Many power programs and discussions revolve around the concept of the "critical power" or CP, which is defined to be the power that one can just hold for some amount of time. For instance, the CP-20 minutes is the power that one can hold for 20 minutes but not longer. There is a common misconception that one can test the critical powers by doing test of the corresponding length. That is, if you want to know your CP-20, you just go out and ride as hard as you can for 20 minutes, but this is sort of silly since pacing over 20 minutes has a pretty big impact on your average power which means that you'd have to know your CP-20 before testing to ride the right effort for the test. If you don't know it, how do you know that you did the right effort? That you couldn't have ridden any harder and sustained it for 20 minutes? If you lose power late in the test, how do you know how much higher the average could be if you started just a little easier?

Luckily there is another much more effective approach. Rather than trying to test for exactly a particular amount of time, try testing for exactly a particular power until you can't hold it any more. Now you know your CP for the amount of time you actually survived on the test. After recovering for a few days, repeat the test at a higher or lower power. It turns out that a plot of critical power versus time will be a pretty much straight line so long as you are testing aerobic power outputs. That means that if you plot a few CPs and draw a line through them, you can predict the CP for neighboring times just by reading them off the graph.