Threshold in 20min test v NP

[Blackie]

Wondering if any of you can help. In a couple of 20min TT tests I've done to establish my threshold to use the Hunter Allen programs, I get about 300W. Yet for 20min NP today in a race I had 335W (though the avg was 284W), and a couple of weeks ago was 315W (avg 279W).

So what value should I use (the program choice is 300W, 320W, 340W) (based on 20min power, not 1 h power).

Cheers

 

[Orange Fish]

Wondering if any of you can help. In a couple of 20min TT tests I've done to establish my threshold to use the Hunter Allen programs, I get about 300W. Yet for 20min NP today in a race I had 335W (though the avg was 284W), and a couple of weeks ago was 315W (avg 279W).

So what value should I use (the program choice is 300W, 320W, 340W) (based on 20min power, not 1 h power).

Cheers

Initially I was thinking of using 320W, but based on your race NP, I would be tempted to go wtih 340W. Races are always a good way to establish your power at threshold because you will probably be pushing harder in a race than you would in a solo 20 min TT test. It appears that your 20 min NP increased from a few weeks ago, so maybe 340W would be a better threshold value to use just based on the information in this thread so far.

 

[NM87710]

I've found it more useful to use actual power instead of NP when setting FT. My workouts are setup to hit certain power ranges and NP really isn't helpful during workouts - you can only see NP post workout. Using NP for FT is like clothing manufacturers labeling an actual 34" waist as 32" so an aging population thinks they have not gained weight...

 

[BlueJersey]

I think NP is useful on a very rolling course or the difference between NP and AP is so great. From my experience, if the course is flat, your AP is a better predictor of your actual FT.

I've found it more useful to use actual power instead of NP when setting FT. My workouts are setup to hit certain power ranges and NP really isn't helpful during workouts - you can only see NP post workout. Using NP for FT is like clothing manufacturers labeling an actual 34" waist as 32" so an aging population thinks they have not gained weight...

 

[peterpen]

Wondering if any of you can help. In a couple of 20min TT tests I've done to establish my threshold to use the Hunter Allen programs, I get about 300W. Yet for 20min NP today in a race I had 335W (though the avg was 284W), and a couple of weeks ago was 315W (avg 279W).

So what value should I use (the program choice is 300W, 320W, 340W) (based on 20min power, not 1 h power).

Cheers

I'd use 300w at the most, but I'd bet it's closer to 280. I've seen 1 hr NP's of 30-50w higher than my threshold on rare occasions, certain courses will mess with the formula. If you consistently see 1 hr NP's from a variety of situations over 300w, then maybe you should adjust your threshold, but I certainly wouldn't do so based on 20' NP.

 

[ZimboNC]

Unless I'm mistaken, I think the original poster is saying that the number used for the purposes of the training program is supposed to be his 20-minute power number and not the regular 1-hour threshold power number. I'm not sure why that would be, but if that's the case then 340w is probably accurate enough as a start.

If you find yourself too fatigued to follow the plan using 340w as the number, you can always drop it to 320 or 300, right?

--Steve

 

[Blackie]

Thanks guys. Your replies and commonsense tells me to go with the programme I can complete! Will start conservatively with 300W and work up.

 

[acoggan]

Wondering if any of you can help. In a couple of 20min TT tests I've done to establish my threshold to use the Hunter Allen programs, I get about 300W. Yet for 20min NP today in a race I had 335W (though the avg was 284W), and a couple of weeks ago was 315W (avg 279W).

So what value should I use (the program choice is 300W, 320W, 340W) (based on 20min power, not 1 h power).

Cheers

The normalized power algorithm is pretty good, but tends to overestimate actual sustainable power during shorter efforts in which anaerobic capacity can play a significant role. That likely explains the 5-12% overestimation you seem to have observed. Given your apparent decision to use 20 min power as the metric by which you judge your metabolic fitness (which I definitely do not recommend), you should rely on the average power, not the normalized power.

 

[RapDaddyo]

The normalized power algorithm is pretty good, but tends to overestimate actual sustainable power during shorter efforts in which anaerobic capacity can play a significant role. That likely explains the 5-12% overestimation you seem to have observed. Given your apparent decision to use 20 min power as the metric by which you judge your metabolic fitness (which I definitely do not recommend), you should rely on the average power, not the normalized power.

And the problem is compounded if the anaerobic efforts are on/off, because AWC partially recovers following each anaerobic effort. As Andy suggests, I don't use NP in performance tests unless the duration is at least ~30 minutes. I would especially not use NP is the NP/AP ratio was greater than ~1.03.

 

[acoggan]

I would especially not use NP is the NP/AP ratio was greater than ~1.03.

IMHO that's being a bit too conservative. Normalized power from a maximal ~1 h effort is usually w/in ~5% of somebody's functional threshold power, even if the raw data are considerably more variable than what would be acceptable under the above criteria. If for some reason +/- ~5% can't be considered "close enough", then formal testing of some form or another is called for, and you might as well ignore the normalized power data entirely.

 

[RapDaddyo]

IMHO that's being a bit too conservative. Normalized power from a maximal ~1 h effort is usually w/in ~5% of somebody's functional threshold power, even if the raw data are considerably more variable than what would be acceptable under the above criteria. If for some reason +/- ~5% can't be considered "close enough", then formal testing of some form or another is called for, and you might as well ignore the normalized power data entirely.

Perhaps so. It probably reflects my newfound performance testing course. I only recently discovered that I have the perfect test course 1.6 miles from my house. This is an 11 mile straight road with a 1.6% steady grade into a prevailing 10-15mph headwind. After the climb, there is a 0.6% downgrade for another 9 miles. The prevailing headwind offsets the downgrade so it's easy to maintain power on this segment as well. I now use this course for all performance tests up to 1 hour and my NP is usually within 2W of my AP. I find that other courses that are much more variable grade can result in higher NPs than I typically do with rock steady power management, which I attribute primarily to AWC and AWC recovery. So, I'm probably just defining my criteria based on my newfound test track.

 

[peterwright]

The normalized power algorithm is pretty good, but tends to overestimate actual sustainable power during shorter efforts in which anaerobic capacity can play a significant role. That likely explains the 5-12% overestimation you seem to have observed. Given your apparent decision to use 20 min power as the metric by which you judge your metabolic fitness (which I definitely do not recommend), you should rely on the average power, not the normalized power.

Not sure I follow here. Is 20 mins considered a "shorter"effort for which NP is not a good indicator ?

Surely by estimating FTP @ 95% 20 min NP we are doing the same thing in using a 20 min effort as the metric to judge met fitness ?

I routinely use tjhis %ge of a 20 min max effort to estimate and measure FTP - am I wrong here ?

 

[acoggan]

Not sure I follow here. Is 20 mins considered a "shorter"effort for which NP is not a good indicator ?

Surely by estimating FTP @ 95% 20 min NP we are doing the same thing in using a 20 min effort as the metric to judge met fitness ?

I routinely use tjhis %ge of a 20 min max effort to estimate and measure FTP - am I wrong here ?

Let me try answering your question(s) this way: the minimum duration over which I'd recommend placing trust in normalized power depends upon:

1) the purpose for which you're using the number,

2) the precision that you need/expect (related to #1), and

3) the characteristics of the individual (relatively high neuromuscular power and relatively high anaerobic capacity appear required to "bust the algorithm").

So, while 95% of 20 min normalized power might provide a reliable estimate of functional threshold power for one person, for another it might not, either because the algorithm tends to overestimate under such conditions, or because that particular person's ratio of 20 min to 60 min power is significantly greater than or less than 0.95.

Does the above make sense?

 

[ZimboNC]

The fact remains that if you go to the web site where the Hunter Allen power training courses are offered (http://www.trainingpeaks.com/hunter/) there is a link on the left side of the page that says how to determine your threshold wattage for use in these training programs: it says to warm up and then use a 20-minute effort.

--Steve

 

[acoggan]

The fact remains that if you go to the web site where the Hunter Allen power training courses are offered (http://www.trainingpeaks.com/hunter/) there is a link on the left side of the page that says how to determine your threshold wattage for use in these training programs: it says to warm up and then use a 20-minute effort.

--Steve

That is correct. However, it is worth emphasizing that Hunter recommends using 95% of average power, not 95% of normalized power (which therefore implies that you should use a course that results in a low "variability index", i.e., ratio of normalized power to average power).

 

[peterwright]

Let me try answering your question(s) this way: the minimum duration over which I'd recommend placing trust in normalized power depends upon:

1) the purpose for which you're using the number,

2) the precision that you need/expect (related to #1), and

3) the characteristics of the individual (relatively high neuromuscular power and relatively high anaerobic capacity appear required to "bust the algorithm").

So, while 95% of 20 min normalized power might provide a reliable estimate of functional threshold power for one person, for another it might not, either because the algorithm tends to overestimate under such conditions, or because that particular person's ratio of 20 min to 60 min power is significantly greater than or less than 0.95.

Does the above make sense?

Yes - thanks. I generally use either a PT300 indoor or a steady hill - so VI is low and it seems to provide a fairly good estimate. I then allocate zones as per CP but adjust quickly as we start sessions. Seems to provide a good starting point.