Here are two related stories based on Whoop, knowing my readership I’m hoping that at least one of them will be of interest to most people.
Related: Whoop 4.0 Strap Review
Whoop & Broadcasting HR
The first two versions of Whoop couldn’t broadcast HR using any standard method. That’s changed with the v3 released earlier this year as Whoop 3 now broadcasts over BLE. Yay. As if proof were needed, here you go…
That’s my second 945 in the image, above. The first was the subject of ‘Buttongate’. I thought I was OK at first and was delighted that I didn’t suffer from the rather prevalent button-pressing issue that everyone else seemed to be having. But eventually, mine gave up too 🙁 and here’s the replacement. Unfortunately, I reckon that even with this new one (it’s reconditioned) every button presses subtly different to the other 4…but at least they ‘press’.
Whoop hasn’t got any buttons to press, which is probably a good thing.
So, you can now pair Whoop to some gym machines and to many Android/iOS apps; or, of course, to any GPS watch or bike computer that supports BLE.
This somewhat opens up Whoop Review to ‘be tested’. When I did my original review, Whoop was very kind and provided me with my HR stats from their own backend system. I’m reasonably sure that no other reviewers had that data at the time. Well, now we all do!
Of course, Whoop know that wrist-based HRMs are subject to the vagaries of motion artefacts during exercise and consequently they also sell an “Impact Series Upper Arm Sleeve” and an “Impact Series Full Arm Sleeve” to encourage you to wear it elsewhere at key times. I’m not so sure how many Whoop users use those arm sleeves (where I would expect GREAT results). So let’s see what happens with the wrist readings.
I’m planning to do some tests on Whoop’s HR accuracy. I should be able to get MANY more data points than before. Let’s hope it’s good.
BEDTIME READING Whoop Review
Whoop & Alcohol
Source: Whoop 43 Podcast and an email they sent around today
NO, it’s not a new kind of drink 😉
Next up is a piece of research by Whoop about the effects of alcohol and the time taken to recover from it. It’s an old piece but it’s doing the rounds again.
We all intuitively know that alcohol is not great for athletic performance mainly because everyone else says so. Perhaps only a few of us have tried to quantify that? I’ve not looked in any detail at the effects of alcohol on me as I don’t drink that much. However, I have noticed very poor overnight HRV readings after drinking but my subsequent night’s sleep always ‘seemed’ OK. And I left it at that.
Whoop did some research with their users who logged alcohol ‘events’ alongside sleep and workouts (as usual). In a nutshell, the finding was that alcohol affects us all differently but that effect can last for up to 5 days in some people. Here’s a bit more detail:
- 1-day post alcohol consumption, 74% of athletes showed reduced recovery metrics
- 2 days post alcohol consumption, 29% of athletes still showed reduced recovery metrics
- 3 days post alcohol consumption, 19% of athletes still showed reduced recovery metrics
- 4 and 5 days post alcohol consumption, 7% of athletes still showed reduced recovery metrics
Whoop’s 2015 study also found that, from a population of 148 athletes, 16.2% had higher rHR and 22.7% had lower HRV due to the after-effects of alcohol.
Must Read: Check out the latest WHOOP Review
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.
Shame they didn’t make it dual BLE and ANT+. That would have made it a reasonably compelling upgrade.
Be interested to see what you find on HR accuracy. Mine has been woeful over the years I have had it. Any slight HR spike, gives you 4.5 strain for the day e.g. getting out of bed.
Re alcohol, I’ve found that I record my best Whoop recovery metrics when in fact I’ve been terribly hungover and almost unable to move. I think their algorithm prioritizes sleep to generate their recovery score over HRV, so if it thinks you slept a lot, then you get a green score.
There are lots of bugs with their algorithms e.g. if you shorten your sleep (because it inaccurately recorded it) it will actually tell you that you had longer deep sleep than previously. ?
I like what they are trying to do, but I’ve so many reservations about the quality of the data, that the recommendations are meaningless.
yes i think some of what you say applies to many companies in this space.
i dont think ant+ is that relavent for them. I suspect MANY users are app-based.
those that are also garmin-based will likely (on the whole) have dual-band recieving watches.
i’ll look at what you say about how the app prioritises the inputs when alcohol is involved…i’m looking forward to testing the effects of alcohol 😉
Interestingly, never happened to me. Heavy drinking at night, and I’m waking up to “red recovery”. Has been proven time and again.
But then, see below… WHOOP uses primarily HRV for your recovery and measures it during what they perceive as deep sleep, more specifically slow wave of N2/N3 phases — their model for when to measure HRV is more continuous than simply light or deep. Still, if your stages are off, perhaps due to sensor placement, they might pick up a period when you have an elevated HRV. I think Polar’s approach with Ignite and taking an average of the first 4 hours is better, but I’m sure they all have their reasons.
Connect your Whoop account to Training Peaks and you’ll see how totally random the Recovery Score actually is.
I’ve been using WHOOP for more than 18 months, fist 2.0 and now 3.0, which did go through some early growing pains with respect to HR broadcasting, but now is rock solid.
Re: HR accuracy. Your mileage will vary. That being said, for me it is as accurate as Garmin/Polar watches on wrist and most of the time as good as Polar OH1 on biceps (in fact, sometimes I wonder if it is WHOOP and not OH1 that is correct). And yes, I have both the older style biceps band and a new impact sleeve, which is terrific. An absolute must purchase! One good thing about the sleeve is how easy it is to put WHOOP sensor in and out and put the whole thing on your shoulder. It is also totally secure there, so if you are involved into any impact sports (doh!), you don’t have to worry about the sensor moving around.
Yeah. I actually resorted to wearing it permanently on the bicep as that was the only way to get reliable HR. But that introduces a different set of issues:
1/ it frequently thinks you are asleep when in fact you are only reading in bed (as I guess it relies on the accelerometer to determine whether you are sleeping.
2/ i guess it over-estimates recovery because it is frequently wrong about whether you are awake or not
3/ i would also suspect that the sleep stages guesses are less accurate because it probably relies somewhat on accelerometer movements to determine what stage of sleep you might be in
4/ charging is less convenient
5/ i seem to get more instances of cadence lock on my bicep than i ever did when i tried to wear it on my wrist. obviously totally unscientific, but just a feeling gathered over time
again, i like what they are doing, just don’t feel the accuracy is there yet to sustantiate the recommendations.
*EVERY* vendors’ sleep stages are guesses.
interesting about the cadence lock on bicep. i’ll watch out for that
Re: cadence lock. Interestingly enough, for me it’s a new “feature” I’m only seeing in 3.0 and latest firmware versions. Don’t remember it ever doing this to me with 2.0 band. Still, not as bad as Suunto and Valencell and strangely, worse with steady walking than running.
Personally, I don’t think heart rate during a day they capture is that important, which is why I go with wrist. Even if you lead a fairly hectic life, non-workout daily contribution to strain appears to be minimal. Since WHOOP only uses HR to calculate daily strain, there’s little need for extra accuracy, come to think about it. It’s not like they are trying to derive HRV to analyze stress — this only happens during sleep. So, overall they should come to fairly decent number as far as strain is concerned. Regardless, it’s all about trends and baselines, not specific number here and now.
My sleep analysis looks spot on (maybe because I’m wearing it on wrist, and they optimize for that). In fact, I’d say WHOOP beats everything I tried before in terms of tracking when I’m awake, when I fall asleep, etc. Their sleep stages breakdown is roughly in line with Oura and what I feel, i.e. if I sometimes wake up from a vivid dream, WHOOP would often show REM was there right before I woke up. Ditto for naps.
Also, WHOOP say recovery model is based primarily on HRV and HR, with sleep duration playing a minor role and sleep stages playing no role at all (other than to determine a period for HRV measurement). Hence, if you trust their HR, you can trust their recovery number.
There are other things I find questionable about their approach:
1) They use the HRV from the last period of continuous deep sleep, which means effectively they will be measuring it at different times of night… But apparently there’s science to back up that particular approach.
2) Any HRV increase is treated as a good thing. So, if my baseline values are 60-70, and my HRV suddenly jumps to 100+ during a period of heavy training (acute HRV change, probably due to impeding overtraining), they give me thumbs up and 100% recovery instead of telling to scale back.
3) You better make sure your max HR is set correctly in the app. They would try to determine it automatically, and the value they gave me was laughably incorrect. I don’t ever remember seeing my HR higher than 182-183, and they gave me something beyond 192. Needless to say, your strain will be calibrated to that ceiling.
Nick. Great comments / discussion.
How did you get them to change your maximum HR? I’ve done the V02 max test to exhaustion on treadmill with face mask etc and so have a more accurate number.
I actually moved Whoop to my bicep because I would routinely get 6 – 8 strain by just getting out of bed, doing the dishes, using my PC etc. I’ve seen so many posts on Twitter where the same thing is happening to other users.
So given the HR was inaccurate during the day, I didn’t have confidence that any of the data was accurate. If you can’t get the base data right then…..
Moving to my bicep gives me very clean HR mostly, with all the other downsides noted. But I agree with you entirely about the importance of looking at the baseline / trend over time.
The problem I think though about inaccurate data during the day, is that Whoop use a log scale for strain. So it becomes progressively harder to build strain during the day / exercise. So if you exercise at night, one day and already have 10 – 12 of meaningless strain, I’m not sure how that affects the strain calculations derived from your workout. I didn’t have the patience to analyze it / compare and it may be nothing, but I use a HR strap / sports watch to allow me to calibrate effort more accurately between sessions.
Re HRV, I do an Elite HRV and HRV4Training test every morning on waking for 5 mins with 30 seconds HR stabilization. As close to same protocol each morning as possible. I also do the Polar orthostatic test every morning afterwards. Routinely, the Whoop HRV measurement is the lowest of all of them – by maybe 20 points. Often the trends do mirror each other, with Whoop consistently being the lowest.
The other issue (not solely a Whoop issue) is that I can get 12 strain for a gentle 1 hour run – say 4:20 per km. But cycling flat out for 2 hours to the point where I can barely stand, only gets me 10 strain. I guess this is a problem where you have 1 very dominant sport and then another one where you don’t have the same muscular endurance.
Finally, I actually find it amusing they advertize golfers using Whoop when I can’t imagine any scenario where optical HR is accurate during golf swing. Again not a Whoop issue per se.
Majority of players on recovery field uses Time domain parameters of HRV (some include HR as well – EliteHRV, ithlete, …). Mysasy.com uses as one of the few, frequency domain LF, HF in combination with Time domain rMSSD. If you like to experiment with different thing its worth trying.
@the5krunner What is the revision (written on the FR945 box) of the replacement watches? I’m still waiting for a good revision to send my 945 for replacement. According to Garmin forum, revision D seems to have button problems resolved.
mine came with a clear lens protector and one (not two) clear circular sellotape stickers on the outside. so i think it is a refurb…not happy. the buttons all feel different to each other (but all work)
product number is 010-02063-01 …so presume 01 is the revision?
I am seeing more and more evidence of cadence lock over the past month and since the last firmware update. I only wear on the wrist. This can really skew the Strain numbers.