How about this idea... With a set number of plausible moves by your opponent, you draw an arrow indicating your opponent's move. Then you draw an arrow for your move. Repeating the process, you draw an arrow for another plausible move, and then your own reply.
Given on that order, your variable premove would be made based on your opponent's move.
Given on that order, your variable premove would be made based on your opponent's move.