TSS for running?

[matthew dunn]

Wondering why for all my rides I have a TSS score, but all my runs are 0's for TSS, Anyway to get a running TSS?

Also any suggestions for videos on how to set up GC? I have all the basic charts and have a vauge idea on how to read them, but I would like yo know how to really set GC up so I can read and understand my information!

ALSO...
Being both a runner a cyclist I was wondering if there was a way to seperate the graphs and charts I would like for running be available when I upload my run workout,and when I upload my bike workout the graphs and charts for cycling.

Especially for my run workouts I have a lot of boxes that say no data available, and I am like of course I am not going to have a power graph when I run!

Any help setting this up would be great!

[Ale Martinez]

El jueves, 22 de octubre de 2015, 23:48:11 (UTC-3), matthew dunn escribió:

Wondering why for all my rides I have a TSS score, but all my runs are 0's for TSS, Anyway to get a running TSS?

Not TSS, but you can use TriScore for the same purpose and you can find a PMC(TriScore) on the chart sidebar in Trends view. Just be sure you have Pace Zones defined for the dates of your running workouts and they have Sport set to "Run".

If your athlete is migrated from a previous version you may need to use View->Reset View both in the Trends and Activities views to upgrade them and to add TriScore/xPace to Summary Metrics, Peak Pace xxx to bests, Pace to intervals, sport filters, etc. Or just create a new athlete and import your workouts from the previous one for this to happen automatically.

 

Also any suggestions for videos on how to set up GC? I have all the basic charts and have a vauge idea on how to read them, but I would like yo know how to really set GC up so I can read and understand my information!

Have you seen the tutorials at the main site? http://www.goldencheetah.org/#section-tutorials 

The Wiki is also a great resource: https://github.com/GoldenCheetah/GoldenCheetah/wiki/UG_Main-Page_Users-Guide

ALSO...
Being both a runner a cyclist I was wondering if there was a way to seperate the graphs and charts I would like for running be available when I upload my run workout,and when I upload my bike workout the graphs and charts for cycling.

Especially for my run workouts I have a lot of boxes that say no data available, and I am like of course I am not going to have a power graph when I run!

Any help setting this up would be great!

Try using a new athlete and see if that's is a better fit for your needs, some customization to remove unwanted graphics to add new ones may also help, some charts adapt itself according to the sport and all of them can be filtered by sport in Trends but, at the moment, there is no way to have separate charts for running and cycling.

[Ron George]

Ale,

Just reviving this thread for not wanting to replicate elsewhere.

Devices like Stryd have started implemening a RSS scoring system for running based on power. Not sure of the mathematical expression but given Coggan is consultant, it's probably the same implementation as TSS. 

TriScore is currently using GOVSS as a running score. Since I do many treadmill runs, I don't know see utility of this metric.

Primary scoring I would prefer would be based on running power.

My running workouts with power are missing the TSS metric, probably because TSS is recognized only for cycling in GC?  How can I get GC to automatically give me a power based running metric with which to display a stress PMC? Right now I'm manually editing every ride to add in a TSS value, which is tiring. Thanks!

Ron

[Ale Martinez]

Hi Ron, if you want TSS for running workouts with power -even when it is defined only for cycling- you can define a Custom Metric using the formula 100 * (Duration/3600) * IF^2 as value.

OTOH if you want the new Stryd's RSS the formula would be slightly different.

[Steve Mansfield]

As an alternative, you could use TRIMP score - just need to make sure HR zones are defined.

[Ron George]

Hi all,

TRIMP is based on HR. So not preferred since I got my running power measurement now.

RSS mathematical implementation is different from TSS. 

The essay on RSS is now provided on Stryd blog. http://blog.stryd.com/2017/01/28/running-stress-score/

They state : 

RSS = 100 x training duration x (Power/CP)^K

"The coefficient K accounts for the fact that different intensity levels spur different physiological adaptations. In order to achieve your best potential, it is important to have the right training variety across all intensity levels, while allowing the proper recovery time for each.

Note that, RSS shares a similar form as the Training Stress Score (TSS) used in cycling. Yet, cycling and running are fundamentally different sports. In particular, run training is often constrained by mechanical stress. Increasing running intensity results in higher stress on our body than comparable increases in cycling. Indeed, the coefficient K for running RSS is significantly higher than that of cycling TSS, reflecting the higher stressing effect of intensive run workouts."

It is not known what the "k" co-efficient is.

From what they state, TSS for running and RSS for running will be different, how different I don't know. I'm ok living with a TSS for running mainly to avoid splitting hair over small differences, and listen to body signals for placing added recovery after intensive runs. Just want the PMC chart to look reasonably meaningful. 

[Mark Liversedge]

On Wednesday, 1 February 2017 08:14:36 UTC, Ron George wrote:

The essay on RSS is now provided on Stryd blog. http://blog.stryd.com/2017/01/28/running-stress-score/

"Critical power (CP): your maximal average sustainable power over one hour of running, also referred to as “lactate threshold” power; this is used as your performance baseline"

I despair. When will this shit ever stop?

Mark 

[Ale Martinez]

On Wednesday, 1 February 2017 05:14:36 UTC-3, Ron George wrote:

Hi all,

TRIMP is based on HR. So not preferred since I got my running power measurement now.

RSS mathematical implementation is different from TSS. 

The essay on RSS is now provided on Stryd blog. http://blog.stryd.com/2017/01/28/running-stress-score/

They state : 

RSS = 100 x training duration x (Power/CP)^K

"The coefficient K accounts for the fact that different intensity levels spur different physiological adaptations. In order to achieve your best potential, it is important to have the right training variety across all intensity levels, while allowing the proper recovery time for each.

Note that, RSS shares a similar form as the Training Stress Score (TSS) used in cycling. Yet, cycling and running are fundamentally different sports. In particular, run training is often constrained by mechanical stress. Increasing running intensity results in higher stress on our body than comparable increases in cycling. Indeed, the coefficient K for running RSS is significantly higher than that of cycling TSS, reflecting the higher stressing effect of intensive run workouts."

It is not known what the "k" co-efficient is.

If you are a Stryd's customer may be you can ask them, but looking at RSS/min vs Power/CP table there seems not to be a single K that fit that equation to those data...

An exponential expression seems to be a better fit:

From what they state, TSS for running and RSS for running will be different, how different I don't know. I'm ok living with a TSS for running mainly to avoid splitting hair over small differences, and listen to body signals for placing added recovery after intensive runs. Just want the PMC chart to look reasonably meaningful. 

 

As I commented in my first post you can build a custom metric to use TSS, RSS or whatever you want as input for your PMC, here you can find a how-to: http://forum.slowtwitch.com/forum/?post=6183696#p6183696 

[Ron George]

Thanks Ale, 

Thought k would be constant. Apparently not. Too much complexity to implement as a custom metric. Think I'll be fine with TSS.

-Ron

[Ale Martinez]

Hi Ron, if you are Ok with the exponential fit the custom metric implementation would be very similar to the original formula with fixed K

[m.renne...@gmail.com]

How would the formula for such a running TSS custom metric actually look like ?
I mean how would the syntax be that I would have to provide, because obviously I don't want to calculate aerobic points as shown in the how-to you posted ;-)

Every help would be very much appreciated.

[Ale Martinez]

El viernes, 21 de abril de 2017, 7:40:24 (UTC-3), m.renne...@gmail.com escribió:

How would the formula for such a running TSS custom metric actually look like ?
I mean how would the syntax be that I would have to provide, because obviously I don't want to calculate aerobic points as shown in the how-to you posted ;-)

If you use the exponential fit to model the RSS/min to relative intensity curve and IF as a proxy for relative intensity, the value formula would be like this: value { (Duration/60) * 0.075808 * exp(3.129738*IF); }

[m.renne...@gmail.com]

Thanks a lot Ale !

I truly agree with you, that your exponential fit seems at least as good as the Stryd Function with the K-coefficient.
Eventhough I perfectly understand the mathematics you did, my C++-skills would never have been sufficient to make up a custom metric from your fit. Is this even C++ ? ;-) and what does the semicolon do ?

All the best
Michaela

P.S.: You don't really have to answer these last two questions ;-) - I'm perfectly fine now !

[Frederic Bard]

Reviving this thread as I discovered this feature in the GoldenCheetah

When I look at GC's source code, I found the implementation of the GOVSS algorithm from Dr Skiba and was wondering where the spreadsheet you mention here

can be found so I can look at the numbers in it.

Thanks!

Frederic

[Ale Martinez - No direct email please]

That spreadsheet is not available online anymore, AFAIK.

[Frederic Bard]

Do you happen to have a local copy by any chance that you could share ?

Thanks

[Ale Martinez - No direct email please]

No, I can't.

[Frederic Bard]

Ok, i understand if you can't share it.

But could you explain the rationale behind your code ?

You multiply Ckin and Caero and Cslope by the weight but I don't see that anywhere in the litterature.

return (cAero + cKin + cSlope*eff) * speed * weight;

Actually, I only see Cslope being multiplied by the user's weight, am I misunderstanding something ?

Example,here 

P=(C​aero​​+C​slope​​⋅η⋅m+C​kin​​)⋅v and mm is the runner weight

Another example here :

power=(cAero + cKin + cSlope*eff*SUUNTO_USER_WEIGHT)*speed;

Additionally, in Skiba's pdf

I don't see the weight being multiplied to CAero and Ckin

Power(W/kg)=(Ci!nV-(Ci!nV!(.5!(V!8.33-1))))+Caero!V+Ckin!V(Eq.7)

Thanks

[Ale Martinez - No direct email please]

El jueves, 17 de octubre de 2019, 20:28:42 (UTC-3), Frederic Bard escribió:

Ok, i understand if you can't share it.

But could you explain the rationale behind your code ?

Sure

 

You multiply Ckin and Caero and Cslope by the weight but I don't see that anywhere in the litterature.

return (cAero + cKin + cSlope*eff) * speed * weight;

Actually, I only see Cslope being multiplied by the user's weight, am I misunderstanding something ?

Example,here 

P=(C​aero​​+C​slope​​⋅η⋅m+C​kin​​)⋅v and mm is the runner weight

 

Another example here :

power=(cAero + cKin + cSlope*eff*SUUNTO_USER_WEIGHT)*speed;

Additionally, in Skiba's pdf

I don't see the weight being multiplied to CAero and Ckin

Power(W/kg)=(Ci!nV-(Ci!nV!(.5!(V!8.33-1))))+Caero!V+Ckin!V(Eq.7)

Because this formula is computing relative power (W/kg) and our running_power function is computing absolute power in W = (W/kg) * kg

[Frederic Bard]

Thank you