Maze Run

Navigate the maze. Reach the exit.

Game by 16Yongjin · Licensed under MIT · Source

About Maze Run

Maze Run is a top-down maze navigation game. A randomly-generated maze fills the screen. You start at one corner. The exit is at the opposite corner. Move through the maze using arrow keys or touch controls. Reach the exit to clear the level. There are no enemies, no time limit, and no score. Just the maze.

The format predates digital games entirely. Printed mazes have appeared in books and magazines for over a century. The first computer maze games appeared in the 1970s on early arcade hardware. The mechanic is one of the simplest in gaming: navigate from start to end through a constrained path. The simplicity is what has kept it relevant.

This version is from 16Yongjin's MIT-licensed HTML5 games collection. The maze is procedurally generated each time the page loads. So every run is a different maze. The algorithm produces solvable mazes only, so there is always at least one valid path from start to exit.

The game is more contemplative than competitive. There is no leaderboard, no score, no time pressure. You just walk through the maze. Most casual players spend two to four minutes per maze, depending on the maze's complexity and how lucky their first guesses are.

Controls are arrow keys on desktop, on-screen direction buttons or swipes on touch. The character moves one tile at a time in the direction of input. Walls block movement. Open corridors allow it. There is no jumping, no diagonal movement, and no walls you can break.

The visual style is minimal. Walls are dark tiles. Open paths are light tiles. The character is a small sprite. The exit is marked with a goal indicator. Nothing else on screen. The simplicity makes the maze itself the entire focus.

Strategy for maze navigation is well-studied. The most reliable rule for any solvable maze is the right-hand rule. Place your right hand on the wall to your right at the entrance. Walk forward, keeping your hand on the wall at all times. Eventually you will reach the exit. The rule works for any simply-connected maze, which this generator produces.

The right-hand rule is not the fastest path, only the most reliable. It often takes you down dead ends and back, and through long detours that a smarter route would skip. But it always works, and it requires no memory of where you have been.

Faster strategies require pattern recognition. After exploring a few corridors, you start to recognize dead ends without walking all the way to the wall. A corridor that looks like it has no branch points is usually a dead end. A corridor with multiple branches is more likely to be on the main path. This intuition is wrong sometimes but right often enough to save time.

A common bad habit is wandering randomly. Random walks in a maze take much longer than systematic exploration. Even just consistently turning the same direction at every junction is better than random choices. Pick a rule and stick with it. Then optimize.

The actual hardest maze technique is mental backtracking. When you reach a dead end, you have to retrace your steps to a previous junction and try a different direction. People with poor spatial memory tend to forget which direction they came from at junction points and get lost in the maze. The fix is to make a deliberate visual marker each time you make a junction choice, even if it is just a mental note like "I went left at the L-shaped intersection."

The game has no save state. Each maze is a single session. If you walk away from your computer mid-maze, the same maze is still there when you return. There is no time penalty. Take as long as you want. Most mazes are about three to five minutes for an attentive solver, longer for the unlucky.