If i remember it correct so is the median value five numbers with three attempts attacking.

This should be optimal.

Real once post that past numbers matters, the five past numbers with no repeat, then the assumption is that for the next three spin one of does five numbers will hit again.

I think i save this topic, but Laurance Scott later stated it only work with certain wheels with certain conditions.

Will see if i can find the material among my AP folders ...

Cheers