We write the best move when we find an exact node and a refutation move when we find a lower bound. We erase this if we end up writing an upper bound. I think the best move is likely still a decent move after the upper bound is written, and doesn't take up any extra space (we already have it allocated for moves elsewhere) so let's try keeping it around.
We write the best move when we find an exact node and a refutation move when we find a lower bound. We erase this if we end up writing an upper bound. I think the best move is likely still a decent move after the upper bound is written, and doesn't take up any extra space (we already have it allocated for moves elsewhere) so let's try keeping it around.