An alternative arbitrary work (AU) calculation

Other Discussions Training Discussion
1

A while back (BBM podcast 39), I got interested in monitoring my training workload for the purpose (among other things) of managing fatigue and reducing the possibility of injury. Most of the resources I’ve seen use training session time (e.g. in minutes) to help calculate the workload as “arbitrary units” (AU) for the purposes of recording session, acute (~weekly) and chronic (~monthly) workload, which can then be used to calculate the acute-chronic workload ratio (ACWR) for an athlete.

For a couple of reasons, I didn’t find session length time to be a useful metric to record (my sessions are sometimes interrupted or truncated, split over the day, or start/stop times are unclear). I sometimes tear through training sessions, and I sometimes take my time. After some thought and experimentation, I came up with a different method of calculation for session workload/AU:

  1. Workload = (Session total tonnage / 1000) x Session average RPE x Session RPE I feel like this calculation includes a reasonable mix of objective measures (session total tonnage) and subjective measures (average RPE for all sets, as well as session RPE). Including session RPE allows for situations like training feeling especially hard due to short rest times or lack of sleep. Session tonnage is trivial to measure, and using an average of the RPE of all sets as well as session RPE helps smooth out situations like peaking periods (where average RPE might be high but session RPE is often low).

I’ll be the first to admit that this measurement is more for academic interest than practical at present, as I’ve only used it to influence training a couple of times (when work has spiked quickly). However, it’s been interesting to chart my workload and correlate ACWR to my current “feeling” of fatigue.

Does anyone else track and use session RPE and/or workload in training? Alternatively, anyone read the algorithm above and note any obvious flaws or improvements?

A couple of references, for those interested:

  1. BBM podcast 39: https://www.youtube.com/watch?v=V43mSQEjZY8
  2. Acute:Chronic Workload Ratio
  3. https://bjsm.bmj.com/content/bjsports/early/2019/02/21/bjsports-2018-099422.full.pdf
  4. Training Load and Fatigue Marker Associations with Injury and Illness: A Systematic Review of Longitudinal Studies - PMC
2

I’ve thought about this for a few months now and haven’t come up with anything good. It seems difficult to reliable mix objective and subjective measurements. I’m of the opinion for now that only subjective feelings of fatigue are important. But I agree there are some issues with using session time, since I’m not always consistent with rest times, times between different lifts, going to the bathroom, getting interrupted, etc.

My $0.02:

Using tonnage would bias towards high rep / volume work as being much more fatiguing. Doing 5 sets of 10@8 would produce a way bigger tonnage than 5 sets of 5@8, and assuming both produce a similar sRPE, the 5x10 day would cause a big spike in ACWR, but I’m not sure if a 5x10 is actually more fatiguing.
I thought about using average intensity, but I think then it biases low rep strength work too much as being much more fatiguing than volume work. But I don’t know if that’s true.

session RPE helps smooth out situations like peaking periods (where average RPE might be high but session RPE is often low)

Interestingly, I find the exact opposite to be true for me. I can do a session of all 8s, 10s, 12s at RPE 8-9 and rate it a low-average sRPE, but asked me to do a 1@8 and a 4@9, and my sRPE is quite high. I guess I just subjectively perceive high intensity work as more fatiguing. Which relates to my first point: what’s more fatiguing, a 5x5@8 or a 5x10@8? I’m not sure, and I think it could possibly be individual to the lifter. Or maybe my subjective feelings towards high intensity don’t matter, and it does produce less fatigue given volume is modulated (hence, a ‘peak’) :slight_smile:

I am starting to record my session average RPE, to see if it at all correlates with my subjective feelings of fatigue.

What I think I will start recording is just a subjective “how fatigued do I feel today? How’s my motivation to train? What’s my attitude towards training today? My general well being?” lumped into a simple 1-10 scale, and see if there’s any correlation to ACWR, sRPE, average session RPE, volume, average intensity, tonnage, etc …

1 Like
3

I thought the same thing - either something can be objectively measured or only subjectively. And mixing them togeather can produce either unreliable or simply uninformative results. As far as I understand, the original method for ACWR was studied on soccer players who spend a lot of time in general in training. I think for such professional athletes, when we are talking about hours spent training, maybe this metric can identify sharp spikes and unsustainable increase in workload, but for an average gym goer? I am not sure. My sessions are almost always 70-90 min and that length depends not only on how focused I am, but what kind of GPP I tack on to the end of that session.
I liked the podcast where this was discussed, but my takeaway from it was to only track sRPE and make sure I am not in 9-10 for more than one or two sessions in a row.

2 Likes
4

Good points by all. I mostly noodle around with this stuff out of interest, and wanting to make something useful out of the couple of years-worth of structured training data I’ve collected.

For myself, I find a set of 8 @ RPE 8 much more fatiguing than a set of 3 @ RPE 8. Heavy sets of 1-3 are finished so fast that I just don’t get that burning-muscle out-of-breath feeling that a set of 8-10 delivers. I think that’s why I tend to rate sRPE for peaking so much lower than for strength and (especially) hypertrophy blocks. 530 tempo squats for sets of 10, anyone? Therefore, I tend to rate sRPE somewhat higher for high-volume training.

I messed around with increasingly-complex algorithms to see if I could reduce the bias of higher tonnages (e.g. trying to derive an AU value for each set) but gave it up in the end as trying to be too cute.

3 Likes
5

For me it is most important to consider how many sets are in the @8-10 range (RPE not reps), and if it’s a competition lift or not. I would consider just adding the RPEs together and weighting the competition lifts by some arbitrary factor.

So if, for example, if I SQ x1@8 x4@9 3x4@8 I would add (8+9+8+8+8)*1.5 = 61.5 AU. But then if later I do CGB 3x8@8 I would add (8+8+8) = 24AU. The 1.5 is the arbitrary factor representing how a competition lift is more physically demanding than an accessory lift.

Just thinking out loud here. Maybe the multiplication factor can be adjusted based on how you are feeling that day.

1 Like
6

There is a fairly good argument to be made for only summing RPEs, as you suggest. Those high-intensity sets contribute a disproportionate amount of fatigue in a session. I find a single set of @9 more immediately draining than multiple sets @7, as I think many people do. I don’t personally expect to be able to perform more than two sets of @9 across without reducing weight on the bar.

That said, I don’t usually include any “accessory” movements or GPP when I calculate AU (only the competition lifts or close variations). Training mostly at home, I tend to do accessory work on rest days. I think that you could go either way (including accessories or not) and it probably wouldn’t matter as long as you were consistent about it.