Heart Rate vs Power Trace

[RecceDG]

Here's another one - a trace of the ratio of heart rate to power, so one can see if it is constant or changing. Per Joe Friel's book.

DG

[old_man_biking]

Here's another one - a trace of the ratio of heart rate to power, so one can see if it is constant or changing. Per Joe Friel's book.

DG

Joe Friel's training bible? It's been a while that I read that for the last time.
Could you provide a few more words of explanation.

[RecceDG]

4th Edition, pages 54-56.

He talks about heart rate "coupling" as a function of "cardiac drift"

In a nutshell, if during the course of an effort, heart rate vs power remains constant within 5% (eg, a given power effort produces a given heart rate), then the workout it "coupled" and it is representative of a good level of aerobic fitness over that period.

If, however, heart rate varies with a given power output, (where it is expected that heart rate will increase with time at a given power output) then the opposite is true.

Joe recommends, during Base period, a bi-weekly aerobic threshold effort starting at 20 min and getting increasingly longer. When the effort is two hours long and heart rate remains coupled to power throughout the effort, you know your base has been properly established.

WKO+ provides a power-to-heart-rate shift statistic where it divides the ride into two, calculates avg power vs avg heart rate for both halves, then divides the first half by the second have to produce the "shift ratio". Joe then shows visually comparing an average heart rate trace vs an average power trace to see if the traces remain parallel over the required range.

But rather than looking for parallelism visually, it would be better to provide a trace of avg power vs avg heart rate over time (or distance) so that the change in the ratio can be explicitly seen. Line stays flat? Coupled. Line angles up or down? Decoupled. And by having a trace, you can see exactly where in the workout the decoupling occurred.

I tried a straight up X-Y scatterplot of power vs heart rate, and it doesn't provide the same information.

Is that a good enough explanation?

DG

[old_man_biking]

Not bad

One question remains:
WKO simply calculates averages of power andhr for 1st and 2nd half.
When calculating avg avg hr/avg power over time, what would be a reasonable averaging time span?

[RecceDG]

That's an excellent question.

And the other aspect is the delayed nature of heart rate - important on intervals, less so on sustained efforts.

Perhaps make the length of the averaging time window be user-selectable?

DG

[old_man_biking]

Ha ha, that's what I've expected:
"If you don't have a clue, make it configurable."
For one plugin (weather) I decided to follow this approach and ended up with a settings page brimming with check boxes.

The analogous approach in politics:
"If you don't have a clue, set up a subcommittee."

Btw, the Training Analysis already uses a "time shifted" heart rate for correlations, to model the delayed response of the cardiovuscular system to applied load.
Appropriate time shift is auto-calculated as a best fit by using least squares sum approximation.

Cheers,
OMB

[RecceDG]

The other side of that though is "if there is debate on a feature, hardcoding it to X ensures that everybody will want Y"

I saw that delay feature elsewhere in the plugin, and it is pretty cool how well it seems to work - which is why I think it might be an option here. Although I wonder if it means that if the heart rate is shifted left in time, if that erroneously shifts the point in time when the heart rate decouples from power (assuming it does) too early.

Best make that an option too.

DG

[stavros]

FYI - You can get the basic decoupling ratio from the Calculated Fields plugin. Doesn't give you the trace, but still provides useful insight over time.

There's a standard function for the Speed:HR decoupling ratio for active splits - DECOUPLINGRATIOACTIVE. See screen shot below for setup. This works well for running, but might give a slightly different result to a Power:HR decoupling ratio for cycling.

If you're inclined, you could create a new Power version. Or request it be added as a standard function? Something like:
Power:HR Ratio = ({FirstHalfActivePowerHRRatio}-{SecondHalfActivePowerHRRatio})/{FirstHalfActivePowerHRRatio}*100
using Virtual expressions for
FirstHalfActivePowerHRRatio = {ACTIVERANGEPower(Distance,0,{HALFDISTANCE})}/{ACTIVERANGEHR(Distance,0,{HALFDISTANCE})}
and
SecondHalfActivePowerHRRatio = {ACTIVERANGEPower(Distance,{HALFDISTANCE},{DISTANCE})}/{ACTIVERANGEHR(Distance,{HALFDISTANCE},{DISTANCE})}

You can then include the custom field in your activity report layout. This can give interesting insights on relative fitness over time, if not actually within an activity.

[hill]

any news or motivation to intigrate this into training analysis? would be great!!!

[old_man_biking]

Post a link to a thorough explanation of decoupling and I'll see.

[hill]

ich denke, hier ist das alles erklärt: http://home.trainingpeaks.com/articles/cycling/aerobic-endurance-and-decoupling,-by-joe-friel.aspx

meinen wunsch aus dem deutschen forumsbereich könnte ich noch um eins erweitern, nämlich damit, dass man evtl. mehr als eine de verteilungskurven auf einmal anzeigen kann. z.b. cadence und power nebeneinander, dafür den track ausblenden oder so...
das ist bei wko+ wirklich sehr übersichtlich und sicherlich in sporttracks integrierbar!

[RecceDG]

Joe describes it again in detail with his new power meter book.

http://www.amazon.com/Power-Meter-Handb ... ower+meter

It would be REALLY very good if either the Training Analysis plugin or SportTracks natively was capable of doing the analysis described in this book.

DG

[the5krunner]

FWIW I think this would be useful as well.