@Allis ... Thanks for your help. Yes I do understand about 'masks.'

And, obviously if I'd listened more carefully to Mac I would have looked more closely in the style maker and known that we're talking 2, 1 and 1/2 bars!

I think I now have enough of the basic idea to just play around with it to get a better feel for what happens.

I was thinking about the probabilites as '5' being neutral, i.e., an equal chance of being selected as any other '5' ... so if you had 5 cells of '5' each selectioni each cell would have a 20% chance of being slected ... where as other cells excluding 1 and 9 would have a slightly higher or lower chance of selection. For example (according to my idle speculation!) if you had 4 cells of '5' and one of '6' the '5s' might have a 18% chance and the '6' a 28% chance ... to be modified if there were 3 '5s' and 2 '6s' in some fashion to something like 16% for '5s' and 26% for '6s.'

However, I don't think this is true 'random selection,' otherwise the 'no duplicates till ... ' function would be useless. Probabilities must be affected by prior selection.

You might imagine ... yes I am fascinated by probability theory and statistics.

Thanks ...

Prado