lower bound of rlh in dual response none settings

How to calculate the lower bound (i.e. random selection / null model) of rlh in CBC/HB dual response none settings.

Do you just add one to the number of stimuli in one choice set?

Thanks a lot in advance.