TSS vs. TRIMPS

doctorSpoc

can someone compare and contrast the pros and cons of each?

is it true that TRIMPS (the zone weighted variety) may be better at predicting training "stress" (i.e. am I going to crash and burn) while TSS might do a better job at predicting training "stumulus" (i.e. how much am i provoking addaption though exersise 'stimulus')?

acoggan

1) TRIMP is based upon heart rate, and thus may be somewhat more variable due to extraneous influences.

2) TRIMP is based upon heart rate, and thus gives you less "credit" for supramaximal efforts (i.e., there is a ceiling effect, in that heart rate can't exceed maximal heart rate).

3) On the other hand, TRIMP places more weighting upon intensity, which may tend to counteract the impact of point #2 above.

Philosophically, you could argue that since power is the independent variable and heart rate is a dependent variable, TSS is a measure of stress, while TRIMP is a measure of strain. If you accept this premise, then what you have concluded above would, at least at first, seem logical. However, not all TSS points are created equal, which makes it difficult to say that TSS is a direct measure of the training stimulus (i.e., you have to take the composition of said TSS into account). Moreover, which is a better measure of stress vs. strain hinges in part on the considerations listed in 1-3 above. I would therefore say that the best way to view them is as simply two different ways of estimating the training load, one based on heart rate and one based on power, and essentially leave it at that. (Certainly, when used as inputs for the impulse-response model of Bannister there isn't a lot of reason to favor one over the other...although in the few cases I've examined, which happen to include a lot of supramaximal intervals, TSS seems to work a bit better.)

kmavm

One of my fantasies for cyclingpeaks is a pie graph breaking down TSS by zone (rather than TSS by hours). In fact, I'd like to be able to use TSS as an "x axis" in a lot of places: ride charts, etc.


Andy, my silly pubmed attempts at finding this "Bannister impulse-response model" are turning up nothing like a "model." I'd be really interested in a reference to anything that offers enough detail, e.g., for the creation of an excel spreadsheet.

acoggan

Given the way TSS is (currently) calculated, that really isn't possible. That is, TSS is based on what you did for the ride as a whole, and isn't easily decomposed into various parts.



Send me an e at acoggan at earthlink dot net (as a reminder) and I can send you a lengthy list of references. You'll need something a little more powerful/flexible than Excel to do the calculations, though.

srwings

srwings

fas_cat

Also, here is a web based list of related IR references w/ links to the full text articles

http://www.fascatcoaching.com/TSTWKT/

srwings

Excellent reading. Thanks.
I didn't see anything in there about converting a TRIMPS score (HR-based) to a TSS score (power-based). I'm a Noob to systematic training so I initially thought there was probably some sort of conversion formula much like converting Fahrenheit to Celsius. Now I see it's more like comparing apples to oranges. Hopefully someone will prove me wrong and suggest a method of conversion.

Regards.

fas_cat

[QUOTE=srwings]Excellent reading. Thanks.
I didn't see anything in there about converting a TRIMPS score (HR-based) to a TSS score (power-based).

As far as I know there is not. As an athlete or a coach I believe you need to choose either/or up front beforehand. Naturally, for single sport cyclists -- go TSS all the way.

If I were or I coached multisport athletes, I'd give TRIMPS a real go. Collect HR data for all 3 sports (nice and consistent) but also use power for the bike and TSS. You could actually study the TRIMP to the TSS.

john979

One could make a strong argument that TRIMPS is really a TSS and TSS is really a TRIMPS.

acoggan

I imagine that they are fairly interchangable except if you're doing a lot of very high intensity interval training - see the last couple of plots here for my reasoning:

http://www.cyclecoach.com/andycogga...ad/uksport1.ppt

john979

Andy;

Hence my comment, as TSS is more responsive to "impulse." Moreover since external factors such as heat affect HR, the overall "score" provided by TRIMPS is indicative of the total workout stress, not just the response the stress. Hence, TRIMPS would underscore a ride, say with many short, steep hills and overscore a steady-state ride on a hot day, at least in terms of the work and its training effect.

doctorSpoc

this kind of aludes to where i was coming from in the first place... although my language was a bit off...

i.e. TSS is really a good indicator of Training Stress (i.e. how much your training benefit you get from a workout) where as TRIMPS is a good indicator of training strain i.e. how bagged you will be the next and coming days (but not for supermaximal type stuff)... sometimes they track really well but sometimes, like in the example you gave in extreme heat, humidty etc. they are not necessarily the same thing.

e.g. i did a long 100km crit in 2006.. it was the 1st really hot day. max daily temp before that was around teens to 20C and then on the day of the crit it went up to 32C (60sF -> 90F) an humid. i thought i was going to die! when i looked at my power meter it was in tempo territory but HR was more indicative of how i was feeling. and the next few days i was in a bad way. TSS score gave me how much training benefit i got from the workout but if i had looked at my TRIMPS score i would have likely seen just how much that workout had given me a kick to the head.

JustCurious

Who tracks TRIMPS and how do you calculate it?