How to play
- 1
Swipe, tap the D-pad, or use arrow keys / WASD to move.
- 2
Reach the exit before time runs out. Hold or swipe far to slide.
- 3
Shorter paths score more — beat the clock and find the shortest route.
What this test measures
This game measures spatial navigation and pathfinding under time pressure. Efficient routes score better than wandering.
Maps in the head, from rats to people
Edward Tolman (1948) argued from rat experiments that animals learn a map of a maze, not just a chain of turns. Rats that ran a maze for days without finding food there made far fewer errors once food appeared, so they had been learning the layout all along. Rats whose usual path was blocked tended to pick a new one pointing toward where the food had been. He called this a cognitive map, as opposed to a strip map that holds a single route.
Brain recordings later found part of the machinery: some cells in the rat hippocampus fire mainly when the animal is in one particular spot (O'Keefe & Dostrovsky, 1971). That line of work earned John O'Keefe a share of the 2014 Nobel Prize in Physiology or Medicine, with May-Britt and Edvard Moser, who discovered grid cells.
This game shows the whole maze from above, so it relies less on an inner map and more on planning a route with your eyes. When people in one study solved mazes on a screen, their eyes followed the path along its course, and their response times rose with the length of the path and its number of turns (Crowe et al., 2000).
One route, a bigger grid every two mazes
Each maze is carved by a randomized depth-first search, which links every cell without making a loop. That leaves exactly one route from the start, top left, to the door, bottom right, that never doubles back: the shortest path your steps are measured against.
Mazes 1 and 2 are 7 by 7 cells with 20 seconds on the clock, mazes 3 and 4 are 9 by 9 with 25 seconds, 5 and 6 are 11 by 11 with 30, and 7 and 8 are 13 by 13 with 35. From maze 9 on, every maze is 15 by 15 with 40 seconds. Running out of time ends the run; Finish run ends it on purpose.
Clearing maze 1 pays 500 points and each maze after that pays 500 more than the one before, plus up to 200 for time left and up to 200 for efficiency, which is the shortest route's length divided by your steps. Maze 10 pays 5,000 before bonuses, so getting through matters more with every maze. Grades count mazes cleared, not points: 4 or 5 is Average, 6 to 8 Wanderer, 9 to 11 Explorer and 12 or more Pathfinder.
Finding the route before you move
Look before you move. With no loops, any corridor that ends is a dead end, so trace forward from the start and back from the door, ruling out dead ends as you go, until the two traces meet. Every step into a dead end costs two, one in and one back out.
If you get lost, keep one hand on a wall. In a maze without loops, following one wall always reaches the exit, but it walks into every dead end on that side, so treat it as a rescue that costs time and efficiency.
Slide only toward walls you want to stop at. A slide, from holding a D-pad button, a long swipe or Shift plus an arrow key, runs straight until the wall ahead and passes side openings, so it can carry you past the turn you needed. Near junctions, take single steps; the green trail shows where you have been.
Score guide
| Grade | Range |
|---|---|
| Pathfinder | 12+ mazes |
| Explorer | 9–11 mazes |
| Wanderer | 6–8 mazes |
| Average | 4–5 mazes |
| Lost | 2–3 mazes |
| Stuck | 0–1 mazes |
FAQ
- How is the score counted?
- Each maze awards points for finishing, leftover time, and how close you stayed to the shortest path.
- How do I move on a phone?
- Swipe on the maze or tap the D-pad. Hold a direction to slide until a wall.
- Is there a time limit?
- Yes. Each maze has its own timer. Fail the clock and the run ends — or tap Finish run to cash out.
- Is this a clinical navigation test?
- No. It is a maze game for fun and practice, not a medical tool.
Sources
- Tolman, E. C. (1948). Cognitive maps in rats and men. Psychological Review, 55(4), 189–208.
- O'Keefe, J., & Dostrovsky, J. (1971). The hippocampus as a spatial map: Preliminary evidence from unit activity in the freely-moving rat. Brain Research, 34(1), 171–175.
- Crowe, D. A., Averbeck, B. B., Chafee, M. V., Anderson, J. H., & Georgopoulos, A. P. (2000). Mental maze solving. Journal of Cognitive Neuroscience, 12(5), 813–827.