Power meters · Public discussion

Playing with TSS formula

Started by sugaken · · Last activity · 6 posts · 1,604 views

Thread navigation

Jump through the discussion

Go to the original post, the replies on this page, or the latest preserved contribution.

Thread details

What we know about this thread

Original section
Power meters
Published
13 April 2006
Last activity
14 April 2006
Original author
sugaken
Posts
6
Discussion status
Public discussion
Total views
1,604
Views / 30 days
0

The navigation and discussion metadata provide context. Posts remain in their original chronological order.

Showing posts 1–6 of 6
Posts remain in their original chronological order.

Text size
  1. This is not a question. It's something I find interesting which I found while playing with the formula to calculate TSS. You may or may not find it interesting, but I don't really care.😉 I'm just sharing in case any of you find it interesting.

    All of you here would probably recognize the algorithm below that's copied from Dr. Coggan's "Training and racing using a power meter: an introduction" paper written in 2003. I added my comments bracketted in []:

    1) starting at 30 s, calculate a 30 second rolling average for power
    2) raise the values obtained in step 1 to the 4th power
    3) take the average of all the values obtained in step 2
    4) take the 4th root of the number obtained in step 3
    [which is Normalized Power, or NP]
    5) divide the normalized power obtained in step 4 by the individual’s power at LT – the resulting decimal value is the IF
    [IF = NP / FTP]
    6) multiply the normalized power by the duration of the effort (in s) to obtain the normalized work performed (in J)
    [Normalized Work = NP * T]
    7) multiply the normalized work by the IF (step 5) to derive the “raw” TSS
    [raw TSS = NW * IF = NP * T * NP / FTP = NP^2 / FTP * T]
    8) divide the “raw” TSS by the amount of work that could be performed in one hour at threshold power (i.e., threshold power x 3600 s) and multiply by 100 to obtain the final TSS
    [final TSS = raw TSS / (FTP * 3600) * 100
    = NP^2 / FTP * T / (FTP * 3600) * 100
    = NP^2 / FTP^2 * T / 3600 * 100
    = IF^2 * t * 100
    where t is the work duration measured in hours; i.e. t = T/3600]

    So, in simplified form, TSS is really IF^2 * training duration in hours * 100.

    What this means is, it makes the training stress levels at intensity higher than FT look "quadratically more expensive" while doing the exact opposite to those at intensity lower than FT. While I'm not sure if the quadratic function is the best match to the actuall stress levels of training at various intensity levels, I'm sure it's much better than linear.

    Also, the simplified formula above makes it much simpler (well, for me, anyway) to estimate rough TSS of any given training session beforehand. You don't even need to know your own FTP to do this.

    TSS_L = 0
    TSS_H = 0
    For each training level (L1-L7)
    TSS_L = TSS_L + (the lower bound %FTP of the level in fraction)^2 * (duration you aim to work in that level) * 100
    TSS_H = TSS_H + (the upper bound %FTP of the level in fraction)^2 * (duration you aim to work in that level) * 100

    TSS_L will be the lower bound of TSS roughly estimated for your training session, and TSS_H the upper bound.

    By "%FTP of the level in fraction", I mean, for instance, if your target level is L3 (76-90%FTP), the upper and lower "fractions" will be 0.76 and 0.90. So, for instance you target to work for an hour at L3, your TSS for that particular part of session will fall in the range of 57-81 (0.76^2 * 1.0 * 100, and 0.90^2 * 1.0 * 100, respectively).

    Of course it's way too simplified calculation of TSS (for one thing, variable power is not considered at all), but when I try to do tempo work for an hour, I usually end up with TSS somewhere between upper 60s and upper 70s so it's a good estimate I presume.

    Cheers,

  2. sugaken said:

    in simplified form, TSS is really IF^2 * training duration in hours * 100

    Correct. I thought that this was common knowledge, but based on comments such as yours now I realize that it is not. I therefore regret the fact that we didn't cover this in the book...we'll have to fix that in a future edition. In any case, I'm glad to see that you've really dug into the formulae, to understand all the implications (and to make sure that I didn't make any mistakes!).

    sugaken said:

    the simplified formula above makes it much simpler (well, for me, anyway) to estimate rough TSS of any given training session beforehand. You don't even need to know your own FTP to do this.

    TSS_L = 0
    TSS_H = 0
    For each training level (L1-L7)
    TSS_L = TSS_L + (the lower bound %FTP of the level in fraction)^2 * (duration you aim to work in that level) * 100
    TSS_H = TSS_H + (the upper bound %FTP of the level in fraction)^2 * (duration you aim to work in that level) * 100

    TSS_L will be the lower bound of TSS roughly estimated for your training session, and TSS_H the upper bound.

    By "%FTP of the level in fraction", I mean, for instance, if your target level is L3 (76-90%FTP), the upper and lower "fractions" will be 0.76 and 0.90. So, for instance you target to work for an hour at L3, your TSS for that particular part of session will fall in the range of 57-81 (0.76^2 * 1.0 * 100, and 0.90^2 * 1.0 * 100, respectively).

    Of course it's way too simplified calculation of TSS (for one thing, variable power is not considered at all), but when I try to do tempo work for an hour, I usually end up with TSS somewhere between upper 60s and upper 70s so it's a good estimate I presume.

    The only addition or correction I would make would be to point out that normalized power is used in the calculation of TSS, in which case the "bounds" for, e.g., a typical level 3 training session are an IF of 0.85-0.95, not 76-90% (which is based on average power).

    Oh, wait: I see that in effect you already recognize this ("...for one thing, variable power is not considered at all..."😉. Still, I'm going to leave the above just in case its helpful to others.

  3. Thanks, Doc. I was kinda hoping to get a reply from the source.😉

    acoggan said:

    Correct. I thought that this was common knowledge, but based on comments such as yours now I realize that it is not.

    While it didn't take a rocket scientist to figure this out, it took me quite a while to realize it. And I can imagine there are people who would never bother with this kind of stuff.

    acoggan said:

    The only addition or correction I would make would be to point out that normalized power is used in the calculation of TSS, in which case the "bounds" for, e.g., a typical level 3 training session are an IF of 0.85-0.95, not 76-90% (which is based on average power).

    Oh, wait: I see that in effect you already recognize this ("...for one thing, variable power is not considered at all..."😉. Still, I'm going to leave the above just in case its helpful to others.

    Yes, you're right that I was aware of the NP vs AP issue; in my "way too simplified", unrealistic model for estimating TSS, one's power output is assumed to be constant which would result in NP equal to AP. For my purpose of estimating TSS for a given training session (or a set of sessions), I thought this should be good enough but I guess it would be more accurate to use the typical IF values listed in your paper.

    Anyway, I'm glad that my original post didn't get dismissed and ignored as trivial or utterly uninteresting.😎

    Thanks!

  4. frenchyge said:

    I use the following as a 'general guide':

    L2 = IF ~.8, TSS = 65 pts/hr
    L3 = IF ~.9, TSS = 80 pts/hr
    L4 = IF ~.95, TSS = 90 pts/hr
    L5 = IF ~1.0, TSS = 100 pts/hr


    cyclingforums.comt326335.html

  5. Correction:

    sugaken said:

    when I try to do tempo work for an hour, I usually end up with TSS somewhere between upper 60s and upper 70s

    The figures are not tempo for an hour. It's more like TSS of a tempo ride for 45-50min plus ~15min total of easy pedaling periods in which I "commute" between my home and the Cycling Road where I train.

    BTW, thanks, frenchyge, for your table. I will no longer need Excel to estimate TSS for my training plans.🙂

  6. sugaken said:

    While it didn't take a rocket scientist to figure this out, it took me quite a while to realize it. And I can imagine there are people who would never bother with this kind of stuff.

    Don't feel bad...for quite some time I didn't realize it myself! Thus, what I meant by "common knowledge" is that I thought that the reduced formula was well known, not that it was obvious from the equations that they could be reduced.

Active in the last 60 minutes

Active in this thread

0 users · 0 guests ·0 bots ·0 total

No signed-in users are active right now.

No known search crawlers active right now.