Prince of Persia (1989) Level Replica

My role

I replayed and researched the original first level, then rebuilt its movement, gates, buttons, falling floors and encounters in Unity/C#.

Ledge grabbing was the hardest mechanic. Recreating the original pauses and animation timing with free Unity assets required approximation.

This is a non-commercial educational project and is not affiliated with or endorsed by the owners of Prince of Persia.

Visual assets in the public build are AI-generated placeholders. The movement, level logic and gameplay code are mine.

Assessed 5 out of 5 over 162 tracked hours, in a university procedural-content-generation research context. Completed 2 April 2024 and assessed by the supervising researcher against final-year project expectations, with autonomy and problem-solving noted. The Unity prototype focused on recreating the first level’s movement, gates, pressure plates, falling floors, hazards and enemy behaviour.

My contribution

• Rebuilt the first level’s movement, gates, buttons, falling floors and encounters in Unity/C#.
• Authored ledge triggers with hand positions, standing positions and facing rotations for grabbing and climbing.
• Compared the level repeatedly with the original to adjust traversal timing, camera framing and interaction order.

Design decisions

Mechanics before embellishment

The exercise prioritised fidelity to the level’s interaction sequence and pacing instead of expanding the original design.

2D reference, 3D implementation

Camera framing and level geometry preserve the original side-view read while using a 3D environment.

What was hard

Translating a tightly timed 2D reference into 3D without losing its structure.

Broke the level into mechanic groups and used repeated comparison with the original first-level flow.

Outcome and evidence

I completed the supervised educational recreation and released a short Windows build. This non-commercial exercise is separate from the original game’s creators and rights holders.

• The project record describes repeated reference play and a complete start-to-finish comparison of the recreated level.

What I learned

Authored ledge positions solved the level’s immediate needs, but tied the controller to that geometry. For another level I would test reusable ledge detection while keeping the reference timing as a separate tuning problem.