# Question about NP and shorter efforts

Discussion in 'Power Training' started by whoawhoa, Nov 27, 2007.

1. ### whoawhoa New Member

Joined:
Oct 28, 2004
Messages:
1,029
0
I understand NP is not necessarily accurate, but if NP is greater than AP for a certain max effort, is that a good indicator that the AP for the effort could be higher given a better pacing strategy (but not necessarily as high as the NP was for that effort)?

the effort in question:

Duration: 9:04
Work: 206 kJ
TSS: 19.8 (intensity factor 1.145)
Norm Power: 389
VI: 1.03
Distance: 6.156 km
Min Max Avg
Power: 0 590 379 watts
Speed: 17.7 52.3 40.7 kph

Tags:

2. ### frenchyge New Member

Joined:
Apr 3, 2005
Messages:
4,687
4
Assuming losses in a system increase at a greater than linear rate, the most effective use of energy will be a steady effort at the highest level that can be sustained for the entire duration. NP=AP in that case, so the answer to your question is at least partially 'yes'.

I say 'partially' because your goal in this case seems to be producing the greatest power possible for the given duration. Keep in mind that that doesn't necessarily constitute the 'best pacing strategy' when the goal changes to covering a given distance in the shortest time possible, as any good pursuiter will tell you. Right Alex?

For the variable terrain on which you tested, a VI of 1.03 seems pretty darn flat. You probably went faster by varying power than you would have with a flatter power, even though AP could have been a bit higher.

3. ### Uhl New Member

Joined:
May 12, 2003
Messages:
99
0
The normalized power algorithm is not reliable for durations that short. Here's a quote from the inventor himself, from a post on the Wattage Google Group:

4. ### acoggan Member

Joined:
Jul 4, 2003
Messages:
3,047
9
I'd say that it suggests that the possibility exists, but that there are no guarantees.

In essence, what it boils down to is just how well you can tolerate the 'hit' of going significantly above your average power for a long enough period of time to result in normalized power being markedly higher than average power. Simply put, some(times) people can, and some(times) people can't.

5. ### rmur17 New Member

Joined:
Oct 5, 2004
Messages:
1,066
0
this time I have to say it's necessary to look at the file and how power was distributed over time. The aggregate measures hide too much ...

here are some simple examples for a 20-min effort averaging 300W but with markedly different time profiles. I've split it into two segments 10-min long and assumed isopower for each.
1. 300/300W yields 300W/300N VI=1.000 (best possible pacing and effort)
2. 315/285W yields 300W/302N VI=1.005 (not bad)
3. 330/270W yields 300W/305N VI=1.017 (getting worse)
4. 345/255W yields 300W/311N VI=1.035 (poor pacing)
5. 360/240W yields 300W/318N VI=1.059 (very poor pacing)
Now if we assume 300W is 20MMP ... it's clear that a modest VI of only 1.017 hides the fact you rode half the ride at 110% of sustainable pace. A VI of 1.035 masks an overpower of 115%.

I can't say exactly what those over-power excursions are going to cost you - only that I feel they will. Looking at the virtual NP alone tells you that.

Summary: Simply looking at VI or comparing NP to AP can hide a lot of pacing errors. You need to look at the file.