I think based on my analysis of the lab “Simple RNA Switch” that if the switch is from one hairpin on a base loop to another hairpin on a base loop with an aptamer in the middle in the 2nd state the hairpin before the aptamer needs to have a high pairing probability on the dot plot. This would be visible by a darker color of the leg associated with the specific stack for the hairpin on the dotplot. The stack connecting the aptamers loop to the main loop does not need to have as high of a pairing probability. I would thus award designs with a higher pairing probability for the hairpin on the aptamer in the 2nd state and penalize designs that have a low pairing probability for the hairpin in the 2nd state.
Link to paper to support:
https://dl.dropboxusercontent.com/u/87351147/Lab%20Analysis_Simple%20RNA%20Switch_Rev1.pdf