Originally Posted by

**bitingduck**
A different way of saying this is that for a particular race you're measuring the total energy spent (and when) and how much error can you tolerate in that measurement? In a flying 200 m, you're only racing 10-12 s, so at 2 samples/sec you're only getting 20-24 data points during the timed part, and if you look all the way back to the final acceleration it's maybe 15-18 s, and 30-36 data points. But your power might be varying a lot over the course of 2-3 seconds before you hit the first tape, and you want really good data there. But you only have 4-6 points to fit, so there might be a peak that gets clipped and you could have the energy spent in those 2-3 seconds off by 10 or 15% if the samples are at exactly the worst times. So you need a faster sampling rate. In physical terms, if you're pedaling at 120 rpm and sampling at 2 samples/sec, you're getting one power sample per revolution, which probably isn't nearly enough-- given that you're doing two "power strokes" with your legs during that one revolution, you need at least twice the sampling rate, preferably 4-5 times. The power meter could still get the total power right by using an analog integrator for each point, but if you're a sprinter you probably want the details of the peaks.