At the end of January this year I purchased a power meter in an attempt to improve my training and general speed. I also performed the obligatory barrage of tests for 5 second, 1 minute, 5 minute and functional threshold power.

The results where not unsurprising but my peak one second power was. On two different occasions I recorded a 1853 watt peak and 1618 watt peak (with two weeks separating each tests). Upon seeing this I almost sent the power meter back as I just assumed the numbers were way out of my league and thus wrong. Some one later told me not to focus on the actual number as its of very little consequence when compared to other peoples?

At the time I had just finished 7 cyclo cross races and later on found out that my Shimano cranks (and stages power meter) had the wrong bottom bracket (causing the cranks to 'rock' from side to side) which could be the reason for the anomalous results. I then went into training my early base period and left training my sprint until a week ago. The highest peak test I have achieved since return is 1213 watts, would it be sensible to think I can retain that peak power again or is the possibility of those two peaks being bad data just too high?

mean maximal power curve

Strava power

  • 2
    One second is two, maybe three pedal strokes, basically "how hard can you kick". Even if the data is good, that doesn't tell you anything useful unless you're in a bar fight with a horse.
    – Móż
    Aug 11, 2014 at 1:54

2 Answers 2


Bad data, no question whatsoever. Muscle fatigue develops quite rapidly, but even so there is minimal decrease in power for at least a handful of seconds. (Indeed, even the fitted curve you've shown seems to be in error, likely as a result of the way GC's fitting algorithm will chase "noise".)

  • Just as a side note the curve is from strava.
    – user95786
    Aug 10, 2014 at 14:40
  • I was shocked to see that no one recognised you.
    – andy256
    Jun 24, 2016 at 1:42

Putting aside the crank fitting problems, you can forget about using data from a Stages power meter for reliable neuromuscular (very short duration) power assessment. It's simply not suitable for that task (not many power meters are).

  • 1
    Are you able to provide a reference to back this? If I were to make a counter statement how would someone know who was correct.
    – mattnz
    Aug 11, 2014 at 23:45
  • Welcome to Bicycles Alex! While some people may realize that you are a professional cycling coach, many (especially international) readers will not. So while you have lots of expertize to share, your voice will be heard more clearly with evidence-supported answers. Such answers certainly take more work to write, but become more durable references.
    – andy256
    Aug 12, 2014 at 0:21
  • Sure. See this sample mean maximal power charts to explain what I mean: dcrainmaker.com/images/2013/09/max_power_curves_thumb.png Aug 12, 2014 at 8:12
  • Use of a power meter for maximal pedal force - pedal speed testing (which is of keen interest to sprint ability) requires the power meter to perform well and accurately under such high loads and in an accelerating scenario. Unfortunately very few power meters tests address this element of testing. Aug 12, 2014 at 8:23
  • Plotting the PF-PV line for a maximal effort of up to ~ 6-seconds should reveal a linear plot, thereafter losing linearity as fatigue sets in. This has been demonstrated many times in cycle sprint testing, e,g.: ncbi.nlm.nih.gov/pubmed/17562069 but you'll need to examine underlying data If your meter doesn't show this PF-PV relationship, then it suggests it's perhaps not up to the task of correctly measuring neuromuscular power. Stages of course also suffers from the normal lack of symmetry in maximal force efforts, i.e. assumption power = left crank x 2. Aug 12, 2014 at 8:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.