10x10 othello game in python, with a Minimax based AI

Source available on github.

During my time in CPP, we had to create an othello game in a 10x10 variant, the standard being 8x8. One of the goal was to personalise the user interface, and I went with a Iron maiden theme.
We then had to develop an AI and took part in a class wide tournament in which I placed 4th out of 70.
My take on the matter was to use the element of random introduced by the dictionaries in Python, and create a simplified version of what is called the Montecarlo tree search. The adventage of this methode is to allow my program to see futher into the game than a standard deterministic tree search can allow, in the same timespan, and thus while not loosing too much in accuracy. My personal take allowed me to reach between 4 and 5 steps ahead of my opponents.
Yet, the point is not only to see far into the game but also to choose wisely which play to make and that was a difficult task. The function I used for this was a little to simple and not very well calibrated.
I haven't modified it since March 2016, and the online 'compétition en ligne' button is totaly out of order. My teacher had created an easy to use tool so that we coud play our AI matches, and disabled the server he used soon after our tournament.
Although I can assure you it is functionnal, it is absolutely not able to scale with different resolutions. If your's is greater than 1366x768, it will look ugly, and if smaller, it will be straight out not working.
Despite that, I am realy proud of it as an archivement.