I managed to win the local crit again - yes folks, D-grade. Well I had to work at it! I had to avoid falling (one rider down as a faster grade caught us on a corner - nasty!) and I had to watch for breaks (just one semi-serious attempt, easily caught). And I had to patiently wait for the impetuous youngster to start the sprint.
More importantly it gave me more race data. So I can confirm that last week's 1400W burst was indeed an error on the ibike's part, as expected. I'll show you the data later but every lap we went over a small hill, and each lap the hill got steeper. Or so the ibike thought. When 'corrected' it's still a 900W effort (bridging a last-lap gap). This week's data is much more consistent and the peak power a more miserly 800W. I was careful not to expend too much energy in short bursts, rather I anticipated accelerations and smoothly bridged. Each lap the hill registered between 300 and 5ooW effort and 42% of the race was above 200W. If you trust the ibike, of course!
It's a slightly downhill sprint so although I briefly hit 55kmh the power was just on 600W.
I have upgraded to ibike firware v1.16. I always reset after a ride and do a re-tilt when changing bikes. I have a battery of coast-down data to tap into a well. It's not perfect, it certainly goes awry when the barometer is moving around, and if you lift the bars or otherwise drastically alter your weight distribution during a ride then it can generate some flaky figures... but it works well enough to be a great tool for the data junkie on a budget.