W' Expended greater than 100%

[Salvatore Iovene]

Hi,
on Saturday I was on a very hard group ride, with people who are way
too fast for me. I got dropped after 46 minutes with an average power
of 224 W and an xPower of 240 W (rolling hills).

My CP and W' settings are 232 W and 18 kJ respectively.

Now, my Model data (GoldenCheetah 3.1) says that W' is 16.5 kJ and my
CP is 231 W.

However, my W' Expended (%) is 126, and in red.

Do you suggest I should update my W'?

Thanks!
Salvatore

[Mark Liversedge]

Have you performed a test to exhaustion at 2-20 mins ?

The model will only reflect the workouts you have performed and if they do not contain any exhaustive efforts you may have W' and/or CP underestimated.

FWIW, an exhaustive effort can include sprinting/crunching to the top of a climb and getting there just before you stop exhausted with you heart pounding and your legs weak and trembling.

I have a few of those, but typically only in the 1-2 min range ;)

Mark

[Salvatore Iovene]

Thanks, Mark.

No, I've not performed a test to exhaustion in the 2-20 minute range.

Will the model adjust automatically if I do one?

Thanks,

Salvatore

[Mark Liversedge]

Yes :)

I know its boring, but test protocols are the only reliable way to get an estimate of things like CP and W'.

Trying to derive them from data is a fools errand when you don't know if the data truly reflects an exaustive effort, amd even if it is recent enough. That's not just me, that's most practising sports scientists !

Having said that Andy F is exploring some cool stuff on identifying TTE efforts and we could combine the W'bal model and the 'intervals' code in 3.2 to model on TTE efforts only .. but thats at least a year away !

Mark

[Chris C.]

I have the same problem, my last MTB race I had a W' of 88 (TSS of 106)  but on  a trainning session where I was doing 2x30' sets of openers I had a W' of 227 (in red and a TSS of 103).

A few days before I did a 20 min FTP test and that had a W' of 174  (TSS of 84)

When I extracted and calculated my FTP/CP and W' using the calculator, I used the values of this last FTP test. Should I have done otherwise? I'm trying to sort out all those false value that I had from my inaccurate Wahoo Kickr power value that seem to be off by 40Watts and to get a sense of where my outside summer training is going.

Thanks
Christian

[Hamish Ferguson]

Is there a way to extend the W' balance to above 39,999J? I have some World Class track cyclists that go well over this figure. 

Hamish

[Mark Liversedge]

Oh snap. There really is a limit. What the hell was I thinking ?

We need to fix that pronto, its a stupid mistake.

You can edit the value in power.zones for now using an editor.

Will see if we can fix it quickly.

Damn.

Mark

[Mark Liversedge]

On Tuesday, 26 August 2014 22:19:06 UTC+1, Hamish Ferguson wrote:

Is there a way to extend the W' balance to above 39,999J? I have some World Class track cyclists that go well over this figure.

I found a way to get round this for now.

If you use kj as units it doesn't hit the stupid 40000 limit.

So enter 60 and you get 60,000

I changed the code to automatically adjust when users entered in those units because it was such a common error.

You won't be able to deal in sub kJ units, but frankly, is anyone /that/ precise ?

Cheers

Mark

[Michael Zehr]

I'm also seeing negative W' balance rides using the new tool, even after excluding rides with auto-pause on.  I've mentioned some of this on wattage, but after reading that the model should predict exhaustion within a kj or two, and seeing -9kj W' balance on some rides, I'm viewing it as more of an anomaly.

My power curve very nicely matches the CP curve over a wide range of times.  (If you say I haven't done any full efforts over 30 minutes... guilty as charged.)

I'm letting GC calculate W'.  But I hand calculated it for a number of points on my power curve. I get values for W' ranging from 14-19kj for various times from 30 seconds out to 5 minutes, so as with CP, seems consistent.

The only other factor that goes into the W' balance calculation is the tau, which parameterizes the rate of replenishment.

First, here's my CP display: (My actuals match predicted curve so closely I have a lot of confidence that it's correct.)

And here is the ride in question (with 30 second smoothing):

The effort ending at 51 minutes was truly "to exhaustion" as it put me into a pretty severe asthma attack.  It's not the first time it's happened on this same ride, going up that slope, about 20 minutes in.  (I'm not sure I've done a short interval to exhaustion.  I've certainly done them to pain, even to agony, but perhaps not to exhaustion.  In part because I usually plan on doing multiples of them.)

What might be relevant are the two sections of reduced power at 45 and 47 minutes into the ride.  If W' should have hit 0 instead of -10kj a few minutes later, it must have recovered much more in those two short sections.  I can find other rides in which I hover in negative W' territory for a number of minutes.

I think I can field test the idea that my tau might be lower than predicted.  Do a 3 minute exhaustive test.  Verify I get W' about 18kj and CP about 185-190.  Do another one five minutes later.  If W' balance recovery is as predicted, I should get a lower W' on the second one.  I'll see if GC shows a negative W' balance during the second one.  If so my assumption is that I recovered more of my W' balance during the five minutes than GC is estimating.

(I am already dreading the second 3-minute exhaustion effort, but it's for science, right?)

If any of the following helps with any physiological explanations for the apparent inconsistency, I'm 50, over weight, didn't start cycling seriously until age 45.  I have always been much better at short intense efforts than endurance efforts.  (Why am I spending so much time training an endurance sport?  Excellent question....)  (Could being over weight help recovery?  My blood plasma volume in theory is higher than that of a lighter athlete, so I might have more of a store of whatever I need to replenish W'.)

Thanks,

Michael

[Mark Liversedge]

Hi Michael,

If you use the Extended Model (its under model in chart settings) is the W' estimate higher ?

With a CP of 190w and a max power of over 1000w I would expect W' to be estimated much higher than 14-19kj 

Mark

[Michael Zehr]

Extended model gives W' of 19.8.

I read somewhere else that micro intervals can throw off W' balance calculations.  The start of this ride has short steep rollers and some tight turns that result in frequent 10-15 second periods of high intensity.

I also tried 3-parameter model.  It gives W' of 31.4 and CP of 171.  That's enough of a difference to make W' balance stay positive for all my rides.

Michael

[Mark Liversedge]

Microintervals have interesting characteristics, I don't really understand all the science behind why, but hopefully there will be more stufff published on this. 

I believe its a combination of enhanced recovery during the first 30 seconds of off with reduced impact/fatigue from intense efforts due to their short duration.

Mark

[Luther Gulseth]

Could his CP also be much higher? When I have gotten dropped from an extra fast group ride, my power levels are much less than a race...try 20min max around 30-40W higher. Somehow, that race thing always makes you dig deep into a very dark place :)

in Christ crucified,
Luther

--

_______________________________________________
Golden-Cheetah-Users mailing list
golden-che...@googlegroups.com
http://groups.google.com/group/golden-cheetah-users?hl=en
---
You received this message because you are subscribed to the Google Groups "golden-cheetah-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to golden-cheetah-u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Andy Coggan]

On Friday, August 29, 2014 4:51:46 AM UTC-5, Michael Zehr wrote:

I read somewhere else that micro intervals can throw off W' balance calculations. 

They can (and every time it happens, a kitten dies). I could explain exactly why, but, well, Mark doesn't seem to think I have a thorough understanding of the physiology of exercise, so why bother? 

[Mark Liversedge]

On Friday, 29 August 2014 20:13:47 UTC+1, Andy Coggan wrote:

On Friday, August 29, 2014 4:51:46 AM UTC-5, Michael Zehr wrote:

I read somewhere else that micro intervals can throw off W' balance calculations.  

 I could explain exactly why

Please do, it would be useful to have something to add to the science section to expand regarding this.

[Mark Liversedge]

I also forgot about this from Dr Skiba, that may help too: http://physfarm.com/new/?page_id=1092

Mark

[Andy Coggan]

Ask yourself this: aside from 4th-order weighting, what do normalized power and xPower have in common?

[Anton Jacobs]

Hi guys, anyone with some help, would be greatly appreciated.

Model says CP 247, W' on Summary are >100% CP per W'Bal says minimum CP 276.

It is a bit confusing or am I looking for too much?

[Mark Liversedge]

The stress chart is suggesting that for the W' you have set CP would need to be 276 to avoid going negative.

It may be that your W' is higher. Have you performed maximal efforts in the 2 and 20 minute range?

Mark

[Anton Jacobs]

Hi Mark,

Firstly thanks.

Okay cool, yes I've performed numerous intervals between those ranges. 

W' Currently on 27,1. CP curve with extended model "now" suggests W' 28.

Will adjust W' accordingly.