My entry for the Mario AI Championship 2012, implementing adaptive procedural level generation. Holds 1st place at the all-time points and percentage tally.
For the algorithm, it would be very convenient to have all chunks to a fixed length of 2. We can predict how many iterations we need to fill a level of a given length. That also ensures that the WorkerState genotype of any two levels of the same length has also the same length. It enables us to use genetic shenanigans.
For the algorithm, it would be very convenient to have all chunks to a fixed length of 2. We can predict how many iterations we need to fill a level of a given length. That also ensures that the WorkerState genotype of any two levels of the same length has also the same length. It enables us to use genetic shenanigans.