FAQs for Running and Swimming

[Ale Martinez]

I was thinking about to add sections for FAQs specific for Running and Swimming to the wiki, I have some ideas about what questions I would like to answer but, probably, it would be more useful to know what questions users would like to ask, if you have some questions to contribute please post them here.

Thanks

[Fa Bio]

How do i set swimming pool lenght?

[Fa Bio]

How do i see critical velocity for swim and run in trends-tracker?

[Ale Martinez]

El miércoles, 21 de diciembre de 2016, 6:05:59 (UTC-3), Fa Bio escribió:

How do i set swimming pool lenght?

In most cases it should be automatically set when you import lap swimming files, you can see/edit it using "Pool Length" (meta)data field which should be automatically defined and showed on Details>Workout tab for new users-

You can add it for pre-existing users, see: https://github.com/GoldenCheetah/GoldenCheetah/wiki/FAQ-DATA#how-can-i-addremove-a-field-to-edit-on-the-details-screen

The tool Edit > Fix Lap Swim from Length Data also allows to change Pool Length and modify pace and distance accordingly.

[Ale Martinez]

El miércoles, 28 de diciembre de 2016, 3:39:18 (UTC-3), Fa Bio escribió:

How do i see critical velocity for swim and run in trends-tracker?

Adding a formula config(scv) or config(cv) will show them in kmh, if you want pace units they can be converted in the same formula, s.t. 6/config(scv) for swim CV in min/100m and 60/config(cv) for run CV in min/km, for more deatils on formulas see https://github.com/GoldenCheetah/GoldenCheetah/wiki/UG_Special-Topics_Formula-Syntax-and-Expressions

[Fa Bio]

Thanks Ale

[tigg...@gmail.com]

Hello Ale,

Is there a way I can change the x axis on the CV chart from time to distance?

Thanks

[Ale Martinez]

El jueves, 29 de diciembre de 2016, 16:20:31 (UTC-3), tigg...@gmail.com escribió:

Is there a way I can change the x axis on the CV chart from time to distance?

 

No, it is just a Speed/Duration curve, but you can see the corresponding distance at the tooltip.

OTOH, It should be relatively simple to draw a Speed/Distance or Pace/Distance curve on "R".

 

[tigg...@gmail.com]

Thanks but simple and "R" don't go hand in hand as far as I'm concerned!

A pace / distance curve would be great though - I have it on Excel where I also plot a line of best fit (log trend line) to see what pace per 100m I can do for different distances and where I am above/below the trend line. That can then let me see what times are "soft" and I can target them when I want a harder effort.

[Ale Martinez]

El jueves, 29 de diciembre de 2016, 16:42:17 (UTC-3), tigg...@gmail.com escribió:

A pace / distance curve would be great though - I have it on Excel where I also plot a line of best fit (log trend line) to see what pace per 100m I can do for different distances and where I am above/below the trend line. That can then let me see what times are "soft" and I can target them when I want a harder effort.

I've been there, I will give it a try in R.

[Ale Martinez]

El jueves, 29 de diciembre de 2016, 16:42:17 (UTC-3), tigg...@gmail.com escribió:

Thanks but simple and "R" don't go hand in hand as far as I'm concerned!

May be, but It is way simpler than to modify the GC CV chart ;-)

A pace / distance curve would be great though - I have it on Excel where I also plot a line of best fit (log trend line) to see what pace per 100m I can do for different distances and where I am above/below the trend line. That can then let me see what times are "soft" and I can target them when I want a harder effort.

Here is my first attempt:

 

## R script will run on selection.

##

## plot meanmax

GC.page(width=800,height=600)

par(mar=c(6,6,6,6))

## get swim data for the selected date range

mm <- GC.season.meanmax(all=FALSE, filter="isSwim<>0", compare=FALSE)

# mean maximal speed from 1s to 1h

speed <- mm$speed[2:3602]/3.6

# pace from speed in min/100m

pace <- 6/mm$speed[2:3602]

# distance from mean maximal speed

distance <- speed*(1:length(pace))

# plot best pace for distance

plot(distance, pace, pch=16, col="red",

xlab="Distance (m)",

ylab="Pace (min/100m)")

# fit an exponential model

model <- lm(log(pace) ~ log(distance))

# plot model line

lines(distance, exp(predict(model, list(distance))),

lwd=2, col="red")

 

[Ale Martinez]

Reversed y-axis plus a legend to show the pace reduction when distance doubles based on the fitted exponent:

Drag and drop of the attached file on GC should recreate this chart on Trends.

[tigg...@gmail.com]

Brilliant!

That worked well.  Just dragged it in and it works first time.

Thank you

[Ale Martinez]

Similar chart for Running:

[tigg...@gmail.com]

Thanks for that.  The only comment that I have is that it does not show pace slower than 5 mins per km - but that's easy to change, I just have to run faster!

[Nigel Van de Velde]

I have some questions on running analysis in GoldenCheetah

1. On most of my runs xPace and average pace seem to be very similar. 

This morning for example I did a run of 19km on flat terrain. I did 4 intervals of 2km at 4:30 min/km and the rest was around 5:30 min/km. Average pace was 5:06 min/km, xPace was 5:04 min/km. 

If I would do a similar training on the bike, I believe xPower and average power would be very different. xPower would be a good indication of how hard the workout was, which doesn't seem to be the case for xPace. 

I thought the concept of xPace was similar to xPower, right? xPace would take pace changes and altitude changes into account and simulate the pace of an equally hard effort on flat terrain at a steady pace, correct? Or does it only take altitude changes into account? In that case it would explain the small differences between xPace and average pace, as all my runs are on flat terrain. Are there other pace based metrics that can indicate the toughness of a workout?

2. I have a CV speed duration chart to track my progress, but it seems hard to select a suitable model:

- 2 parameter model: it seems only one point is being used to fit the data to a model. CV seems a little high when looking at the according pace zones

- extended model: the model doesn't fit the data very well. I always have a "bump" at around 3-4 minutes and at durations above 20 minutes the model is a lot higher than actual data. I have good data between 2 and 40 minutes. I don't think many people test to exhaustion for longer duration runs. CV again seems rather fast to base my pace zones on. Possibly the model doesn't fit my data very well because the "long aerobic search interval" can't be set lower than 3000s. I don't have good efforts in that range.

- 3 parameter model (with changed search intervals; anaerobic 120-200 and aerobic 420-1200). Same remark as for the 2 parameter model. Idem dito for the Ward-Smith model

- The Multicomponent model: model doesn't fit data at all, too high at low durations, too low at high durations.

Should I even use these models, or are they only "validated" for bike power data? Maybe it would be better to select no model at all and possible add the best TPace value in the overlay. TPace seems to be a good estimation of my CV. When I base my pace zones on TPace, I can complete my intervals as intended.

For swimming these CV models are even more troublesome. I guess most people do several sets during swimming with resting periods in between. So the models on the CV chart don't really give useful information concerning CV. It would be nice if there would be some sort of "TPace" metric for swimming as well.

3. It seems GoldenCheetah does some estimations of running power based on pace as well. On the activity summary I see "LNP" and "RTP" for each run. I guess they are similar to NP and CP for bike power. 

It seems to me that LNP seems high in regard to RTP. Generally I see LNP values of over 250W for easy endurance runs, while my RTP is generally around 310-330W. If I compare this to my bike activities containing power data, then rides with an NP of 50W below my CP are definitely not easy endurance rides. How are these values of LNP and RTP calculated on the summary page? 

I tried using the "estimate running power" tool, but that gives values that are way too high. The CP curve that I get doesn't go lower than e.g. 380W for an activity of 1h30, while RTP is shown as only 330W.

I guess these LNP and RTP values seem great to track toughness of activities and progress over time, but I'm not sure if I can trust these data.

Thanks for setting up a wiki on running and swimming analysis. I hope it will help to improve GoldenCheetah even further!

Kind regards,

Nigel

  

[Marc Meyer]

I am not sure whether this is relevant to your post. but as I just asked this question in the group now...

Is it possible to create or identify intervals with a workout by GC detecting the LAP press?

In other words for running , when doing intervals with recovery jog in middle, I would like to be able to identify when the 300M hard effort starts and when the 200M recovery ends.

Thanks

On Monday, December 19, 2016 at 3:49:04 PM UTC+2, Ale Martinez wrote:

I was thinking about to add sections for FAQs specific for Running and Swimming to the wiki, I have some ideas about what questions I would like to answer but, probably, it would be more useful to know what questions users would like to ask, if you have some questions to contribute please post them here.

Thanks

Please consider the environment before printing this email

[KJA]

Hi. I am very new so GC is a bit blurry.

What data/fields are possible and can be explored further, and what are definite 'no no belongs to cycling' to me as a runner (wihtout power meter)? - CTL, ATL, TSB, RR, xPower, W', CP, Stress, PMC, IF, TRIMP, Triscore, etc (UG_Glossary).. What should be prio for me to begin to learn? (As a runner and maybe a newbie :))

I have imported via tapriik, is it possible to see running dynamics from garmin connect? I can see mean running cadence, what about a chart for the entire workout?

In the beginning i feel kind of clumsy when i handle the charts? Is it possible to more easily zoom in/zoom out? Every time i press the charts im creating selections. Any tips? Also, how to increase the text?

The pace zones are based on CV, is there any easy test you do to easy find it?

[Ale Martinez]

On Wednesday, 1 February 2017 11:02:46 UTC-3, Nigel Van de Velde wrote:

I have some questions on running analysis in GoldenCheetah

Sorry for the late response, I was on Holidays and missed this.

  

1. On most of my runs xPace and average pace seem to be very similar. 

This morning for example I did a run of 19km on flat terrain. I did 4 intervals of 2km at 4:30 min/km and the rest was around 5:30 min/km. Average pace was 5:06 min/km, xPace was 5:04 min/km. 

If I would do a similar training on the bike, I believe xPower and average power would be very different. xPower would be a good indication of how hard the workout was, which doesn't seem to be the case for xPace. 

I thought the concept of xPace was similar to xPower, right? xPace would take pace changes and altitude changes into account and simulate the pace of an equally hard effort on flat terrain at a steady pace, correct? Or does it only take altitude changes into account? In that case it would explain the small differences between xPace and average pace, as all my runs are on flat terrain.

The algorithm accounts for changes in pace even when includes some averaging as explained in the Skiba Paper, OTOH running pace is a lot less variable than cycling power in typical conditions.

 

Are there other pace based metrics that can indicate the toughness of a workout?

Not that I know. The "gold standard" for this is RPE (if it feels hard, it is hard), Session RPE is supported.

2. I have a CV speed duration chart to track my progress, but it seems hard to select a suitable model:

- 2 parameter model: it seems only one point is being used to fit the data to a model. CV seems a little high when looking at the according pace zones

You can play with the search duration but you need at least 2 max effort runs, one in each search interval, for example a 1000-1500 run in the range of 3-5 min and 3000-5000 in the range of 10-15 min.

 

- extended model: the model doesn't fit the data very well. I always have a "bump" at around 3-4 minutes and at durations above 20 minutes the model is a lot higher than actual data. I have good data between 2 and 40 minutes. I don't think many people test to exhaustion for longer duration runs. CV again seems rather fast to base my pace zones on. Possibly the model doesn't fit my data very well because the "long aerobic search interval" can't be set lower than 3000s. I don't have good efforts in that range.

Yes, the extended model has a tendency to overfit for speed data, I found the "bump" can be avoided or, at least, diminished, playing with search durations. For longer durations I found good fits when half-marathon and/or marathon data are included, if your longer max efforts don't go over 3000 secs I would suggest to stay with 2/3 par or WS models.

 

- 3 parameter model (with changed search intervals; anaerobic 120-200 and aerobic 420-1200). Same remark as for the 2 parameter model. Idem dito for the Ward-Smith model

Se above.

 

- The Multicomponent model: model doesn't fit data at all, too high at low durations, too low at high durations.

These model don't work at all for my speed data, I don't use them.

 

Should I even use these models, or are they only "validated" for bike power data? Maybe it would be better to select no model at all and possible add the best TPace value in the overlay. TPace seems to be a good estimation of my CV. When I base my pace zones on TPace, I can complete my intervals as intended.

2/3 par model are used for CV in the literature, not just for CP. In the cloud there is a chart I've uploaded with an exponential fit which you can find useful, it would be relatively easy to change it to use a Daniels curve.

For swimming these CV models are even more troublesome. I guess most people do several sets during swimming with resting periods in between. So the models on the CV chart don't really give useful information concerning CV. It would be nice if there would be some sort of "TPace" metric for swimming as well.

Well, if you want a CV curve you need some continuous (Test) sets even when your training is by intervals... A practice suggested for some coaches for non elite swimmers is 200 and 400 max effort tests to determine CV (also called CSS) which can be "validated" with a 1500-2000 long duration test afterwards. See for example http://www.swimsmooth.com/training.html

3. It seems GoldenCheetah does some estimations of running power based on pace as well. On the activity summary I see "LNP" and "RTP" for each run. I guess they are similar to NP and CP for bike power. 

It seems to me that LNP seems high in regard to RTP. Generally I see LNP values of over 250W for easy endurance runs, while my RTP is generally around 310-330W. If I compare this to my bike activities containing power data, then rides with an NP of 50W below my CP are definitely not easy endurance rides. How are these values of LNP and RTP calculated on the summary page? 

I tried using the "estimate running power" tool, but that gives values that are way too high. The CP curve that I get doesn't go lower than e.g. 380W for an activity of 1h30, while RTP is shown as only 330W.

I guess these LNP and RTP values seem great to track toughness of activities and progress over time, but I'm not sure if I can trust these data.

LNP and RTP are intermediate values to compute xPace and GOVSS, if you want a "power" based approach try some of the new running "power meters" s.t. Stryd or the included Edit>Estimate Running Power tool. 

[Ale Martinez]

On Thursday, 2 February 2017 14:18:52 UTC-3, Marc Meyer wrote:

I am not sure whether this is relevant to your post. but as I just asked this question in the group now...

Is it possible to create or identify intervals with a workout by GC detecting the LAP press?

In other words for running , when doing intervals with recovery jog in middle, I would like to be able to identify when the 300M hard effort starts and when the 200M recovery ends.

Yes, manual lap markers are used to identify intervals on GC. 

[Ale Martinez]

On Thursday, 2 February 2017 16:21:00 UTC-3, KJA wrote:

Hi. I am very new so GC is a bit blurry.

What data/fields are possible and can be explored further, and what are definite 'no no belongs to cycling' to me as a runner (wihtout power meter)? - CTL, ATL, TSB, RR, xPower, W', CP, Stress, PMC, IF, TRIMP, Triscore, etc (UG_Glossary).. What should be prio for me to begin to learn? (As a runner and maybe a newbie :))

For running Pace based metrics include xPace, TriScore, D', CV, and PMC based on TriScore, HR and RPE based metrics are sport independent.

 

I have imported via tapriik, is it possible to see running dynamics from garmin connect? I can see mean running cadence, what about a chart for the entire workout?

You need to import the original .fit file to have running dynamics, I *think* they are not included in the generated TCX files.

 

In the beginning i feel kind of clumsy when i handle the charts? Is it possible to more easily zoom in/zoom out? Every time i press the charts im creating selections. Any tips?

Once you select an area you can use the left sidebar to zoom that area and to delete the selection.

 

Also, how to increase the text?

See Options/Preferences. 

The pace zones are based on CV, is there any easy test you do to easy find it?

You need 2 max efforts runs s.t. 1000-1500m and 3000-5000m to estimate CV, the alternative would be to use Daniels TPace as a proxy based on your best recent running race, see Tools menu.

[KJA]

I'm trying to set up a new chart on Trends. I have average stride lenght, ground contact time etc. I have mostly done running activities and one workout on trainer on bike. It seems that the bike activity interfering with the other, and gave me a really high running stride lenght. Is there any way you can choose what type of activity (running) when Add Chart?

[Ale Martinez]

Yes, you can specify a filter for the whole chart at Basic Settings or for each curve separately at Curve Settings, isRun<>0 allows to select only runs.

[Ale Martinez]

I created a first version, reviews and contributions are welcome: https://github.com/GoldenCheetah/GoldenCheetah/wiki/Running-&-Swimming

[Ale Martinez]

Previous link is not longer valid since I've updated the page to follow the existing FAQ format (FAQ RUNNING & SWIMMING), also added an index to FAQ for easier reference.

[KJA]

Can i do a similar test for Running Critical Power as Critical Velocity whitout using power meter? I'm playring around with "Estimate Running Power" (which i assume is for them who don't have a power meter), but Relative intensity = xPower/Critical power, but CP is set to default on 250W so i must test CP or everything related to the estimated running power get incorrect?

[Ale Martinez]

Yes, you can use Estimate Running Power on your CV test(s) and then use the power valúes as input to estimate CP and W', several but not all power related metrics depend on having on CP value.

[KJA]

- What power values? Average or maximum watts?

Should i use tool: CP and W' Estimator?  -->  In CP Estimator, should i set it on Bike (instead of run, i only get critical pace otherwise)?
And why is it only 60 min? (My CV measurement was on half-marathon as it could according to wiki).



Thanks for your help.

[Ale Martinez]

El sábado, 22 de julio de 2017, 10:56:53 (UTC-3), KJA escribió:

- What power values? Average or maximum watts?

Average 

Should i use tool: CP and W' Estimator?  -->  In CP Estimator, should i set it on Bike (instead of run, i only get critical pace otherwise)?

Yes, the CP model is not specific to cycling

 

And why is it only 60 min?

Due to technical reasons, look for a Critical Power Model explanation if you want to understand this. 

 

(My CV measurement was on half-marathon as it could according to wiki). 

You can use your average power during the half increasing it a little if it is way longer than an hour, for example for a 1h30' HM something like 2.5% higher.

[KJA]

I did a CP test now (3-9 according to https://www.trainingpeaks.com/blog/running-with-power-how-to-find-your-run-ftp/).
I got 289 watts on 3 minutes and 233 watts on 9 minutes.

Should i use the average of the two and then 90% (as in the article) (CP = 235 watts) or Golden cheetah estimation (CP = 205 watts?) What do you think?. I think its a big diff so a iam unsure :)

I see that the long rung should at least be 15 min according to GC but surely that are cycling and 9 minutes should work anyway for run?

Also.. on chart "Ride" i can't find Pace (min/km), only speed?

[Ale Martinez]

El viernes, 28 de julio de 2017, 6:09:05 (UTC-3), KJA escribió:

I did a CP test now (3-9 according to https://www.trainingpeaks.com/blog/running-with-power-how-to-find-your-run-ftp/).
I got 289 watts on 3 minutes and 233 watts on 9 minutes.

Should i use the average of the two and then 90% (as in the article) (CP = 235 watts) or Golden cheetah estimation (CP = 205 watts?) What do you think?. I think its a big diff so a iam unsure :)

It's not GC in particular but 2 parameter CP model which gives 205watts with those data, the 90% of the average may be a way to account for you are not completely fresh for the longer test using that protocol, but I have no idea which base it has.

I see that the long rung should at least be 15 min according to GC but surely that are cycling and 9 minutes should work anyway for run?

A shorter "long" test tends to overestimate your CP not to underestimate it, I would do a longer test but when fresh, not after other maximal effort. 

Also.. on chart "Ride" i can't find Pace (min/km), only speed?

https://github.com/GoldenCheetah/GoldenCheetah/wiki/FAQ-RUNNING-&-SWIMMING#how-can-i-see-pace-in-activity-charts 

[KJA]

Also.. on chart "Ride" i can't find Pace (min/km), only speed?

https://github.com/GoldenCheetah/GoldenCheetah/wiki/FAQ-RUNNING-&-SWIMMING#how-can-i-see-pace-in-activity-charts 

Is it possible to get pace on the axis instead of kph?

What is pmax (some kind of anaerobic performance in ml.kg-1.min?) and how do i test it? Is is it relevant for running?  (i see it on the Run: set power zone)

Thanks for all your answers.

[Ale Martinez]

El miércoles, 2 de agosto de 2017, 5:06:11 (UTC-3), KJA escribió:

Also.. on chart "Ride" i can't find Pace (min/km), only speed?

https://github.com/GoldenCheetah/GoldenCheetah/wiki/FAQ-RUNNING-&-SWIMMING#how-can-i-see-pace-in-activity-charts 

Is it possible to get pace on the axis instead of kph?

No 

What is pmax (some kind of anaerobic performance in ml.kg-1.min?) and how do i test it? Is is it relevant for running?  (i see it on the Run: set power zone)

Max power in watts. Methods based on GPS data are not any good at estimating it, I'm not sure about the available running power meters but I would not put much stock on it.

[Nigel Van de Velde]

Hi,

The most recent firmware on the Garmin 735 (8.30) now saves the performance condition into the fit file. It imports fine into GoldenCheetah.

Although the algorithm behind it is unknown, I always found this a useful metric in regard of how I felt during a run. So probably useful to track on longer term as well.

Kind regards,

Nigel

[Faisal Ghani]

Garmin Swim Drill Mode.

Does swim drills especially with fins have any effect on CV model? I swim faster or at the same pace as my Threshold Pace with less effort when using fins and sometimes paddles. 

[Ale Martinez - No direct email please]

El jueves, 13 de junio de 2019, 3:18:23 (UTC-3), Faisal Ghani escribió:

Garmin Swim Drill Mode.

Does swim drills especially with fins have any effect on CV model? I swim faster or at the same pace as my Threshold Pace with less effort when using fins and sometimes paddles. 

I'm afraid yes, currently drills and swimming are not separated in speed-duration calculations nor do we have different curves for different strokes, it would be something interesting but it is not currently available.

You can use filters to exclude sessions were you have done long intervals using fins if you mark them via metadata (s.t. Equipment="Fins"), but the whole session will be excluded, currently there is no way to exclude only the fins part.