Interesting Team Pursuit file. I've not had a powermeter on the track bike for long, but it's struck me how much variance there can be in the power in the line between efforts. I didn't realise the technical demands were quite so high before and how much you can stand to save in a zero yaw (just about) situation compared to TTT efforts on the road. Here's a screenshot from a relatively slow drill (4:48) with three riders at London. The off/on is a bit more pronounced. Apologies for having the altitude data on and the drop outs in data.