TRIMPS - what exactly is the weighting factor?

[normZurawski]

I see TRIMPS defined as...

TRIMPS = exercise duration x average HR x an intensity-dependent weighting factor

What exactly is "an intensity-dependent weighting factor"? I'm guessing there is no "exactly" to answer this question. Are there any guides to this factor? Should I just use RPE? Or do you simply make it up as you make it up and try to be consistent?

I searched the forums and the internet at large, and there's not a whole lot written about TRIMPS or even how to use it. I know TSS is what the power people like. Maybe next year that'll work out, especially if this Revolution turns out to be worth the investment. But then, TSS has the same weighting factor.

How to determine this weighting factor?

 

[frenchyge]

The Polar Performance software includes exercise dependent weighting factors. IIRC, whole-body sports like swimming produce higher HR's at a given intensity because the blood demands are more widespread, when compared to cycling. I think it's only necessary when comparing HR data from different sports, and that one sport would be the baseline, while another has a weighting factor of .9 or 1.1, or whatever.

If you're only dealing with cycling, I don't think you'd need to use it, but I'll try to remember to post the default Polar values when I get home tonight.

 

[kmavm]

The Polar Performance software includes exercise dependent weighting factors. IIRC, whole-body sports like swimming produce higher HR's at a given intensity because the blood demands are more widespread, when compared to cycling. I think it's only necessary when comparing HR data from different sports, and that one sport would be the baseline, while another has a weighting factor of .9 or 1.1, or whatever.

If you're only dealing with cycling, I don't think you'd need to use it, but I'll try to remember to post the default Polar values when I get home tonight.

I think something may have gotten lost in translation. The idea isn't that you multiply the whole TRIMPS score by some sort of weighing factor, but that you use a >1 weight factor for heart rates above a certain threshold. It functions like the ^4 in TSS.

 

[SolarEnergy]

For males, TRIMP = training duration x [(HRavg - HRrest)/(HRmax - HRrest)] x 0.64exp[(HRavg - HRrest)/(HRmax - HRrest) x 1.92], where HRmax is maximal HR and HRrest is resting HR.

For females, TRIMP = training duration x [(HRavg - HRrest)/(HRmax - HRrest)] x 0.86exp[(HRavg - HRrest)/(HRmax - HRrest) x 1.67].

 

[frenchyge]

I think something may have gotten lost in translation. The idea isn't that you multiply the whole TRIMPS score by some sort of weighing factor, but that you use a >1 weight factor for heart rates above a certain threshold. It functions like the ^4 in TSS.

Sorry, you're right. For some reason I read "exercise-dependent weighting factor."

 

[normZurawski]

I found this post by Coggan:
http://lists.topica.com/lists/wattage/read/message.html?sort=&mid=907028483&start=1

Which says this:

Typical IF values for different events or training sessions:

<0.75 level 1 recovery rides
0.75-0.85 level 2 endurance training sessions
0.85-0.95 level 3 tempo rides, aerobic and anaerobic interval workouts
(work and rest periods combined), longer (>2.5 h) road races
0.95-1.05 level 4 intervals, shorter (<2.5 h) road races, criteriums,
circuit races, 40k TT (by definition)
1.05-1.15 shorter (e.g., 15 km) TTs, track points race
1.15 prologue TT, track pursuit, track miss-and-out

For those people who use TSS, are these details you calculate or is it all done with the software?

SloarEnergy,

Do you actively use TRIMPS or is that just something you know how to calculate?

 

[normZurawski]

For males, TRIMP = training duration x [(HRavg - HRrest)/(HRmax - HRrest)] x 0.64exp[(HRavg - HRrest)/(HRmax - HRrest) x 1.92], where HRmax is maximal HR and HRrest is resting HR.

To clarify, does that 1.92 multiply the denominator (max-rest) or the result of the fraction (avg-rest)/(max-rest)?

 

[frenchyge]

Typical IF values for different events or training sessions:

<0.75 level 1 recovery rides
0.75-0.85 level 2 endurance training sessions
0.85-0.95 level 3 tempo rides, aerobic and anaerobic interval workouts
(work and rest periods combined), longer (>2.5 h) road races
0.95-1.05 level 4 intervals, shorter (<2.5 h) road races, criteriums,
circuit races, 40k TT (by definition)
1.05-1.15 shorter (e.g., 15 km) TTs, track points race
1.15 prologue TT, track pursuit, track miss-and-out

For those people who use TSS, are these details you calculate or is it all done with the software?

The software calculates it all, but I use the above guidelines to estimate the TSS for a future workout. The TSS formula mathematically simplifies to IF^2 x 100 x duration in hours, so from the above guidelines I estimate:

L2: IF~.8, 65 TSS/hr
L3: IF~.9, 80 TSS/hr
L4: IF~.95 (incl. rest periods for interval workout), 90 TSS/hr
L5: IF~1.0 (incl. rest periods for interval workout), 100 TSS/hr

 

[normZurawski]

Thanks frenchyge. How do you figure L6? Roughly the same as L5? Or is it too short to merit its own classification?

I'm trying to figure out how (if) to measure some metric next year. I'm holding out on the PM for 2 reasons. First, next year will be my first fully structured season and I want to get that experience under my belt before I buy a PM. Secondly, I'm holding my breath that the MicroTech Revolution is worth the investment, since I'll be able to use that on any of the bikes I have.

I'm considering using a generic TSS baseline like you present and estimating everything as a rough guide to complement my plan. The more I think about it the less I want to strap the HRM on every time I go out there. Plus, as you well know, HR response is just not the same from day to day, especially for someone who likes to block train, like me.

On the other hand, many of my rides are off-road, especially the long ones. That probably makes estimating TSS that much more difficult.

Your thoughts are appreciated.

 

[frenchyge]

Thanks frenchyge. How do you figure L6? Roughly the same as L5? Or is it too short to merit its own classification?

A little bit of both. Measuring L6 workouts in *hours* is too scary to even think about, and it's difficult by definition to produce IF>1.0 for an hour anyway.

Your thoughts are appreciated.

As you can see from the approximations above, TSS is more influenced by the number of hours you ride than the intensity of the ride, especially when you consider one's power v. duration curve. L5/6 rides are going to be less than 1 hour, L4 rides will be 1-2 hours, rides over 2 hrs are going to be L3 (max), and rides over 3 hours are probably going to be L2, unless it's a long race. Looking at the weekly training mix, it'd be possible to estimate a weekly IF just based on ride composition. Since we typically have more hours in the lower levels, the average weekly IF favors the lower values, and will probably be between .8 and .9 for most people over the entire season (except maybe when you're really trying to hammer out a peak before tapering). That means 65-80 TSS/hr is a good, rough weekly guide over the course of the whole season (actually, averaging IF=.9 for a week is some pretty tough riding, so 65-75 probably works during most of the year).

 

[SolarEnergy]

To clarify, does that 1.92 multiply the denominator (max-rest) or the result of the fraction (avg-rest)/(max-rest)?

Sorry NormZurafwski, I posted something pasted from a webpage, this one http://jap.physiology.org/cgi/content/full/95/4/1575#REF4

Though they don't seem to use the constant e in their calculation. I find it a bit strange. There must be something I'm missing.

But here, look at page 26 of this other one here http://multisport.qc.ca/app/DocRepo...ka_aterlier6.pdf#search="trimp 1.92 weighted"

I think it's presented in a more understandable way, and everything seems to be there

 

[frenchyge]

To clarify, does that 1.92 multiply the denominator (max-rest) or the result of the fraction (avg-rest)/(max-rest)?

As written, I'm pretty sure the 1.92 should multiply the entire fraction (ie, the numerator) rather than the denominator. Mathematically, the parentheticals should be resolved first, then, since multiplication and division are of equal precedence, work left to right.

Brought to you by Mr. Math.

 

[normZurawski]

A little bit of both. Measuring L6 workouts in *hours* is too scary to even think about, and it's difficult by definition to produce IF>1.0 for an hour anyway.

Aw, come on. I'm stupid enough to try! Ok, maybe not.

As you can see from the approximations above, TSS is more influenced by the number of hours you ride than the intensity of the ride, especially when you consider one's power v. duration curve. L5/6 rides are going to be less than 1 hour, L4 rides will be 1-2 hours, rides over 2 hrs are going to be L3 (max), and rides over 3 hours are probably going to be L2, unless it's a long race. Looking at the weekly training mix, it'd be possible to estimate a weekly IF just based on ride composition. Since we typically have more hours in the lower levels, the average weekly IF favors the lower values, and will probably be between .8 and .9 for most people over the entire season (except maybe when you're really trying to hammer out a peak before tapering). That means 65-80 TSS/hr is a good, rough weekly guide over the course of the whole season (actually, averaging IF=.9 for a week is some pretty tough riding, so 65-75 probably works during most of the year).

This is interesting. I'll probably opt for this system ultimately, but try to break it down a little more to start just to get a feel for what the numbers look like. I guess the key ingredient is the mix, which is something I don't quite have worked out yet. Or maybe I do, I'm not sure.

Another question. You note that the TSS includes rest between intervals. I assume this does not include warmup and cooldown?

Another thing is working out how to apply this to your (read: my) standard MTB ride. While the downhills are zero pedalling, I'm often just as tired on the fast descents as the climbs. But then again, I'm not sure that qualifies as "training stress" in the way TSS aims to quantify the workout. Maybe I'll need to experiment a little with the HRM to get some rough feel for how the scores cross-over and go from there. My best guess is somewhere between L2 and L3 when all is said and done.

In (roughly) calculating my highest TSS over the past few weeks, I'm shocked by paltry my numbers are as compared to some people. On the other hand, I've probably got plenty of room to improve going into next year.

Solar,

Thanks for the links!

Mr Math,

My bad. Various high school and college professors would be disgusted with me.

 

[acoggan]

As you can see from the approximations above, TSS is more influenced by the number of hours you ride than the intensity of the ride, especially when you consider one's power v. duration curve.

Sorry, but this assertion is incorrect. First, TSS = duration (h) x intensity factor^2 x 100. Second, I can send you numerous files with high TSS values generated via strictly level 5+ training (indeed, my wife's CTL when training almost entirely this way was ~120 TSS/d).

 

[acoggan]

Since we typically have more hours in the lower levels, the average weekly IF favors the lower values, and will probably be between .8 and .9 for most people over the entire season

For people who ride outside all, or at least most, of the time, it's usually closer to 0.75.

 

[normZurawski]

For people who ride outside all, or at least most, of the time, it's usually closer to 0.75.

For lack of a better word, .75 seems atrocious. So the average IF for outdoors rides is less than an L2 ride?

 

[acoggan]

For lack of a better word, .75 seems atrocious. So the average IF for outdoors rides is less than an L2 ride?

By the time you include recovery rides, warm up, cool down, chatting w/ friends, any group ride that isn't a real hammerfest, etc., etc., etc., the answer is "yes". (And now you know why, for three decades, I've chosen to largely train alone, and often on an ergometer...I've simply never had the free time to be so inefficient.)

 

[normZurawski]

By the time you include recovery rides, warm up, cool down, chatting w/ friends, any group ride that isn't a real hammerfest, etc., etc., etc., the answer is "yes". (And now you know why, for three decades, I've chosen to largely train alone, and often on an ergometer...I've simply never had the free time to be so inefficient.)

Ok, makes sense when you put it that way. This is also why I train alone. A 2:00 off-road ride becomes a 2:40 ride when I go with people. I guess this gets at the heart of the idea that every ride should have a purpose.

 

[frenchyge]

For people who ride outside all, or at least most, of the time, it's usually closer to 0.75.

You certainly have access to a greater sample size than I do, but last night I checked my monthly IF values over the last season. My lowest was .80-something, and my highest was .88-something. I'm counting the season from last Oct. through the present, with predominantly trainer riding from Oct - Mar, and predominantly outside riding from Apr - Present.

Admittedly, I'm not a fan of just riding around. I like to go fast.

 

[frenchyge]

Sorry, but this assertion is incorrect. First, TSS = duration (h) x intensity factor^2 x 100. Second, I can send you numerous files with high TSS values generated via strictly level 5+ training (indeed, my wife's CTL when training almost entirely this way was ~120 TSS/d).

I know what you're saying, and I think you know what I'm saying. As we've been discussing, over the course of a week or more, IF values tend to fall in a pretty narrow range, and that range is low enough that it's possible to log any number of hours up to a person's time limitations. In that sense, I think my statement is fairly representative of the situation for most people, but I don't disagree with what you said.