Beginner question: FTP targetted training?

jbank said:

5min long hill and do an average of 370watts, or about 5.5 watts/kg.

Good.

jbank said:

So I've got two questions. First, does it sound like I'm doing that
baseline testing right?

You're new to training with power. One of the first lesson you may learn is that now, several training can be considered as testing. "Training is testing, testing is training" (I believe this quote may be from Andrew Coggan).

If you're on the impression that for this test, type of course wasn't ideal for generating sustained power, then result might have been different on some other course. That doesn't matter all that much, because as you continue logging in some training, best sustainable power over 20min will be reached

jbank said:

Do people manage to get consistent watts
for 20 minutes in rolling hills?

I do not. I let it drop while descending.

This is why it is very important to not care about AVG power over 20min as much as Normalized Power over 20min. Did you use Normalized Power in your analysis for this 20min test?

jbank said:

Second, assuming that my testing
is right, it seems like my 20 min is relatively worse than my 5 min
based on the charts I've seen. What sort of rides would help
target FTP to improve the 20 min power? I mainly ride triathlon's

If you ride triathlons, olympic distances, then I'd say that you're fortunate enough to have the Bike leg having a duration that is very close to that prescribed for FTP definition.

I'd say that any work lasting 20min or more with Normalized power that is close to max for this duration will pay off a lot. The 5.5 w/kg for the 5 minute test suggests that you are relatively gifted on the fitness side (especially Vo2Max), now you need to use wko+ along with your powertap to "tap" into triathlon specific type of work.

My favorite test is quite simple: 60min all out for this duration. Period. Because that factors in my Pacing ability. But anywhere between 20min intervals to 90-120min rides (tempo) will greatly benefit to improve your time-trial skills.

Thanks for the response.

I don't have wko+, I've just used the PowerAgent program that
came with the PowerTap. I don't think that gives Normalized
Power (at least if it does it isn't obvious). I gather from your post
that wko+ is the right tool to go along with the power meter.
I guess I'll give the trial version a spin...

Thanks,

-Joe

Your Powertap purchase was the first and most important step in allowing you to train smartly, now the second most important step is using it along with the appropriate software, and that is wko+

Without being in the *knows*, I suspect Saris to not be in a hurry to improve their agent, since they know there`s something out there that is doing much more than they could ever hope.

What I am trying to say is that I am not even sure that significant r&d is allocated to try and match wko+

If I was them, I wouldn't even try.

- - - - -

Now, you still don't know the result of this 20min test on rolling terrain. Avg is irrelevant. You'll have to wait and get computed average which is what's called Normalized Power. This important variable will put more weight, give more importance to intense work, thus raising the average so that it can precisely quantify what this training dose / test actually was.

You basically get a value that represents what this workout might have been if performed on a flat terrain (somehow).

In the mean time, there would be nothing wrong in using this 275w as a tentative FTP baseline. Worst case, later down the road you may need to adjust it. Or you may start with a value that is little underestimated (that is what I prefer, for what it's worth).

When you first install Wko+, you'll quickly be asked about your FTP, since it is probably the most important parameter for the good functioning of the program. What about plugging in 270 and see if it makes sense later after 2 weeks of workouts.

But again, I know many won't agree with me on that. So take it as you want. But as a triathlon specialist, to me, there's just one very valid FTP test and it is a flat 60min on a course that allows for riding this duration with no interruption. I do them on a FormulaOne race track (I am very lucky). You live near New-York, friends of mine love testing on Mount Whiteface. There you have a 60min climb with no cars (from 6pm or 6:30 pm).

Or even better mate, just check in to an event of secondary importance, race with your powertap. Perform the Tri as if there was no running leg, and take it easy during the 10k. That will give your your FTP right there, with the motivational edge in bonus.

SolarEnergy said:

Now, you still don't know the result of this 20min test on rolling terrain. Avg is irrelevant. You'll have to wait and get computed average which is what's called Normalized Power. This important variable will put more weight, give more importance to intense work, thus raising the average so that it can precisely quantify what this training dose / test actually was.

You basically get a value that represents what this workout might have been if performed on a flat terrain (somehow).

In the mean time, there would be nothing wrong in using this 275w as a tentative FTP baseline. Worst case, later down the road you may need to adjust it. Or you may start with a value that is little underestimated (that is what I prefer, for what it's worth).

When you first install Wko+, you'll quickly be asked about your FTP, since it is probably the most important parameter for the good functioning of the program. What about plugging in 270 and see if it makes sense later after 2 weeks of workouts.

But again, I know many won't agree with me on that. So take it as you want. But as a triathlon specialist, to me, there's just one very valid FTP test and it is a flat 60min on a course that allows for riding this duration with no interruption. I do them on a FormulaOne race track (I am very lucky). You live near New-York, friends of mine love testing on Mount Whiteface. There you have a 60min climb with no cars (from 6pm or 6:30 pm).

Or even better mate, just check in to an event of secondary importance, race with your powertap. Perform the Tri as if there was no running leg, and take it easy during the 10k. That will give your your FTP right there, with the motivational edge in bonus.

If terrain is rolling, then as suggested consider a longer test than 20-min and look at Normalised Power.

You can calculate it without WKO+, just need to export the file from poweragent to a spreadsheet and run the calcs on it there. It's not a difficult calculation in a spreadsheet. But it's a pain if you wanted to do it for every ride (which for other reasons is a very useful thing to be able to do).

WKO+ of course does it automatically and has many other features making it an excelent addition to your "train with power" suite of tools.

If over time, with improved knowledge and experience of the data and your performances, you realise your original FTP estimate wasn't correct, well you can always adjust it post-hoc.

Alex Simmons said:

If terrain is rolling, then as suggested consider a longer test than 20-min and look at Normalised Power.

sounds like you're implicitly (and very kindly) suggesting not to use NP for a plain 20min test?

Would 20min MMP be the border line? 20min and less should use AP and longer should use NP?

SolarEnergy said:

sounds like you're implicitly (and very kindly) suggesting not to use NP for a plain 20min test?

Would 20min MMP be the border line? 20min and less should use AP and longer should use NP?

You might want to read this:
Alex's Cycle Blog: The seven deadly sins
and this:
Alex's Cycle Blog: The Sins of Sins (Testing FTP #2)

So NP should be considered for rides >= 60min long?

Good for 60, but not for 45 for instance? hmmmm I wonder why?

I can't speak for Alex, but I'd say NP is good for 45 but where you start to get into trouble is NP for 20 then layer that with something like NP_20*0.95 as a way to estimate FTP.

Basically that takes a swag at the shape of someone's Monod curve (.95, .92, .97 take your guess but it varies between riders and their AWC to CP ratios) and then uses the NP algorithm which attempts to model overall metabolic stress, not specific intensity at shorter durations and applies it at the short end of its useful range.

Get far enough out onto the flatter portion of the MMP or Monod curve and using NP makes a lot more sense as AWC contributions are a lot less, the curve is much less steep and you're effectively using a tool (NP) that models metabolic stress to estimate an intensity that's almost entirely metabolic.

-Dave

frost said:

You might want to read this:
Alex's Cycle Blog: The seven deadly sins
and this:
Alex's Cycle Blog: The Sins of Sins (Testing FTP #2)

read as much as possible of it.

My only question would be. Why do they call this testing protocol "deadly sins"? I just don't get the pun.

Anyway. Thanks.

SolarEnergy said:

read as much as possible of it.

My only question would be. Why do they call this testing protocol "deadly sins"? I just don't get the pun.

Anyway. Thanks.

Because the original post on the topic by Dr Coggan identified 7 methods for determining FTP, and presumably "the SDS" was a catchy title. It's stuck ever since.

SolarEnergy said:

sounds like you're implicitly (and very kindly) suggesting not to use NP for a plain 20min test?

Would 20min MMP be the border line? 20min and less should use AP and longer should use NP?

There's not a clear cut off point on a reliable duration for NP as a useful input to estimate FTP. Depends to an extent on an individual's physiological abilities as well as the nature of riding they do.

For me it's probably 30 to 40-min and up is OK for the purpose. For others, shorter durations are OK. But for hard ~1-hour highly variable efforts, NP will be closer to FTP than AP.

Some data for just me:

Pre accident I've had 20-min NP of 373W & 371W on separate occasions and there is absolutely no way I had an FTP north of 350W !!

Post accident I've had 20-min NP of 312W twice and again there's no way my FTP is > 295W.

In my case those 20-min NP numbers represent 118% and 116% of FTP at the time, respectively. Or flipping it over - my FTP was/is 84% and 87% of 20-min NP, respectively.

Comparing to mean maximal average power for me:

20-min MMAP was/is 104% and 107% of FTP respectively.
Or FTP was/is 96% and 93% of 20-min MMAP respectively.

Also, based on data for myself and clients, ratio of FTP to 20-min MMAP:
Average:_____92.7%
Std Dev:_____ 2.3%
Min:________ 88%
Max:________ 96%
Avg +/- 1SD: 90% - 95%

The data ain't perfect but it's a pretty reasonable indicator.

Alex Simmons said:

Because the original post on the topic by Dr Coggan identified 7 methods for determining FTP, and presumably "the SDS" was a catchy title. It's stuck ever since.

Thanks Alex for having taken the time for this clarification. Really appreciated.

Alex Simmons said:

There's not a clear cut off point on a reliable duration for NP as a useful input to estimate FTP. Depends to an extent on an individual's physiological abilities as well as the nature of riding they do.

So at the end, all that is being questioned here (like Dave said earlier in the thread) isn't really the NP algorithm as much as its use for estimating FTP based on a short test (20min long in this case).

- - - -

Reason why that got me a bit confused I guess is this: At the end, PCM data is built upon NP data, not AP?

If 20min long is so borderline as a duration for assessing FTP, then why is this duration included in the SDS list at the first place? Why not suggest the 30min long effort for instance?

(anyway, I don't intend to reinvent this complex wheel in any way, just wondering...)

Quoted post said:

...If 20min long is so borderline as a duration for assessing FTP, then why is this duration included in the SDS list at the first place?...

Well, a fixed percentage of 20 minutes isn't actually mentioned in Andy's original seven deadly sins post. He mentions Monod which often uses a 20 minute data point or something in that vicinity and he mentions tracking long repeatable intervals and 20 minutes is a typical duration for those but then he doesn't suggest scaling it by a fixed fraction.

Here's a link showing Andy's original suggestions and some good comments by wattage list regulars: How to determine your Functional Threshold Power | Two Wheel Blogs

IIRC, Alex was pretty clear that he only included the 20 minute * 0.95 method because it is so commonly used and referenced by folks but even then AP and NP aren't in general interchangeable until the durations get longer.

-Dave

Slowly getting there now.

Based on your experience (both of you ideally).

Do you have difficulty to sell to your clients the idea of a formal 60min test as a means of assessing FTP? Should FTP include any *pacing* ability (or lack of it)?

I can't understand why, a triathlete for example, would favor any other protocol over this one. Not that difficult to book a 40k alone, when you're used to perform it after a swim and prior a 10krun?

I might do one full out 40K effort per year outside of actual time trials and don't generally advocate this method for FTP estimation. IME, it usually leads to lowball FTP estimates since sufficient rest, motivation and yes pacing are important to nail a full FTP effort sustained for 40K. Not to mention it can be difficult at best for many cyclists to find 25 miles of road free of traffic interruptions or too many short steep hills that make pacing tough.

No doubt a full 40K at sustained L4 is great training at the appropriate place in the schedule but it's mentally tough to stay focused that long outside of competition and when I've done those (usually in the weeks preceeding an actual 40K TT) I typically end up at 90-95% of what I actually manage in a race. Great training and good pacing practice sure, but IMO not a great way to estimate FTP and not something most folks have the venue or motivation for very often.

I've used MAP tests, Monods and often take a quick look at power histogram bins (one method of ride file inspection) for the previous week or two if I think my FTP has changed but in practice like SDS #6 which is simply to keep track of long (20-30 minute) intervals done in training and take the regular and repeatable power numbers directly as FTP.

Andy offers it as the second best way to estimate FTP coming in second only to an actual 40K TT under ideal conditions which generally means a bit of pre event rest, a lot of race day motivation and of course good pacing. I've been using SDS#6 for the past couple of seasons to estimate FTP between other tests or races and when I've tested it tends to agree very well. But the key is the regular and repeatable part, I don't take my best ever 20 minute effort and call that FTP but if I do two or three in a session and I hit them hard I'll often look at the final interval(s) of a set.

Anyway, there are a lot of ways, but for someone coaching a variety of athletes with differing levels of motivation, available time, available long unbroken training roads and differing goals I'd probably adopt something other than a full 40km TT to establish and track FTP. In that situation I'd think strongly about supervised MAP or Monod testing to keep it quick short with the coach on hand to assure motivation and focus and to observe things like body language, onset of labored breathing, etc.

I'd also tune into the training data clues that an FTP retest is in order. For instance I usually know when I'm due to reset my FTP when TSS just racks up too easily or IOW my daily ride IF is higher than normal but I'm not intentionally killing myself.

-Dave

Very enjoyable to read Dave, thanks again for your time.

I might have another little one (if you're not feed up yet;-)

SDS#4 and #7.... wait a minute, I think I get it. #4 would be data recorded during a hard *race* that's not a TT, hence the recommendation to use NP instead of AP. This one is already answered.

I'm cool thanks !

SolarEnergy said:

So at the end, all that is being questioned here (like Dave said earlier in the thread) isn't really the NP algorithm as much as its use for estimating FTP based on a short test (20min long in this case).

Yes, I think that's the case. Discussion on the validity, usefulness, efficacy and/or misuse or misinterpretation of NP is a whole 'nuther ball game.

SolarEnergy said:

Reason why that got me a bit confused I guess is this: At the end, PCM data is built upon NP data, not AP?

Yes, NP.

SolarEnergy said:

If 20min long is so borderline as a duration for assessing FTP, then why is this duration included in the SDS list at the first place? Why not suggest the 30min long effort for instance?

daveryanwyoming said:

Well, a fixed percentage of 20 minutes isn't actually mentioned in Andy's original seven deadly sins post. He mentions Monod which often uses a 20 minute data point or something in that vicinity and he mentions tracking long repeatable intervals and 20 minutes is a typical duration for those but then he doesn't suggest scaling it by a fixed fraction.

Here's a link showing Andy's original suggestions and some good comments by wattage list regulars: How to determine your Functional Threshold Power | Two Wheel Blogs

IIRC, Alex was pretty clear that he only included the 20 minute * 0.95 method because it is so commonly used and referenced by folks but even then AP and NP aren't in general interchangeable until the durations get longer.

Yep, Dave is correct - the 95% of 20-min test is not one of the original SDS, and Andy Coggan quite deliberately didn't include it in his list.

The method was "popularised" by its publication in the book by Allen & Coggan. The protocol for performing the 20-min test (i.e. following a 5-min blowout effort) and taking 95% of that for your FTP estimate is Hunter Allen's suggestion.

For those that use their 20-min mean maximal power (either from a stand alone test or plucked from their training/racing data) as a marker and then want to extrapolate to an FTP estimate, my experience with myself and those I have/do coach is that the value is typically falls within a range as shown below.

Ratio of FTP to 20-min mean maximal average power:

Average:_____92.7%
Std Dev:_____ 2.3%
Min:________ 88%
Max:________ 96%
Avg +/- 1SD: 90% - 95%

and that ratio is not static - e.g. you might have an FTP to 20-min MMP ratio of 90% today but increase that at 93% after a block of training.

Another coach, John Verhuel also reports the typical range as 90-93% with the extremes being 88% and 96%.

Guys, I have been little uncomfortable with the SDS concept, now I understand why. I have one question or two.

The SDS protocol seems to advocate defining FTP (in wko+) not as being the best 60MMP we can do, but rather the best 60MMP we *would* do under ideal conditions (decent TBS, motivation etc).

If I use myself as an example. Right now in wko+, my FTP is set to 255. This, I know I can deliver it for sure, and I know I can most probably do better than this (by at least 10w) under ideal conditions after short taper etc... But 255, I can deliver. Just give me a day off followed by an easy day, and on the third one I can book a full 60min 255w. Will feel hard, heavy breathing all the way etc, but not scaring enough not to attempt and finish it.

So if I use SDS protocol, #7 put aside, I guess I could gather clues that my FTP *would* be 265-270, maybe even little higher. But that FTP would only be confirmed under these ideal condition. IOW, I can't guarantee being able to do it.

I guess I thought it was better or safer to define all these training zones based on FTP we could generate while training.

Should I conclude that it is important to set our FTP as some sort of a Target, that is higher than what we could do in training?

What impact would it have to define it little too high?
What impact does it have to define it little too low?