Too much reduction in NMR

[Masoud Habibi]

I have found some special move the best in depth = 8. imagine that one can do 30 moves after this move. and in real the 20th move in this list will lead to a very good position after for example 4 plies. but because it is the 20th move in the list it will never is check more than one or two ply because it is null move reduced because every four null move only one real move is check so at most it is check for two plies in depth = 8 or 9. How can I fix this? Should I disable extended null move reduction is nullwindow search?

[H.G.Muller]

I don't understand your problem. At d=8, after the move you have d=6 left (reduce 1, because of late-move reduction), then you do a null move, which reduces another 2 ply (so d=3 left, or d=2 when your null-move reduction is 3), and then you play a real move on the 3rd ply, so that you get in this very good position. (Allowing your opponent a move after that normally would not improve your position.) So you fail high, and the branch is re-searched even without LMR.

So it seems to me everything should be fine.