Whoop / Garmin / Polar Comparison
A not very scientific comparison
Kind of a follow up to this and this. Basically I’ve gone back to Whoop and wanted to see how the new hardware / software compared to last time. Hopefully accuracy has improved a bit!
As ever, this won’t be exhaustive testing in controlled environments — I’m just going to make a one-off comparison across a few areas and see how it works for me. That last bit is important as results may vary etc…
Test 1: Stationary bike
Les Mills Sprint #24 if you’re interested. Whoop left wrist, Garmin right wrist, Polar left arm. Last time out the Whoop struggled badly with this one.
Looking at those, the only real issue is at the beginning where the Garmin takes a little time to catch-up then has that weird spike before falling in line with the other two. But that’s a massive improvement for the Whoop over the last time I tried this.
Calorie burn estimation was also interesting. Polar, Whoop, Garmin reported 400, 348, 397 respectively for this workout. Previously, for a similar 30 minute spin session they reported 377, 270, 357. Obviously there are a number of factors here but one consistent observation with Whoop was it’s under-reporting of calories. It’s clearly still at the low end of these three but does seem inline with what I’d expect for this type of workout for me.
Test 2: Walk
Brisk 5k walk, bit of a run up the hill towards the end.
Whoop was back to it’s old tricks here, with way higher HR for the most part and generally just missing the point here. The Garmin, while not perfect, did track the shape of the Polar pretty well.
Definitely some work to do to make sure the Whoop is reading accurate here. I’ll try adjusting on the wrist initially (higher & tighter is normally the advice they give) but it might be the arm-band gets more use here — thankfully they’ve made the connection quick release now — perhaps for this very reason.
Test 3: Hill Sprints, More Walking
Similar walk to yesterday, but interspersed with a short set of hill sprints.
Quite why walking seems to cause so much trouble for these devices I cannot understand. The Polar never registered above 103bpm yet the Whoop and Garmin hit the 120 mark on a couple of (different!) occasions.
For the actual hill sprints:
For whatever reason, these all seemed like a pretty good representation — you’d think the rapid changes would be harder to track than a consistent walk but obviously not. Maybe the Garmin tracks everything slightly lower but they all give the general feel of things. Unsurprisingly, Garmin estimates a higher calorie burn (158) than Whoop (113) despite the lower tracked heart-rate.
Last one is the walk back — all up-hill with one last hill-sprint at the end so seemed worthwhile to include this.
Clearly going for a walk is a struggle for these devices! The Whoop looks pretty good for the most part but then overshoots the initial peak and fails to come back down after the 2nd one. Garmin seem to not only overshoot the initial peak but uses advanced AI to pre-empt it before it even happens. Or it just got it wrong. Catches the descending bit after the second peak OK though.
Test 4: Strength
The main reason I went back to Whoop was for it’s new strength training feature so I obviously had to test this one. Basically it allocates you ‘strain’ (Whoop’s measure of physical effort essentially) for muscular exertion and not just cardiovascular effort.
None of these are miles away — Garmin seems to overshoot things a bit near the beginning (seated rows and shrugs I think) but isn’t miles away. Calorie estimate seems high (319) while Whoop is probably on the low side (151). Polar has me at 252 which just seems, you know — accurate.
Test 5: Sleep
Think sleep tracking is maybe the original USP for Whoop so feels necessary to compare against the Garmin. Obviously no Polar comparison here. Here’s the Garmin first:
First thing to note is Garmin does a nice job of presenting everything in one view — total sleep, stages, wake time — nice job. Second thing to note is that colour scheme is a fucking disaster.
How does Whoop compare? Overall summary looks like this:
So Whoop has me about 30min less overall sleep and 30min more deep sleep. The other bits differ a bit too but for me these are the main elements because the first one is fairly easy to validate and the second one, well, gainz man.
So looking at overall sleep and highlighting the ‘awake’ phase in Whoop gives:
The two things I know for a fact are that I put my phone by the side of the bed just after 23:30 in order to go to sleep. That approx 20 mins or just under at the beginning seems ‘about’ right.
There’s a big difference in wake time though. Whoop says 5:18, Garmin says 6:00. Can confirm I was excited to watch first day of NFL draft and did in fact get up around 5:20am. That 6am bit was I think when I turned sleep mode off on my phone. So almost seems like Garmin tracks sleep by your phone rather than watch.
Second bit I was interested in was the deep sleep element. Mainly because I always seem to have this element reported as quite low (Whoop) or very low (Garmin) and blame this for most of my lack of looking like The Rock despite always being in the gym (I know — there may be a few other inputs I’m missing here). Anyhow, deep sleep looked like this in Whoop:
You can see from the previous charts in any case that Garmin absolutely under-reports deep sleep compared to Whoop — that was obvious from the moment I switched over to Garmin 18 months ago. What’s interesting though is the fact that the level of deep sleep has increased massively according to both devices since the last time I looked. I suppose that explains one of the unique differences between the Whoop and Garmin devices — I can’t tell you how much deep sleep I’ve been getting over the last 18 months. I’m absolutely aware of how much I’ve gotten over the last couple of days.
Wrapping Up
If you’ve read this far, fuck me that is honestly impressive because I’m not sure all this is of any use to anyone except me. And it’s maybe not even that useful for me to be honest. Anyhow thanks and well done if you did actually get this far. My summary is as follows:
Wrist-based optical heart rate devices are iffy.
Arm-based optical heart rate devices are a bit less iffy.
Despite all this, these things might help you smash some goals. It is mostly on you though.
Whoop
