kukusz

I know that the Allen and Coggan wisdom regarding using normalized power to determine something like your FTP is accepted, however I was wondering if there are exceptions to this rule. I was looking at my MMP curve as a normalized curve and found my 20 minute to be much higher than my average 20 minute value (30 Watts). I then looked at the ride that generated the high normalized 20 minute power and found that the average for that portion of the ride was 90 Watts lower than the normalized average. Additionally, this occurred on a day where I was doing 1-2 minute intervals with tempo pace in between the intervals.

With such a difference, I would think that the normalized power curve has failed at this point. Like I mentioned above, my best effort in a controlled environment (rollers) where normalized power was approximately average power was almost 30 Watts lower.

The only reason I care is a really would like to dial in my training zones.

Andrew Coggan

My guesstimate is that, on average, the normalized power algorithm is accurate to w/in +/- ~5% at a duration of ~ 1 h (or longer). At shorter durations, it will tend to be less accurate, because it essentially ignores any contribution from anaerobic capacity.

cslone

Not really contributing much here, but my MMP/FTP chart is within 4 watts.

kukusz

Thank you for the reply, your book was great read. I figured the shorter the duration, the less accurate, but it's good to get an experienced response on exactly how dependable I should consider that value. I will continue to use my average power for my shorter durations. Interestingly, my normalized FTP works out to 95% of my average 20 minute power.