Zitat von
ArneB
Playing with a plan based chess search system against UCI limt strength modes in other non-plan based engines does not really make sense, since the UCI Limitstrength will at some point make occasional sub optimal moves or even blunders to compensate or to reach the desired reduced level of play.
But to adapt the play of a lower ELO you have to play sub optimal moves. The limitation of nodes is in my opinion the best way to get a human like behaviour. You cannot expect Karpov 2100 to play like Karpov and at the same time with 2100 ELO. Wouldn't it be better to look for a setup that adapts Karpov's style at a level of 2800 ELO, and later only reduce the nodes or depth to calibrate the ELO levels?