Rémi Coulom wrote: > with the highest upper bound (which should also be the move with the > highest lower bound if intervals are symmetrical). oops, sorry, this is wrong. Highest upper bound to prove the node is below alpha, highest lower bound to prove it is above beta. Rémi