Code: Select all
VALUEL_P_MG = 0.74 +/- 0.02 [190]
VALUEL_M_MG = 3.03 +/- 0.09 [776]
VALUEL_R_MG = 3.90 +/- 0.21 [998]
VALUEL_Q_MG = 8.59 +/- 0.41 [2200]
VALUEL_P_EG = 0.93 +/- 0.02 [239]
VALUEL_N_EG = 3.04 +/- 0.06 [779]
VALUEL_B_EG = 3.32 +/- 0.06 [850]
VALUEL_R_EG = 6.08 +/- 0.11 [1556]
VALUEL_Q_EG = 11.08 +/- 0.26 [2836]
VALUEQ_BB_MG = 0.13 +/- 0.03 [33]
VALUEQ_BB_EG = 0.23 +/- 0.04 [59]
VALUEQ_PASS_MG = 0.71 +/- 0.18 [183]
VALUEQ_PASS_EG = 3.11 +/- 0.09 [796]
VALUE_WIN_EG = 4.44 +/- 0.15 [1136]
MOBILITY_N_MG = 0.00 +/- 0.00 [1]
MOBILITY_N_EG = 0.00 +/- 0.00 [0]
MOBILITY_B_MG = 0.00 +/- 0.00 [0]
MOBILITY_B_EG = -0.00 +/- 0.00 [-0]
MOBILITY_R_MG = -0.00 +/- 0.00 [-1]
MOBILITY_R_EG = 0.00 +/- 0.00 [1]
MOBILITY_Q_MG = -0.00 +/- 0.00 [-0]
MOBILITY_Q_EG = -0.00 +/- 0.01 [-1]
PST_CENTRE_P_MG = 0.01 +/- 0.00 [3]
PST_CENTRE_N_MG = 0.05 +/- 0.00 [12]
PST_CENTRE_B_MG = 0.05 +/- 0.00 [12]
PST_CENTRE_Q_MG = -0.00 +/- 0.01 [-1]
PST_CENTRE_K_EG = 0.04 +/- 0.00 [11]
PST_CENTRE_N_EG = 0.08 +/- 0.01 [20]
PST_CENTRE_B_EG = 0.05 +/- 0.01 [12]
PST_CENTRE_Q_EG = 0.06 +/- 0.01 [16]
The PST_* terms are just the weight for the centre in the PST (currently the only component in the tables) and are multiplied by the following table to get the piece-specific PST:
Code: Select all
-7 -4 -2 -1 -1 -2 -4 -7
-4 -1 1 2 2 1 -1 -4
-2 1 3 4 4 3 1 -2
-1 2 4 5 5 4 2 -1
-1 2 4 5 5 4 2 -1
-2 1 3 4 4 3 1 -2
-4 -1 1 2 2 1 -1 -4
-7 -4 -2 -1 -1 -2 -4 -7
Code: Select all
Knight : 2 5 4 8 { -3 -2 -1 0 1 2 3 4 5}
Bishop : 7 9 7 13 { -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7}
Rook : 14 14 14 14 { -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7}
Queen : 21 23 21 27 { -20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7}
Has anyone successfully managed to tune mobility weights using test positions? The only explanation I can think of is that it is the interaction with the search that makes mobility a valuable term, but I find that highly surprising. Thoughts?