WHOOP – Ground Breaking Strength Training? Or Damp Squib?
WHOOP’s new strength training feature promises groundbreaking insights into muscular strain, never before seen in a wearable. Yet, whilst the new addition to WHOOP looks awesome and generally works well, it’s somewhat limited and perhaps even confusing in certain aspects.
Must Read: Detailed WHOOP Review – is it accurate?
What is it?
The underlying science is Velocity Based Training (VBT)
The WHOOP smartphone app lets you design a complete strength workout comprising multiple sets, steps, and supersets. The app guides you through each routine and gives you a new type of strain score at the end.
The WHOOP band still tracks your heart rate, but now, its onboard gyroscope and accelerometer are used to better quantify the true effect of each rep. Note that those two sensors combined are referred to as the Inertial Motion Unit or IMU.
So, just to be clear, this new ability only comes into play when following a known strength routine on the app.
Muscular Strain – Truly Scoring Strength Workouts?
Before this week, no other mainstream wearable, including WHOOP, properly scored your strength workouts. Even if you used a super-accurate chest strap to measure your heart rate, the heart rate simply does not give a proportionate insight into the true strain on your body. Sure, it gives some strain, but it does not account for the stressful impact on muscles, bones, joints, and tissues.
Polar’s power-based Training Load Pro calculations include an element of estimated muscular strain, Garmin & Coros both produce muscle heatmaps for strength workouts and Biostrap attempts to auto-recognise strength moves with internal sensors. None are quite the same as WHOOP.
To assess the true impact WHOOP gather a variety of bits of information
- Movement profiles of different exercises vary and affect different muscle groups in different ways. WHOOP determines the relationship between the musculoskeletal system and each exercise profile
- Weight – In your set, you will define the weight used
- Reps – WHOOP measures these where it can
- Intensity – The speed of movement (velocity) is also collected.
Faster reps, more reps and heavier reps each increase the muscular strain.
Thus, the total physical work done for each rep is determined and allocated to the appropriate muscle groups.
Some secret sauce happens here, and WHOOP determines both the normal cardiovascular strain and the new muscular strain.
What appears not to be collected right now is the range of movement from one rep to the next, nor the quality of each rep in terms of its idealised 3D motion.
Remember, there are limitations. If you are wearing WHOOP on your left wrist and doing right-arm biceps curls, then WHOOP isn’t going to record a whole lot of motion data that describes the curl.
What to expect: I’m definitely not the most muscular of people. I see a low muscular strain for the weights I do in WHOOP. Like 3% of the total strain. If you are an Olympic weightlifter, then your muscular score will be significantly higher, but I don’t know if the magnitude difference you will see will be 33% or 93% for muscular strain.
What You Need to Do – Building a Workout
WHOOP’s graphic sums up the process you go through on the app pretty well. As you can see, it’s relatively intuitive to use.
There are some nice graphics as you execute the workout showing your heart rate and stats after you’ve finished showing the strain breakdown and also how your heart rate track maps onto the times when you executed individual sets.
The newly modified WHOOP strain metrics are also accounted for elsewhere in the WHOOP ecosystem for example it feeds into sleep recommendations.
But is it any good?
For a first pass, this is an okay feature from WHOOP. However, considering that many WHOOP users are gym-goers, lifters and cross-fitters, WHOOP definitely needs to go further.
As a somewhat weedy endurance athlete (!), here are my thoughts:
The whole process looks nice, and there is an extensive workout library to build from with good imagery to help you identify the right name for the workout you plan to do. However, choosing an individual workout to add to your set is quite a manual process. I couldn’t see how to either assign a favourite workout or copy a set to repeat it.
Executing the workout required repeated and numerous interactions with the smartphone app. WHOOP would definitely benefit from the ability to recognize a tap gesture to end and/or advance between reps. As it is now, I was putting my smartphone to one side, lifting something, and then returning. WHOOP presumably was detecting the reps but didn’t seem to recognize my movement (or lack of it) as a stop or start to the set.
I only did fairly simple sets, and WHOOP was okay there. I could easily imagine that complex sets would be cumbersome to both build and annoying to execute.
Then, what does the end number mean? Elsewhere I’ve seen some whoop users bemoaning the old, maximal 21 strain score. “Why couldn’t it be 20 or 100?” was the criticism. Fair enough but at least we knew what 21 means. What does a new 21 mean? Then what is the relevance of a strain score for a bench press plus the strain score for squats? Is it really an additive measure? Then is there any meaning to changed strain scores over time? Does WHOOP track the range of motion and velocity of each of your exercises?
Finally, the science behind VBT requires, I believe, that each rep is executed at a maximal velocity. Jo Public simply won’t do that. Does that mean that the science then doesn’t apply?
Don’t Forget
Whilst heart rate alone can never quantify muscular strain, WHOOP’s HRV absolutely did and does correctly measure your body’s reaction to muscular strain.
Take Out
WHOOP claims to have extensively validated its muscular strain results, and we have to take it at its word. When we factor in its acquisition of PUSH, the company has presumably expended significant financial resources, so we can only hope the data capture and calculations have good technical and scientific foundations. There is no way to easily test this.
WHOOP’s novel intentions for muscular strain seem groundbreaking and offer lifters, cross-fitters, and the like the main hope of getting the most accurate assessment of their strain.
However, the new section of the app is very much a Gen 1 product that needs ease-of-use refinements for workout building and execution. I suspect the former is already in progress; however, to better automate workouts without having to continuously interact with the smartphone app might be technically challenging for WHOOP. Time will tell.
Thanks to: Daniel Haenle Admin of WHOOP Athletes Group
Reader-Powered Content
This content is not sponsored. It’s mostly me behind the labour of love which is this site and I appreciate everyone who follows, subscribes or Buys Me A Coffee ❤️ Alternatively please buy the reviewed product from my partners. Thank you! FTC: Affiliate Disclosure: Links pay commission. As an Amazon Associate, I earn from qualifying purchases.
WHOOP doesn’t detect reps at the moment me thinks, not like Garmin, and maybe it’s a good thing. The flow appears to be exceedingly simple: you tap “start set”, go do your thing, then tap “end set”. WHOOP copies planned set weight and reps, which you can modify, and kicks off a rest timer. Rinse repeat until you are done. If there’s any counting done, I personally didn’t notice it, and I was doing quite explosive kettlebell swings, snatches, and squats.
Given how bad Garmin and COROS are at counting reps most of the time, years after these features have been introduced, I’d rather do the counting myself. In fact, since you enter planned rep count when you setup a workout, as long as you hit your numbers, no modifications is necessary. Pretty streamlined! Far less intrusive approach than that rep count confirmation screen Garmin pops up after every set, but which is missing when sets are back to back.
What is needed though is for WHOOP to switch from the obligatory outside wrist location for the sensor. Because they still track heart rate, and as expected, it is thoroughly wanting. I find WHOOP to be off by as much as 4-6 bpm for strength training even with a bicep band, but this new placement pushes things beyond 10-12 bpm, making heart rate tracking mostly a random number. The only hope is the heart rate catches up somewhat during rest periods.
Wrist is also not ideal for any kettlebell exercise involving a flipping kettlebell. Apparently, great minds at WHOOP missed a great deal of kettlebell exercises involve a 50-70 lbs steel ball flying over your hand and slamming into the arm. Maybe not slamming, ok. But there will be enough impact to destroy the sensor eventually. For now the only workaround is to move the sensor to non-exercising hand, which somehow works. So, if an exercising hand isn’t really required, why demand outside wrist placement? How about inside? Better yet bicep?
Still, personally for me this update is the Holy Grail of strain tracking. I shed pious tears every night and am getting ready to sacrifice small furry animals at the altar of WHOOP Machine Learning gods. The workouts which previously tracked with 6.0-9.0 strain, which I knew to be a bad joke, are now 12.0+. The score finally reflects how I feel.
Now, do I need a black puck to tell me how I feel? No. How much recovery I need afterwards? Not really. But having a Big Data Daddy watch you over the shoulder to curb some of the exuberant enthusiasm is valuable. WHOOP can do that. And they are the only game in town that meshes traditional aerobic exercise and muscle bound loads.
hey there
they must count reps in some fashion otherwise this vbt stuff wont work. whether or not they report it is a different matter. I guess that would expose them to claims that ‘reps are wrong’. For example with Garmin, it can detect raising the weight to position and putting it back as reps, rather than preparing for reps.
self counting/counting to plan – yes, that’s fine most of the time. however, again, VBT requires reps to exhaustion and the exhaustion point would be unknown in terms of reps.
I’m not saying you’re wrong here, just questioning (as i did in the article) about the use and validity of vbt in this context
placement – yes, it’s a dilemma. idk the solution there. the wrist is a good place to measure movements but a rubbish place to measure hr.
score of 12+ … just wondering what you see as the strain breakdown? can you share a pic?
They use data from accelerometer/gyro to see how “fast” and “explosive” your “movements” have been. Your reps determine your tonnage and speed adds a factor, both to adjust for muscular strain.
If there’s any rep counting, it is well hidden in the bowels of the algo.
I suppose we can do a simple experiment: do same set of exercises on 3 different days, but use lighter weights on one day and cheat on rep count the other and see what difference in strain you’d get 😉
i’m intrigued to know…especially after discussing in some detail with Daniel from the Whoop group (not an employee)
‘fast’ requires a consideration of distance and that comes from, at least in part, reps
I’m not saying they are not trying to quantify your movement. But I’m not sure that quantification is on the level of movement recognition and rep counting.
For example, their instruction video for kettlebell snatches (raise kettlebell over your head in explosive movement, flip, fix position) clearly shows a lady wearing WHOOP on her non-exercise hand! Well, she won’t be able to have it on her outside wrist otherwise. Needless to say, that sensor hand won’t be moving nearly as fast, if at all. You won’t be able to get much rep counting from that position. I’m not sure you can measure explosiveness of the snatch from non-exercising hand either to be honest.
The more I think about it, the more I want to try to cheat it and see if WHOOP would be able to call it. But then if it doesn’t, what does it mean for me? I just got something that “quantifies” muscular strength 🙂