Pythagoras: Personally I think, that huge engame tablebases would not significantly decide the outcome of a game. Maybe 0.1% of all games could benefit. But to demonstrately use them is an indicator of panic.

Contrary to that big opening books normaly have a huge influence on chess games. But listen, it is very important to first develop a kind of intelligent playing program, relying on its own. Chess knowlegde should be the (timely) last thing to be added to a chess engine. SMIRF is not at all in a final development stage. Thus it would be much too early to implement such moves. Programs, which play badly without opening books, have been "completed" too early.

Actually I have an idea, where the difference in SMIRF's playing strength concerning 8x8 and 10x8 could be caused. It has to do with the method I reuse cached data in SMIRF. So I have some ideas how to improve SMIRF one more time - so I hope.