Recalculate history for remaining moves after each search

[Sergei S. Markoff]

Is it worth to try?
If remaining depth > 2 plies, every search could potentially change history values for remaining moves — because moves for current side could reappear at ply + 2 — so, theoretically, we can improve ordering using data from already searched subtrees.

[F. Bluemers]

Is it worth to try?
If remaining depth > 2 plies, every search could potentially change history values for remaining moves — because moves for current side could reappear at ply + 2 — so, theoretically, we can improve ordering using data from already searched subtrees.

I have played with the same idea to reorder the moves after 4moves had been searched.
Alas this did not help much,i guess the benefits are to small to overcome the extra cost.
Maybe its more effective to use moves further on in the killerlist (ply+2)
for less reducing or reordering.
Best