Originally Posted by
echappist
these comparisons aren't strictly apple to apple, as the average for the last one is higher than that of the first (and second). The average for the last one is 308W, and a good chunk of the speed increase is due to higher overall power output
A more apples to apples comparison would have the additional constraint of same overall average power.
Personally i'd get a good estimate of my 25 min power, shave a few percent off of that for fatigue. That's the average to shoot for the entirety of the climb.
For a climb where the shallow ramps last half as long as the steep ramps, 90% for shallow and 105% for steep. The spread should get narrower if the estimated duration is longer.
If I were doing this race, I'd shoot for 98% FTP for that 25 minutes. This may be on the conservative side of things (to account for fatigue due to the three hours leading into the climb), but it reduces risk of blowing up. After 3-4 minutes on the steeper ramp, I'd get a sense of how well i'm going and can always up the watts.
The average power argument doesn’t always hold because generally you’re capable of higher output over less time duration but power drops off for longer times. But if we did use 308w for both segments the combined time comes to 26:11 in this example, still much slower than the last split example. The reason is that for power vs. speed the speed drops off faster as a percent on a grade than on a flat road. In this example 300w=>270w reduces your speed by 4% on a flat road but by 9% on a 6.3% grade.
But we are talking about the same apples here. Pick a target power, go about 90% of that on the flat and then give it all you got for the next 15-16minutes up the hill, this will produce the best time for whatever condition you find yourself in at that time of the day.
If the duration was longer you wouldn’t change the spread for power but instead you would lower your target power inline with that duration, this is what would give you the best time.