Sense of direction test: method

What it measures

Spatial navigation: wayfinding from a briefly studied map (route efficiency), and keeping track of the start position while moving with a turning, first-person-aligned view (pointing error to the start, a spatial updating measure).

Procedure

Top-down view of a hedge maze at night; every run has new mazes. An optional practice level comes first (lights on, no map phase). Then 5 scored levels. Each level: the whole map (start tent, starting direction, numbered flags) is shown for 8 to 12 s; the player can start walking sooner. The player then walks to the flags in order from memory. Only the tiles within 2.8 tiles of the walker are lit, and the view turns with the walker (forward is always up). A flag counts when the walker comes within 0.5 tiles of its centre, in order. After the last flag, or when the level time runs out, the player turns an arrow to point back to the start and confirms. After each level the map shows the route walked and the pointing direction. The page records the position every 250 ms, plus at each flag and at the end. Input: arrow keys or W, A, S, D on a keyboard; on touch screens, on-screen buttons held to walk or turn. Pointing: drag around the walker, or use the turn keys or buttons; Enter or Space confirms. Leaving the browser tab ends the run. About 5 minutes.

Parameters

Scoring

Per level: route efficiency = shortest route from the start through the flags in order / (distance walked + shortest remaining route if the level was not finished), capped at 1. Shortest routes are 8-way grid paths between tile centres (a diagonal step only when both side tiles are open). Distance walked = sum of straight segments between the recorded positions, up to the last flag. Pointing error = smallest angle (0 to 180 degrees) between the pointed compass direction and the true direction from the walker's final position to the start. Primary score: mean route efficiency over the 5 levels, in percent (higher is better; 100% = the shortest route every time). Secondary: mean pointing error in degrees over the levels with an answer (lower is better; pointing at random averages 90 degrees). Also reported per level: finished, flags reached, time, map viewing time. Both scores are compared with other players: as a band (below typical = under the 25th percentile, typical, above typical = over the 75th) from 100 comparable players, and as the exact share of players beaten from 1,000. Player comparisons use only first runs (no earlier run by the same visitor, account, or IP address and device) without quality flags, with the same scorer version and the same runner settings.

Valid-run rules

All 5 levels must be sent. The server regenerates the mazes from the run's seed and checks the walk: it must begin at the start, never enter a hedge tile, never move faster than the walking speed (15% plus 0.1 tile tolerance), keep time going forward, and stay inside the level's time limit (plus 1.5 s). Otherwise the run is rejected. Unfinished levels are scored, with the shortest remaining route counted as walked. The server rebuilds every run from the raw log (and the run's seed where there is one) and computes the score itself; anything the page says about right or wrong is ignored. A log that breaks the game's rules is rejected, and nothing is stored. Stored runs can carry quality flags that are never shown and never block a result: 10 or more runs from one IP address, a missing visitor id, tab switches, and automation signals (automated browser, impossible speed, no input events, machine-regular timing).

Differences from the classic task

Sea Hero Quest is a 3D boat game with its own levels and art, played on phones and tablets. This is an original top-down maze with a limited, turning view and only 5 levels. Pointing to the start is asked at the end of every level, not only in special levels. Route efficiency here is the ratio to the shortest grid route, a different formula from the Sea Hero Quest trajectory-length measure. There is no real walking, no landmarks and no body turning.

Limitations

Browser game, played unsupervised on very different devices. Keyboard and touch controls, screen size and experience with video games all change the result. Learning the controls matters, especially on the first levels. The link between Sea Hero Quest and real-world wayfinding was shown for that game in a small sample, not for this game; this game has not been validated. The retest reliability of one run is unknown. The comparison players are visitors of the site, not a representative sample. Not a clinical test and not for diagnosis.

Languages and review status

Language-independent stimuli, instructions in all site languages (English plus 99 other languages).