aMaze

You can't stop where you want to

aMaze looks like an ordinary maze until you realise you don't control where the blob stops, only which way it starts moving. Press a direction and the blob slides until a wall or the edge of the board halts it, painting every open cell it rolls across on the way. That single rule is the whole puzzle. You're not steering a path so much as firing the blob down corridors and using the walls as brakes, which flips the usual question: it's never simply where you want to go, it's which wall will catch you when you get there. Controls are whatever's comfortable (arrow keys, WASD, a swipe, or the on-screen d-pad) but the thinking is identical however you send the blob off.

Walls are brakes, and repeats are free

Two facts shape every good route. First, the board edge counts as a wall, so the outer rim of the level is one big brake you can always rely on. Second, re-crossing painted cells costs nothing. You don't need a tidy one-pass route that never doubles back; you need any route that eventually touches every open cell. Shuttling back and forth through painted territory to line up the next stop isn't a mistake, it's how most levels actually get solved, and a route that looks wasteful but covers everything beats an elegant one that leaves a single cell white.

That's also why guessing feels productive early and falls apart later. On small boards, random slides fill most cells by accident, and cells sitting in the middle of open space get painted in passing whether you plan or not. On bigger boards there are cells that only one particular slide will ever cover, and no amount of wandering lands on them by luck.

Find the anchor cells first

Before your first move, scan the level for the cells a slide would skate straight past. Say one open cell pokes off the side of a long corridor. A slide along the corridor paints the corridor and ignores the notch. The only way to claim that cell is to stop level with the opening and then turn into it, which means you need a wall that halts you at exactly that spot. Those are the anchor cells, and they decide the route. Work out which wall you'd have to stop against to reach each one, then plan your earlier moves to deliver the blob onto that wall.

From there, think a slide or two ahead the way you'd read a billiard shot: where does this slide end, and what does that new position let me do next? When a level looks impossible, it's nearly always one or two awkward cells left over at the end. Trace which slide could ever stop on them, build backward from that, and the order of the earlier moves tends to fall into place on its own. One more habit pays off in the late levels: before any slide, glance at what it paints along the way, not just where it stops. The best moves do double duty, mopping up a stray row of cells while travelling to the wall you actually need.

An old trick that keeps working

Slide-until-something-stops-you is one of the great puzzle mechanics. Board game fans know it from Ricochet Robots, a 1999 design by Alex Randolph in which pieces skid across a grid until they collide with a wall or each other, and video game players have been slithering across the frozen floors of ice caves in role-playing games since the 1990s. aMaze bolts one extra demand onto that tradition: the paint. Arriving somewhere isn't enough. You have to cover everything on the way.

There are around fifteen levels, each a little larger and more tangled than the last. The early boards teach the sliding almost by accident, because they're small enough that any energetic wandering covers them. Somewhere in the middle of the run the anchor cells start to matter, and the last few levels are honest planning puzzles where a careless first slide can wreck the shape of the entire route. When that happens, stop pressing and start reading: find the awkward cells, work out their stopping walls, and build the route from them backward.