TSB returning to zero yet I'm sore



grahamspringett

New Member
Feb 26, 2004
263
0
0
After a fairly lengthy dose of SST riding I've now started to do L5 intervals. I did 2 sessions this week, Tuesday 6x4 and Thursday 4x4.

Due to the decreased volume my TSS is falling and my TSB is returning up from -25 on Sunday up to -2 today.

My legs feel heavy and I'm hoping to be recovered for the Saturday morning world champs (why else do we train so hard?) but I'm intrigued that my TSB is so close to zero yet my legs feel rubbish at the moment.

I'm wondering if I should just do one L5 session and keep the volume up ready for racing in early March, or keep going, ignore the CTL and TSB numbers and just go on how I feel.

Perhaps a few weeks of L5 then return to increased volume?
 
grahamspringett said:
After a fairly lengthy dose of SST riding I've now started to do L5 intervals. I did 2 sessions this week, Tuesday 6x4 and Thursday 4x4.

Due to the decreased volume my TSS is falling and my TSB is returning up from -25 on Sunday up to -2 today.
Yeah that's pretty typical with a transition to high end work. It's hard to do both quantity and quality and a block of focused L5 or L6 work tends to cost weekly TSS and with it CTL.
My legs feel heavy and I'm hoping to be recovered for the Saturday morning world champs (why else do we train so hard?) but I'm intrigued that my TSB is so close to zero yet my legs feel rubbish at the moment.
Since the inception of the PMC there's been quite a bit of discussion about ATL time constants or IOW, how to model time to recovery and whether it changes with: age, CTL, training intensity, etc. I wasn't involved in the beta testing phase, but I understand that the beta version allowed dynamic ATL time constants but the idea was scrapped as most folks didn't find it useful. But there are a lot of reasons your actual recovery might not match the model.

FWIW, I looked back over my first year and a half of PMC data a while back and visually matched up my peak performances against TSB with candidate ATL time constants. I got a pretty good in season match by setting the ATL Tc to 9 days vs. the 7 day default. I'm a Master's racer and that value seemed to give me good TSB predictions during the height of the summer season. But I have no doubt that the time to recovery could also vary with local workout intensity and that your recovery might be slower while you do 2 a week L5 sessions, especially while your adjusting to the increased weekly intensity.

But in the end the PMC is a modeling tool, a good one no doubt but still a model that won't always match up with reality. In addition to tuning your ATL Tc, there seems to be some individual variation in terms of the ideal event day TSB and whether it needs to be positive, strongly positive or just not deeply negative. IOW, you might still fell tired at 0 but might feel great after it rises to say +8 then drops back a point or two after an openers ride or two. But that's getting into peaking strategies for important events. I personally wouldn't spend my limited "peaking bullets" on the local weekly hammerfest. I'd save my peaking for sanctioned events I care about but that's a personal choice.
...I'm wondering if I should just do one L5 session and keep the volume up ready for racing in early March, or keep going, ignore the CTL and TSB numbers and just go on how I feel....Perhaps a few weeks of L5 then return to increased volume?
Those are the tough decisions in terms of load balancing, avoiding early peaks, doing some "pull up" blocks during base build etc. There are as many theories out there as there are coaches but at least IMO you're asking the right questions and looking at the right metrics.

One way to understand performance peaking is by looking at your CTL curve. If it starts to droop you'll be pretty fresh but you'll also be stressing your body less on a daily basis. This is the classic way to bring on a peak but it also means your body will over time start adjusting to the lower training load. Yeah it's more complex than that when you consider all the possible variations in training intensity and ways to let that CTL sag. But in simple terms you'll peak when your load drops and you'll slowly start to detrain so the peak won't last indefinitely.

I've found my peaks are sharp and dramatic when I taper from a steep short term CTL build. IOW, after an extended period of flat or rising CTL I punch out a couple of hard days to induce a sharp CTL spike. Then I rest a bit, resume more normal training perhaps with some extra intensity but less load and I get my most pronounced but relatively short lived performance peaks. Tapering more slowly without the sharp CTL spike tends to give me a longer but less dramatic peak and is my choice if I have several weeks of races but not a clear A race.

Anyway, load balancing is key if you want these peaks to come when you want them and not here and there throughout the training year. Personally I hold off on dedicated L5 blocks until I'm outside in the spring so I can offset the lost CTL with some longer Tempo rides and maintain my CTL build. But you could also do a short L5 block as you mentioned then go back into rebuilding prior to spring racing. There are a lot of options but think in terms of overall CTL pattern and of course specific workout intensities to target desired systems. Finding the ideal balance is tough but by paying attention to both specific workout goals and overall load patterns you have visibility into the tradeoffs and can make a plan.

Good luck,
-Dave
 
daveryanwyoming said:
Yeah that's pretty typical with a transition to high end work. It's hard to do both quantity and quality and a block of focused L5 or L6 work tends to cost weekly TSS and with it CTL.
Since the inception of the PMC there's been quite a bit of discussion about ATL time constants or IOW, how to model time to recovery and whether it changes with: age, CTL, training intensity, etc. I wasn't involved in the beta testing phase, but I understand that the beta version allowed dynamic ATL time constants but the idea was scrapped as most folks didn't find it useful. But there are a lot of reasons your actual recovery might not match the model.

FWIW, I looked back over my first year and a half of PMC data a while back and visually matched up my peak performances against TSB with candidate ATL time constants. I got a pretty good in season match by setting the ATL Tc to 9 days vs. the 7 day default. I'm a Master's racer and that value seemed to give me good TSB predictions during the height of the summer season. But I have no doubt that the time to recovery could also vary with local workout intensity and that your recovery might be slower while you do 2 a week L5 sessions, especially while your adjusting to the increased weekly intensity.
+1
I suspect that factors impacting the ATL TC might also include a lengthening of the TC during a period of increased intensity, particularly when higher pedal forces are introduced. This would correlate with the increased perception of fatigue despite a relatively flat CTL when doing a block of L5 / L6 work.
I have also noticed just recently that the trade off for a nice FTP bump is that the ATL TC and CTL TC will play against each other for a while, resulting in a relatively falsely higher TSB. I suspect that TSB is invalid for perhaps one ATL TC duration (or a bit more) due to the maths. Of course, the size of the FTP bump will have a bearing on the relative invalidity of TSB immediately folowing an FTP bump.
 
I thought changing the time constants based on intensity was scrapped, but that the general consensus was that increasing age did tend to increase the ATL constant? In any case, I agree that for this individual the situation decribed above would make a case for increasing the ATL time constant somewhat.
 
daveryanwyoming said:
Yeah that's pretty typical with a transition to high end work. It's hard to do both quantity and quality and a block of focused L5 or L6 work tends to cost weekly TSS and with it CTL.
Since the inception of the PMC there's been quite a bit of discussion about ATL time constants or IOW, how to model time to recovery and whether it changes with: age, CTL, training intensity, etc. I wasn't involved in the beta testing phase, but I understand that the beta version allowed dynamic ATL time constants but the idea was scrapped as most folks didn't find it useful. But there are a lot of reasons your actual recovery might not match the model.

FWIW, I looked back over my first year and a half of PMC data a while back and visually matched up my peak performances against TSB with candidate ATL time constants. I got a pretty good in season match by setting the ATL Tc to 9 days vs. the 7 day default. I'm a Master's racer and that value seemed to give me good TSB predictions during the height of the summer season. But I have no doubt that the time to recovery could also vary with local workout intensity and that your recovery might be slower while you do 2 a week L5 sessions, especially while your adjusting to the increased weekly intensity.

But in the end the PMC is a modeling tool, a good one no doubt but still a model that won't always match up with reality. In addition to tuning your ATL Tc, there seems to be some individual variation in terms of the ideal event day TSB and whether it needs to be positive, strongly positive or just not deeply negative. IOW, you might still fell tired at 0 but might feel great after it rises to say +8 then drops back a point or two after an openers ride or two. But that's getting into peaking strategies for important events. I personally wouldn't spend my limited "peaking bullets" on the local weekly hammerfest. I'd save my peaking for sanctioned events I care about but that's a personal choice.
Those are the tough decisions in terms of load balancing, avoiding early peaks, doing some "pull up" blocks during base build etc. There are as many theories out there as there are coaches but at least IMO you're asking the right questions and looking at the right metrics.

One way to understand performance peaking is by looking at your CTL curve. If it starts to droop you'll be pretty fresh but you'll also be stressing your body less on a daily basis. This is the classic way to bring on a peak but it also means your body will over time start adjusting to the lower training load. Yeah it's more complex than that when you consider all the possible variations in training intensity and ways to let that CTL sag. But in simple terms you'll peak when your load drops and you'll slowly start to detrain so the peak won't last indefinitely.

I've found my peaks are sharp and dramatic when I taper from a steep short term CTL build. IOW, after an extended period of flat or rising CTL I punch out a couple of hard days to induce a sharp CTL spike. Then I rest a bit, resume more normal training perhaps with some extra intensity but less load and I get my most pronounced but relatively short lived performance peaks. Tapering more slowly without the sharp CTL spike tends to give me a longer but less dramatic peak and is my choice if I have several weeks of races but not a clear A race.

Anyway, load balancing is key if you want these peaks to come when you want them and not here and there throughout the training year. Personally I hold off on dedicated L5 blocks until I'm outside in the spring so I can offset the lost CTL with some longer Tempo rides and maintain my CTL build. But you could also do a short L5 block as you mentioned then go back into rebuilding prior to spring racing. There are a lot of options but think in terms of overall CTL pattern and of course specific workout intensities to target desired systems. Finding the ideal balance is tough but by paying attention to both specific workout goals and overall load patterns you have visibility into the tradeoffs and can make a plan.

Good luck,
-Dave
Danged nice post Dave. You back in winter writing mode again? :D

fwiw, I've noticed the same effect from a short burst of 2-3-4d TSS load on top of a nice build. I found that the std. 42d CTL tc tends to mask this somewhat and now keep a custom chart with 42d and 15d (!) CTL overlaid. The 15d CTL shows the spike more clearly ---though not to the same extent as the ATL spike.

Re zero TSS and bad legs for the OP. Hey sometimes the wheels just fall off the wagon: give me 2-3 consecutive days of a negative disconnect between power vs. PE and TSB and I know I need a light week. Typically that's around 30-40% reduction in load measured by TSS.

Also of course, any time you introduce a 'new' more intense workout - the body will likely protest for a while.