only how long they are and in what order. That is the whole puzzle. Your job is to work out positions by comparing the clues of crossing rows and columns until only one arrangement fits. A finished grid satisfies every clue at once, and on a well-made puzzle there is exactly one way to do that.
It helps to say a clue out loud the first few times. Two, gap, four is far easier to picture than a pair of numbers on a screen. After a dozen grids you will stop translating and simply see the shapes, the same way a knitter reads a pattern without counting every stitch.
The overlap trick that starts almost every grid
The single most useful idea in picture grids is the overlap. Imagine a 10-cell row with a clue of 7. Push the run as far left as it can go and it covers cells one to seven. Push it as far right as it can go and it covers cells four to ten. Cells four to seven are shaded in both cases, so they must be shaded. You have filled four cells without guessing.
The same trick works with several runs. Add up the clue numbers plus one gap between each pair. If a row of ten has clues 3 and 5, that totals nine, which leaves one spare cell. Slide everything left, slide everything right and shade wherever both positions agree. On big clues you will often fill most of a line on the first pass.
Start every new grid by scanning for the largest clues and applying the overlap to each. In Workbench Blueprints the 10 x 10 grids usually give you a handful of certain cells straight away, and those cells become the foothold for everything else. A blank grid stops being blank within the first minute.
Crosses are just as valuable as shading
Newcomers tend to treat empty cells as an afterthought. They are not. Marking a cell you know must stay empty is information, and it shrinks the space in which the remaining runs can sit. Our grids have a cross mode for exactly this reason, and drag-painting lets you cross out a long stretch in a single stroke.
When is a cell certainly empty? When every run in its line is already complete, the rest of the line is empty. When a run of three is fixed in place, the cells at either end of it are empty. When a gap between two shaded cells is too small for any remaining run, it must be empty too. Each of these gives you crosses for free.
Crosses also stop you second-guessing yourself. On a 15 x 15 grid it is easy to lose track of which cells you ruled out ten minutes ago. A neat row of crosses is a note to your future self that says this part is settled, and it lets your attention move on to the lines that still need work.
Working the crossings when a line gets stuck
Sooner or later every grid stalls. You look at a row and nothing more can be said about it. The answer is almost always in the columns that cross it. A column might now have enough shaded cells to complete one of its runs, which gives you new crosses, which in turn give the stuck row a new constraint. Picture grids are solved by going back and forth like this.
A good habit is to sweep. Work through every row from top to bottom, then every column from left to right, asking each one a simple question: has anything changed since I last looked? If it has, apply the overlap again with the new information. If not, move on. The clue highlighting in Workbench Blueprints helps you keep your place as you sweep.
If an entire sweep produces nothing, look for edges. A shaded cell right beside the border is a powerful clue, because the run containing it must start at that edge. Cells at the edge of the grid often break a stalemate faster than anything in the middle, which is why experienced solvers watch the border so closely.
Moving up from 10 x 10 to the bigger blueprints
The 15 x 15 grids use exactly the same logic, but there is more of it to hold in your head. The trick is to treat the big grid as a set of smaller problems. Solve the obvious long clues first, then work into one region at a time, perhaps the top of a locomotive or the arms of a silver robot, rather than sweeping the whole sheet at every pass.
Expect a big grid to take three or four times as long as a small one, and enjoy the middle stretch where nothing seems to happen. That is usually when the picture is quietly forming. A sudden run of certain cells tends to follow, and suddenly the outline of a water tower or a robot dog is unmistakable. The final few rows almost fill themselves.
If you are playing with a child, the small grids are a lovely shared puzzle. One person reads the clue, the other counts the cells, and you argue happily about whether that square really must be empty. By the time you reach the bigger grids together, you will both have the overlap trick firmly in your fingers.