First long training run and write-up completed

Have ran a long experiment training an agent to learn via self-play for around a month on my desktop machine. Full evaluation, including results using a forward search on top of the trained policy are complete and the write-up draft if finished.