Login required to started new threads

Login required to post replies

Does this PMC chart look right?
Quote | Reply
I've started using Training Peaks about 6 months ago to track and plan my workouts. I'm tracking TSS scores across all three disciplines so that's what's represented on this chart.
Right before I started logging anything, I did an FTP test and a 10k run race and determined LTHR, pace zones, power zones etc... based on those. With swimming I've set zones based on a 1500yd TT and I use a Garmin Swim watch to track every workout.

There are two peaks on the chart, one corresponding to the training leading to my first race of the season. That race ended up in a crash which resulted in 1 week off training. The second peak is the 2nd race a month later where I've had a good performance and ended up winning my age group. I took a very easy week after that and started my build for IMMT in August. I'm currently 4 (of 16) weeks in and just finished planning my next 4 week block.

1)
Back to the PMC chart, I see a pretty steep increase in CTL in the first 6-7 weeks. Is this normal? Also notice how the TSB (yellow line) is totally in the gutter at the beginning peaking to -76 or so which would indicate a ton of fatigue.
The chart seems to look more realistic after 4-5 weeks but my concern is that the quick ramp-up at the beginning is overly inflating my CTL.

2)
Another thing: I've been using a Kickr with very questionable power data for a couple of weeks. Basically, some of my power data was inflated by 25-30W. At some point at the beginning of January they came out with a patch which allowed me to use the PM to driver the Kickr and get accurate results. I knew that I had some garbage data in there so I went and deleted those 2 weeks worth of questionable workouts. That data was from Jan 1st to about Jan 12. So at this point my training log starts on Jan 12 and is up to date containing accurate data.

I expected the PMC chart to change but didn't notice any change in CTL. CTL was about 105 before and after removing the garbage... Does this make sense?



3)
CTL ramp rates:

7 days: 5.7
Last 28 days: 27.2
Last 90 days: 28.3
Last 365 days: 106.5


My training load has been about 15h for the past 4 weeks and on average 10-11h/week since the beginning of the year with some 5-6 weeks of 15h in there.
Watching a TP video where Gordo Byrn went over PMC data for IM training they mentioned that a healthy CTL ramp rate would be around 10/month. I'm not sure if they were counting just cycling data or overall swim/bike/run in that number. It would seem that I'm either overtraining like crazy or my PMC chart is not exactly accurate.


Thanks y'all!



Last edited by: sp1ke: May 24, 15 9:18
Quote Reply
Re: Does this PMC chart look right? [sp1ke] [ In reply to ]
Quote | Reply
It seems fine to me and and shows a progressive and logical approach for a build. Recall, CTL 7 week average, ATL 7 day average. Your TSB between -25 to -40 seems similar to mine but the question is whether you can maintain that level for the duration of the build which seems to be about 8-weeks.

Just a question, how do you get your planned parameters to display on the chart? do you have to plan so far ahead for it to show or is that a feature I am missing in the charts.


Hope this helps and good luck.
Quote Reply
Re: Does this PMC chart look right? [sp1ke] [ In reply to ]
Quote | Reply
When you start from a CTL / ATL of 0, you're going to get MASSIVE increases right away, which probably aren't the most accurate or representative of what you're actually doing. My advice would be, unless you're truly coming off the couch, start your CTL around 25-50 depending on how much time you took off between training or have been consistently training. The charts once you got to a certain range are looking good!

As for the Wahoo KickR.... do you calibrate the device 5-10min into every ride?? Also, once a Wahoo KickR warms up, it tends to lose it's calibration number. That's why I recommend calibrating once it has a chance to warm up for a few minutes. Or even calibration at 10min in, and then after your 1st interval or 30min effort. In our Wahoo KickR studio, we find that the power can be 5% off of your regular powermeter. Wahoo has never released their % error. So we can only speculate that a Powermeter is ~ 1-3% error, and a trainer can definitely be 2-5%. So right there, you could have a 3-8% error between the two numbers. Best to find the correct mode (Erg / Slope / etc) that lets you be as close to your PERSONAL powermeter as possible!

Hope this helps!

Carson Christen
Sport Scientist / Coach
Torden Multisport
Quote Reply
Re: Does this PMC chart look right? [sp1ke] [ In reply to ]
Quote | Reply
sp1ke wrote:
1)
Back to the PMC chart, I see a pretty steep increase in CTL in the first 6-7 weeks. Is this normal? Also notice how the TSB (yellow line) is totally in the gutter at the beginning peaking to -76 or so which would indicate a ton of fatigue.
The chart seems to look more realistic after 4-5 weeks but my concern is that the quick ramp-up at the beginning is overly inflating my CTL.

Since there is no data before January and CTL is a "42-day exponentially weighted rolling average", you will see that sharp increase for the first 42 days that the chart actually has data. It levels out there and that is a better representation of your actual CTL. You had some level of fitness when you started in January so basically disregard that first sharp curve.

sp1ke wrote:
2)
Another thing: I've been using a Kickr with very questionable power data for a couple of weeks. Basically, some of my power data was inflated by 25-30W. At some point at the beginning of January they came out with a patch which allowed me to use the PM to driver the Kickr and get accurate results. I knew that I had some garbage data in there so I went and deleted those 2 weeks worth of questionable workouts. That data was from Jan 1st to about Jan 12. So at this point my training log starts on Jan 12 and is up to date containing accurate data.

I expected the PMC chart to change but didn't notice any change in CTL. CTL was about 105 before and after removing the garbage... Does this make sense?

If it was more than 6 weeks ago, it shouldn't matter.

sp1ke wrote:
3)
My training load has been about 15h for the past 4 weeks and on average 10-11h/week since the beginning of the year with some 5-6 weeks of 15h in there.
Watching a TP video where Gordo Byrn went over PMC data for IM training they mentioned that a healthy CTL ramp rate would be around 10/month. I'm not sure if they were counting just cycling data or overall swim/bike/run in that number. It would seem that I'm either overtraining like crazy or my PMC chart is not exactly accurate.


The chart looks good to me. I would assume Gordo was talking about cycling only, but could be wrong. The numbers are a representation of fitness and fatigue, and you have to assign the value to them based on how you feel.
Quote Reply
Re: Does this PMC chart look right? [triordie1994] [ In reply to ]
Quote | Reply
If your future workouts have planned TSS you can change the PMC dates to model future values.
Quote Reply
Re: Does this PMC chart look right? [triordie1994] [ In reply to ]
Quote | Reply
triordie1994 wrote:
CTL 7 week average, ATL 7 day average.

This is incorrect. CTL and ATL are calculated as exponentially-weighted moving averages with default time constants of 42 and 7 d, respectively. This means that CTL mostly (~90%) reflects what you've done in the last ~3 mo, whereas ATL mostly (again, ~90%) what you've done in the last ~2 wk. However, everything prior to that going back to your very first workout contributes to some extent, if only to an infinitesimal degree.
Quote Reply
Re: Does this PMC chart look right? [sp1ke] [ In reply to ]
Quote | Reply
sp1ke wrote:
I've started using Training Peaks about 6 months ago to track and plan my workouts. I'm tracking TSS scores across all three disciplines so that's what's represented on this chart.
Right before I started logging anything, I did an FTP test and a 10k run race and determined LTHR, pace zones, power zones etc... based on those. With swimming I've set zones based on a 1500yd TT and I use a Garmin Swim watch to track every workout.

There are two peaks on the chart, one corresponding to the training leading to my first race of the season. That race ended up in a crash which resulted in 1 week off training. The second peak is the 2nd race a month later where I've had a good performance and ended up winning my age group. I took a very easy week after that and started my build for IMMT in August. I'm currently 4 (of 16) weeks in and just finished planning my next 4 week block.

1)
Back to the PMC chart, I see a pretty steep increase in CTL in the first 6-7 weeks. Is this normal? Also notice how the TSB (yellow line) is totally in the gutter at the beginning peaking to -76 or so which would indicate a ton of fatigue.
The chart seems to look more realistic after 4-5 weeks but my concern is that the quick ramp-up at the beginning is overly inflating my CTL.

2)
Another thing: I've been using a Kickr with very questionable power data for a couple of weeks. Basically, some of my power data was inflated by 25-30W. At some point at the beginning of January they came out with a patch which allowed me to use the PM to driver the Kickr and get accurate results. I knew that I had some garbage data in there so I went and deleted those 2 weeks worth of questionable workouts. That data was from Jan 1st to about Jan 12. So at this point my training log starts on Jan 12 and is up to date containing accurate data.

I expected the PMC chart to change but didn't notice any change in CTL. CTL was about 105 before and after removing the garbage... Does this make sense?



3)
CTL ramp rates:

7 days: 5.7
Last 28 days: 27.2
Last 90 days: 28.3
Last 365 days: 106.5


My training load has been about 15h for the past 4 weeks and on average 10-11h/week since the beginning of the year with some 5-6 weeks of 15h in there.
Watching a TP video where Gordo Byrn went over PMC data for IM training they mentioned that a healthy CTL ramp rate would be around 10/month. I'm not sure if they were counting just cycling data or overall swim/bike/run in that number. It would seem that I'm either overtraining like crazy or my PMC chart is not exactly accurate.


Thanks y'all!



1) As others have pointed out, the initially-steep ramp is because you've just started using the PMC, and reflects the absence of data prior to that point in time. Your choices are A) wait until a few months have passed, after which time your CTL will have "caught up", or B) "seed" the PMC with estimated starting values (is that possible in TrainingPeaks? It is in WKO+.)

2) Deleting data should have resulted in a change in your PMC, especially since you don't have much data to begin. I don't know why it didn't appear to do so.

3) If continued for more than a few weeks, a ramp rate of +10 TSS/d/d is rather steep, with such a rapid build-up risking illness, injury, and/or just plain ol' burnout. However, since your PMC lacks data, your ramp rate is as rapid as indicated.

4) One last point: I don't recommend adding TSS, rTSS, and sTSS (or even just TSS and rTSS) to derive a combined PMC. Not only are the three training metrics designed by different people are not directly comparable, to add them together is assuming that responses to training follow an identical time-course regardless of sport, something that is not supported by either the scientific literature or common sense.
Quote Reply
Re: Does this PMC chart look right? [triordie1994] [ In reply to ]
Quote | Reply
Thanks!
The total build is going to be 16 weeks and I'm currently 6 weeks in. I only entered my data for the next block of 4 weeks. I estimate my TSS based on previous data, simulating the workouts using the TrainerRoad custom workout editor and by educated guessing. I like it that when you do that, TP will predict your PMC chart so that's what you see there.
I've gone 14 weeks straight like this before, hopefully I can hold it together. I'm definitely going to take some easy/off days but I'm not planning for those specifically. My Fridays are usually swim only so that gives me a chance to recover a little bit before the big workouts on the weekend.


triordie1994 wrote:
It seems fine to me and and shows a progressive and logical approach for a build. Recall, CTL 7 week average, ATL 7 day average. Your TSB between -25 to -40 seems similar to mine but the question is whether you can maintain that level for the duration of the build which seems to be about 8-weeks.

Just a question, how do you get your planned parameters to display on the chart? do you have to plan so far ahead for it to show or is that a feature I am missing in the charts.


Hope this helps and good luck.
Quote Reply
Re: Does this PMC chart look right? [chrica04] [ In reply to ]
Quote | Reply
Thanks!
About the Kickr, I'm actually using the TrainerRoad feature they now have where you can have your PowerMeter control the load on the Kickr. The data is 100% accurate now because it's actually recording the power meter data not the Kickr's and the algorithm seems to work well as far as matching the load based on the target wattage. PerfPro studio also has this BTW.


chrica04 wrote:
When you start from a CTL / ATL of 0, you're going to get MASSIVE increases right away, which probably aren't the most accurate or representative of what you're actually doing. My advice would be, unless you're truly coming off the couch, start your CTL around 25-50 depending on how much time you took off between training or have been consistently training. The charts once you got to a certain range are looking good!

As for the Wahoo KickR.... do you calibrate the device 5-10min into every ride?? Also, once a Wahoo KickR warms up, it tends to lose it's calibration number. That's why I recommend calibrating once it has a chance to warm up for a few minutes. Or even calibration at 10min in, and then after your 1st interval or 30min effort. In our Wahoo KickR studio, we find that the power can be 5% off of your regular powermeter. Wahoo has never released their % error. So we can only speculate that a Powermeter is ~ 1-3% error, and a trainer can definitely be 2-5%. So right there, you could have a 3-8% error between the two numbers. Best to find the correct mode (Erg / Slope / etc) that lets you be as close to your PERSONAL powermeter as possible!

Hope this helps!
Quote Reply
Re: Does this PMC chart look right? [Andrew Coggan] [ In reply to ]
Quote | Reply
Andrew Coggan wrote:
4) One last point: I don't recommend adding TSS, rTSS, and sTSS (or even just TSS and rTSS) to derive a combined PMC. Not only are the three training metrics designed by different people are not directly comparable, to add them together is assuming that responses to training follow an identical time-course regardless of sport, something that is not supported by either the scientific literature or common sense.

Thank you! I think this is the most important point here as from a "feel" point of view I'm experiencing the difficulty and recovery cost of the workouts differently across the three sports. For example a swim TSS of 87 is pretty much a regular 1h workout which includes some hard work nothing epic. On the other hand a 1h ride accumulating the same TSS will feel much harder to me and the run version of that would be pretty tough.

I wish there was a way to get compensate TSS scores so that 1 TSS means the same thing across all three disciplines. The PMC chart would probably be start to be helpful then.
Quote Reply
Re: Does this PMC chart look right? [sp1ke] [ In reply to ]
Quote | Reply
sp1ke wrote:
I think this is the most important point here as from a "feel" point of view I'm experiencing the difficulty and recovery cost of the workouts differently across the three sports. For example a swim TSS of 87 is pretty much a regular 1h workout which includes some hard work nothing epic. On the other hand a 1h ride accumulating the same TSS will feel much harder to me and the run version of that would be pretty tough.

Even though I don't swim or run, that's entirely what I would expect.

sp1ke wrote:
I wish there was a way to get compensate TSS scores so that 1 TSS means the same thing across all three disciplines. The PMC chart would probably be start to be helpful then.

Keep in mind that the PMC was designed by a cyclist for cyclists. That said, your experience is why I've always suggested that triathletes use Foster's session RPE ("if it feels hard, it is hard") instead of TSS + rTSS + sTSS.
Last edited by: Andrew Coggan: May 24, 15 16:53
Quote Reply
Re: Does this PMC chart look right? [Andrew Coggan] [ In reply to ]
Quote | Reply
But ANDY!!! All of us at FasCat are slaves to the TSS. Frank, especially :D

I take time and try to estimate my TSS for my athletes and I's swimming as close as possible, based off similar sets / intervals / workouts on the bike. But I agree, it's hard to take that metric and relate FTP to LT swim speed. Bummer that TPeaks only can "estimate" off of time and LT swim speed. It's funny if I do a 5k set, and I only have a TSS of like 35!

Carson Christen
Sport Scientist / Coach
Torden Multisport
Quote Reply