An easy way to check would be to give the users gloves of a distinguishable color they are not wearing (it just have to cover the back of the hand) and check for that color to cover the dots.
That's one way (tape on the back of hand so it doesn't impact climbing). Another way is to make it a distinguishable shape and color (white triangle) that a Kinect could be coded to recognize. The problem lies in that for motion detection to be fun, it should be interactive, as in create new shapes on the fly and such. Instantly recognizing them would be tougher to implement, since I have little experience in motion detection.