Is a Simplified Pathfinding Implementation Acceptable for a Bachelor’s Thesis?

Be careful with the terminology. A 1,000-line codebase is usually a particular optimized implementation, not the algorithm itself. You may be able to implement the core ideas much more simply, but removing certain steps could change the algorithm’s guarantees or behavior. If the result is substantially different, call it a simplified or inspired implementation rather than claiming it is the complete algorithm.

A simplified implementation can be completely reasonable, especially if the algorithm is supporting the game rather than being the main subject of the thesis. The important parts are to define exactly what you implemented, explain what you omitted, and avoid presenting it as a full or faithful implementation of Polyanya if it is only a simplified version. Your supervisor ultimately has to decide whether that scope meets your program’s requirements.

Switching to A* or another manageable method is not a failure. A thesis does not become stronger merely because the supporting algorithm is more obscure or difficult to reproduce. Since your actual contribution is the game and its evaluation, a well-understood algorithm that you can implement, test, and explain is likely better than an ambitious algorithm you cannot properly validate.

Ask your supervisor before committing to the simplified version. They can tell you whether the thesis needs an original technical contribution, a faithful reproduction, or simply a working method that supports the game study. Bring them a concrete scope: the core features you will implement, the optimizations you will exclude, and how you will test that the resulting pathfinding works.