The rectangular seeds of Domineering, Atari-Go and Breakthrough

Abstract—Recently, a methodology has been proposed for boosting the computational intelligence of randomized game-playing programs. We modify this methodology by working on rectangular, rather than square, matrices; and we apply it to the Domineering game. At CIG 2015, We propose a demo in the case of Go. Hence, players on site can contribute to the scientific validation by playing (in a double blind manner) against both the original algorithm and its boosted version. I. ALGORITHMS Given a stochastic AI, we can check its performance against a baseline program (possibly itself) as we vary the random seed. I.e. we can generate K different random seeds, and for each of these seeds play Kt games against the baseline. We can then plot the winning rates, sort, and compare the variations to the standard deviations. Results are presented in Fig. 1 an