Setting ATL & CTL Constants

[LT Intolerant]

I am struggling with setting the PMC constants. I've searched for a thread in regards to the conditions under which someone should change their PMC constants and I've read the PMC info on the CP site but I'm still unsure about what to do here (maybe nothing).

For what it's worth I've been using the std PMC constants (i.e., 42/7) and find that my bests for 1, 5, and 60 min power occur when my CTL is > 100 (averaged 105 with a high of 130 in 2007) and my TSB is between -9 and 1.

I'm curious if others have changed their constants. What you changed to and why. I guess my question is under what conditions should someone change their constants?

Thanks in advance for any help.

gene r

 

[Alex Simmons]

Yes, it's an interesting question. Apart from TSB at time of MMP peaks, also have a look at the direction TSB was heading in the days leading up.

In general terms, the shorter the event being targeted, the longer the ATL Time Constant could be. Everyone is different.

CTL is rather less sensitive to changes in the CTL TC and there really isn't much reason to change it from the default 42 day value.

As a suggestion, clone your PMC and reset the ATL TC and see if it matches your performances any more closely. Also see how it matches your diary comments too - e.g. legs were **** today, or feeling great, could rip the cranks off and so on.

 

[NomadVW]

Hey Gene,

I've played around with it a little, but have come to settle on using the defaults more out of convenience.

I look at a couple things when I'm looking back at my data to compare. I run my Excel spreadsheet with a standard PMC running the default constants. But on a separate page I keep a rolling/non weighted 7 day average TSS. My experience this year is that a trending positive TSB is a good thing, but if the 7 day non-weighted average is still going to be above a certain point in relation to the CTL, performance isn't quite as good.


In the spring, my best performances came when my 7 day rolling average was less than ~115ish and for my peak race month it was mostly less than 100 and I placed 5th, 1st, and 1st in my events that month. CTL was in the 120's at the beginning of peaking.

In the fall, my CTL was tipping 140 and the 7 day rolling avg was "good for feeling" in the 110-120 range. I suppose I could run through the numbers and come up with a ATL constant that reflects that a little more closely, but I really don't worry about it much anymore.

 

[LT Intolerant]

Yes, it's an interesting question. Apart from TSB at time of MMP peaks, also have a look at the direction TSB was heading in the days leading up.

In general terms, the shorter the event being targeted, the longer the ATL Time Constant could be. Everyone is different.

CTL is rather less sensitive to changes in the CTL TC and there really isn't much reason to change it from the default 42 day value.

As a suggestion, clone your PMC and reset the ATL TC and see if it matches your performances any more closely. Also see how it matches your diary comments too - e.g. legs were **** today, or feeling great, could rip the cranks off and so on.

Thanks Alex. I'll give that a try.

gene r

 

[LT Intolerant]

Hey Gene,

I've played around with it a little, but have come to settle on using the defaults more out of convenience.

I look at a couple things when I'm looking back at my data to compare. I run my Excel spreadsheet with a standard PMC running the default constants. But on a separate page I keep a rolling/non weighted 7 day average TSS. My experience this year is that a trending positive TSB is a good thing, but if the 7 day non-weighted average is still going to be above a certain point in relation to the CTL, performance isn't quite as good.


In the spring, my best performances came when my 7 day rolling average was less than ~115ish and for my peak race month it was mostly less than 100 and I placed 5th, 1st, and 1st in my events that month. CTL was in the 120's at the beginning of peaking.

In the fall, my CTL was tipping 140 and the 7 day rolling avg was "good for feeling" in the 110-120 range. I suppose I could run through the numbers and come up with a ATL constant that reflects that a little more closely, but I really don't worry about it much anymore.

Thanks for the insight Nomad. I figured worse case I'd use the defaults and just live w the fact that my bests come in a range that is different from the norms that show up in the data Andy C has sunshined.

BTW that's one helluva 2007! Congrats.

gene r

 

[acoggan]

I am struggling with setting the PMC constants. I've searched for a thread in regards to the conditions under which someone should change their PMC constants and I've read the PMC info on the CP site but I'm still unsure about what to do here (maybe nothing).

For what it's worth I've been using the std PMC constants (i.e., 42/7) and find that my bests for 1, 5, and 60 min power occur when my CTL is > 100 (averaged 105 with a high of 130 in 2007) and my TSB is between -9 and 1.

I'm curious if others have changed their constants. What you changed to and why. I guess my question is under what conditions should someone change their constants?

Thanks in advance for any help.

gene r

1. What about your performance at other durations (e.g., 5 s, but also the rest tracked by default by WKO+)?

2. You might try seeing how things line up when you set your ATL time constant to 5 d.

 

[LT Intolerant]

1. What about your performance at other durations (e.g., 5 s, but also the rest tracked by default by WKO+)?.

here are my bests and the corresponding ATL & CTL

5 secs = ATL was 117, CTL was 102, TSB was -1
1 min = ATL was 138, CTL was 123, TSB was 1
5 min = ATL was 140, CTL was 123, TSB was -9
20 min = ATL was 138, CTL was 123, TSB was 1
60 min* = ATL was 110, CTL was 114, TSB was 4 (note: equaled my best 2 months later when TSB was -2)

*I realize now I made a mistake in my OP insofar as my TSB for my 60 min best is concerned. So my TSB range for PBs is now -9 to 4. My apologies.

2. You might try seeing how things line up when you set your ATL time constant to 5 d.

I made the above change and now my TSB range for PBs is still -9 to 4. Interestingly enough my 5 sec, 5 min, and FTP all have TSBs of 4, while my 5 min time remains stubbornly at -9. This does make sense to some extent given that my 5 min power is my strength, or in other words, the highest rung on my power profile.

Thanks for the help Andy.

gene r

 

[Alex Simmons]

Is this over a whole season's or year's worth of data?

 

[peterpen]

I'd also look at two things:
-when did these bests occur? specific testing? racing?
-what is your typical TSB range (and min to max) ?

While over the past two years I've set two bests with high TSB (>+10) (1m and 5m, both during a season break after a taper for a race period) the majority of my bests are all over the place in terms of TSB. I generally track max avgP for3m, 15m, and NP for 60m - I've set bests (top 10 for 2 seasons) on the same date twice, with TSB -6.7 and -13.3. Like 90% of my bests, these were during races.

 

[peterpen]

I'd also look at two things:
-when did these bests occur? specific testing? racing?
-what is your typical TSB range (and min to max) ?

While over the past two years I've set two bests with high TSB (>+10) (1m and 5m, both during a season break after a taper for a race period) the majority of my bests are all over the place in terms of TSB. I generally track max avgP for3m, 15m, and NP for 60m - I've set bests (top 10 for 2 seasons) on the same date twice, with TSB -6.7 and -13.3. Like 90% of my bests, these were during races.

 

[acoggan]

I've set bests (top 10 for 2 seasons) on the same date twice, with TSB -6.7 and -13.3. Like 90% of my bests, these were during races.

So what happens if you race when your TSB is higher?

 

[acoggan]

here are my bests and the corresponding ATL & CTL

5 secs = ATL was 117, CTL was 102, TSB was -1
1 min = ATL was 138, CTL was 123, TSB was 1
5 min = ATL was 140, CTL was 123, TSB was -9
20 min = ATL was 138, CTL was 123, TSB was 1
60 min* = ATL was 110, CTL was 114, TSB was 4 (note: equaled my best 2 months later when TSB was -2)

I was actually thinking of all the durations that are listed in the Ranges box, i.e., 5 s, 10 s, 20 s, 30 s, 1 min, 2 min, 5 min, 10 min, 20 min, 30 min, 60 min, and also the relevant normalized powers (i.e., 5 min, 10 min, 20 min, 30 min, 60 min). That's what I've previously used to get a feel for the optimal TSB, at least on average (i.e., across workouts/races and across individuals) - see http://www.cyclingpeakssoftware.com/pmc_summit.pdf.

(BTW, for simplicity's sake I use the average TSB for those durations at which someone hits the exact same maximal power on more than one occasion.)

 

[peterpen]

So what happens if you race when your TSB is higher?

Depends.

I have set my two of my top five 3m efforts during races with +12TSB and +17.4 TSB (won one race, 11th in the other.) But I've also performed well below par at a stage race I entered with +28TSB (TSB tending upward from -30 three weeks prior) and had many unremarkable performances with TSB > +10.

Hmm, what's my point? I guess that I have yet to find any 1:1 correlation between TSB and performance bests, particularly in races. I will note that I felt my best racing came in late July/August, when I was generally running TSB between -10 and +10. But (and maybe this is what I'm getting at) I didn't set *any* top 10 power bests during that period, despite getting half a dozen top tens in a more competitive field than I usually raced (35+ 1/2/3 vs. cat 3 races.) Conclusion - I had good enough fitness but raced smarter.

btw, in looking back at this data, I thought how nice it would be to have TSB as a column for the WKO+ calendar, where one call also see the Workout Goal and Notes sections.

-Peter

 

[acoggan]

Depends.

I have set my two of my top five 3m efforts during races with +12TSB and +17.4 TSB (won one race, 11th in the other.) But I've also performed well below par at a stage race I entered with +28TSB (TSB tending upward from -30 three weeks prior) and had many unremarkable performances with TSB > +10.

Hmm, what's my point? I guess that I have yet to find any 1:1 correlation between TSB and performance bests, particularly in races. I will note that I felt my best racing came in late July/August, when I was generally running TSB between -10 and +10.

I certainly wouldn't call it a 1:1 correlation, but I would argue that you have, in fact, determined for yourself just how much fitness to give up to gain how much freshness, and that your responses are consistent with those of others (again, see the last few slides here: http://www.cyclingpeakssoftware.com/pmc_summit.pdf).

 

[LT Intolerant]

Is this over a whole season's or year's worth of data?

The data I have goes back to late June when I changed PMs to a more "precise" PM.

When I look at my PBs and other days that were within 5% or less of my PB for select times the TSB range is slightly more expanded. Here is what I came up with...

5 sec = -1, 7, 15, -15, -1, 9
1 min = 1, 21, -1, -15, 12, 11, 5
5 min = 0, -9, 1, 18, 20
60 min = -9, 7, -18, 3, 5, 5, 18

I still need to go back and read through my old notes to see where I felt like my legs were good or lousy and see what my corresponding TSB was at that time.

gene r

 

[LT Intolerant]

I was actually thinking of all the durations that are listed in the Ranges box, i.e., 5 s, 10 s, 20 s, 30 s, 1 min, 2 min, 5 min, 10 min, 20 min, 30 min, 60 min, and also the relevant normalized powers (i.e., 5 min, 10 min, 20 min, 30 min, 60 min). That's what I've previously used to get a feel for the optimal TSB, at least on average (i.e., across workouts/races and across individuals) - see http://www.cyclingpeakssoftware.com/pmc_summit.pdf.

(BTW, for simplicity's sake I use the average TSB for those durations at which someone hits the exact same maximal power on more than one occasion.)

Ahh, got it. I'll go back and take a look and see if can pick up on any trends in the data. Thanks.

gene r

 

[LT Intolerant]

I'd also look at two things:
-when did these bests occur? specific testing? racing?
-what is your typical TSB range (and min to max) ?

While over the past two years I've set two bests with high TSB (>+10) (1m and 5m, both during a season break after a taper for a race period) the majority of my bests are all over the place in terms of TSB. I generally track max avgP for3m, 15m, and NP for 60m - I've set bests (top 10 for 2 seasons) on the same date twice, with TSB -6.7 and -13.3. Like 90% of my bests, these were during races.

I thougth about the fact that I really don't do any formal testing once race season begins, and maybe that's my issue. That is I'm not testing formally when my TSB is trending positively because I'm "saving" my legs for race day.

Some on my bests have come in races, most in hard group rides, and my 60 min on a hillclimb with a small group.

In another post as you can see I have gone back and looked at my TSB range for 5 sec, 1, 5 and 60 min power and there is certainly a more expanded TSB range for PBs and readings within 5% of my PB for a certain duration.

Thanks for the thoughts and ideas Peter.

gene r

 

[Alex Simmons]

The data I have goes back to late June when I changed PMs to a more "precise" PM.

When I look at my PBs and other days that were within 5% or less of my PB for select times the TSB range is slightly more expanded. Here is what I came up with...

5 sec = -1, 7, 15, -15, -1, 9
1 min = 1, 21, -1, -15, 12, 11, 5
5 min = 0, -9, 1, 18, 20
60 min = -9, 7, -18, 3, 5, 5, 18

I still need to go back and read through my old notes to see where I felt like my legs were good or lousy and see what my corresponding TSB was at that time.

gene r

Well that's a neutral to positive bias in TSB for near PB performances.