Hello, somebody thinks about this option to reduce the searchtree?
The given values are: The last boards m_l := 361 * 360 The current board m_c := c * (c - 1) and m_c involved in m_l := m_p (-> m_p is subset of m_l, all lastmove boards with m_c ; see example) 1. Every matrix (every goboard) in m_c will analysed for win/loose situation when anylysing m_p. 2. Every matrix in m_c+1 = (c - 1) * (c - 2) (the next move) will also analysed for win/loose then play the move which gives the best result in m_p+1. Example: Play first move on BLACK D4. m_p = m_l - (m_l - m_c) (-> only look for BLACK D4 played moves) m_p+1 = m_l - (m_l - m_c+1) (-> try all valid moves for m_c in m_p) The idea is to use the final boards and connect them to the current move for a simple win/loose analysis. I don't implemated it yet to my engine first its only a concept hope the main idea are able to get. Constructive critics are welcome. Regards, WSK
_______________________________________________ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
