Some activities are harder than TSS would suggest they are, too. Gym sessions where heart rate remains low for the relative amount of muscle damage, for example, are hard to provide values for…
When I look at the context of that statement, it is in the 3rd of 7 paragraphs talking about the pitfalls of using TSS, PMC, etc. The comment is based on the fact TrainingPeaks uses HR to estimate TSS in the gym. We could say the same about using HR for TSS in cycling, it can be hit or miss.
TSS was defined as a power based metric, taking into account both intensity and duration. Would you ignore the power meter on your bike, and use HR for TSS just because TrainingPeaks allows it?
The question to ask – why does TrainingPeaks ignore intensity/duration when calculating TSS in the gym? The whole point of TSS is to take intensity and duration into account, and in the gym this is sets, reps, and weight as % of 1RM. Having strength training TSS based on HR is just adding a feature to claim support, despite the fact its somewhat arbitrary and perhaps meaningless support of strength training.
“Intensity Number of Lifts” (INOL) is taking the TSS concept of wrapping intensity/duration into a number, and applying it to weight lifting.
1 Like
This…INOL is a pretty good gauge of stress from strength training. The problem that I’ve found with it is that it doesn’t differentiate between 5 sets of 2 with 90% and 2 sets of 5 with 90%.
I have taken an approach where I set a max tss achievable per set (arbitrarily set at 10), then estimate the percent of a rep max at that percent of my 1rm. I then treat this like an IF.
So a set of 5 at 80% 1rm is 80/85 = .94IF, given that a 5rm is roughly estimated at 85% of 1rm.
Then tss for a set is just 10 * IF * IF.
The insight for treating the set as the atomic unit of stress came from the stuff that Greg Nuckols has written, which says that “number of hard sets” seems to be the most predictive of hypertrophy and strength outcomes.
So do you include this in your TR calender/TSS data?
I don’t include it in the TR calendar, as it doesn’t do a good job differentiating between different types of tss.
I’m still using training peaks for that reason…I can see a strength or cycling (or rowing) specific tss / week and ctl calculation.
Hopefully that’s coming to the TR calender soon 
so, attatch a stryd powermeter to the bar? 
I use a fitbit & set it weight training & do the workout. At the end of the session it gets uploaded to TraingPeaks & hTSS is calculated in & around 50 generally. However when the exact same workout file is uploaded to Todays Plan is comes up at hTSS 30ish.
I go with the Todays Plan number as it seems a little more realistic given that TSS 50 is an hour of endurance riding as a comparison.
I think matching cycling TSS & Strengh Training TSS is like comparing apples & oranges but the fitbit / Todays Plan route seems as close as it gets to being somewhat realistic at the moment I think.
Shane 
Similar. 30 minutes, 3 lower body force movements, 3 sets each, 4-10 reps. 30tss. I adjust 1 tss per strength minute if sessions are maintenance (15minutes, 15 tss), or longer.
This is consistent with the cycling TSS calculation if you run the numbers:
TSS (bike)= IF x IF x Time (hrs) x100.
Keeping track of the fatigue from Strength training is a great idea. However like all training responses, it is multifactor with all the usual suspects like training history, intensity, volume, type of lifts (e.g. dead lifts vs curls), nutrition and talent in play. I gave up on trying to put specific numbers on it relative to cycling and just try to align my light, medium and heavy strength days so as not to impact my TR workouts and pay attention to overall fatigue.
With that your model you would have a TSS of 8.836 with just one set of 5 reps at 80% intensity. That’s WAY to generous.
Per INOL, anything over 2 is “Brutal”, the same way in TSS, anything over 100 (for an hour) is brutal…
The approach I take is to calculate all of the INOL scores for all exercises in my workout and sum them together. Then take that score and divide by the max INOL that could have been achieved.
So in your example, if my entire workout was just that 1 set of 5 reps at 80% of my 1RM, my INOL would be .3 with the following formula to calculate INOL:
(sets * reps)/(100-intensity) or (1*5)/(100-80) = .3
With a max possible INOL of 2 per exercise, (and in this example the entire workout since there is only 1 exercise), my intensity of the entire workout would be 0.3 / 2 = .15 (i.e 15%)
So then to convert to the TSS scale, I just take this 15% and multiply by the 100 TSS per hour scale.
So if my workout was 5 minutes long, the MAX tss i would be able to achieve is 8.3 (i.e. 100 tss / 60 mins to get the per minute max tss that could be achieved, then multiply by 5 since my workout was 5 mins).
However, since my effort was only at 15% (per the INOL score) we then apply this to the max tss that could have been achieved… (.15 * 8.3)… Thus your TSS for the workout would be 1.25
Sine we are only talking about 1 set here, I find this is a much closer estimation, however this model assumes you aren’t just sitting around during the workout because time has a clear effect on the scores.
Would love to hear everyone’s feedback as this is the best I’ve been able to get it and I am by no means a mathematician!
My issue with the INOL formula is that it doesn’t take into account the set x rep scheme used. 10x2 is very different than 2x10 at the same weight. They clearly don’t produce the same stress. This is why I largely abandoned INOL
Certainly 11 sets of 5 squats at 80% of true 1rm (totaling around 100 TSS) would be brutal…to the point of probably not being possible. At least for me, it elicits the same feelings of dread and doubt as an hour at ftp does.
I think it depends greatly on the exercise in question though, the more muscle groups, the greater the stress in my experience. I’ve taken to normalizing by my bodyweight, so squats and deadlifts (above my bodyweight) produce more TSS per set than e.g. press.