In Part 3's exploration of mathematical thresholds we were looking at methods of approximating a rider's power curve, which allows us to estimate maximal power for a given duration. However, that's rarely how we ride (or run, except on the flattest courses). The models discussed assume that an athlete stays at precisely the same power for the duration of their effort, but in the real world only TTers even come close. So this post will look at ways of dealing with variation.
