This review really focusses on the optical HR data collection with my newly purchased Polar Vantage V3 watch.
Last autumn, I bought a Garmin Vivoactive 5 watch, largely to collect data that was then pulled in by Athletica.ai, such as sleep, resting HR and HRV. Watches have never really been a maor part of my sports data collection - because I almost explusively train on a bike, I generally collect data via a Garmin Edge bike computer (currently an 840), or via trainer software such as Zwift. So, why go back to Polar, specifically the Vantage V3?
I was mostly using the Garmin watch (and before that a Polar Ignite 3 watch) to collect those lifestyle parameters. But I find the data as processed by Garmin Connect to be rather unhelpful, and I generally prefer the interface on the Polar watches. I also prefer the Polar magnetic charging puck over the charging socket on Garmin watches.
I ordered the watch up at £320 from Amazon, about £100 less than usual (and the list proce at the Polar website is over £500). This was a “return”, but the carton seemed unopened and sealed - the contents were all present and correct. The Vantage V3 was released some time back, which may mean a new version is imminent, but I felt that the firmware updates since the watch was released may have enhanced the watch's performance.
I've been playing with the watch and trying to get a handle on its good and bad points, focussing on how to maximise the accuracy of the optical heart rate measurement of the watch. Some prior reviewers had highlighted issues with HR accuracy (DC Rainmaker and The Quantified Scientist). I’ve generally found wrist-based optical HR a bit variable in accuracy. This may be because of my thin and bony wrist, particularly my left wrist where I normally wear a watch - using a Polar Verity Sense optical HR sensor on my forearm or upper arm works perfectly. It's my impression that my wrist is just too thin for a good optical signal from blood flow, and coupled with an inflexible watch band can make for pretty erratic heart rate measurement. I have found that pushing the watch a couple of centimetres up my forearm makes the data far more accurate. This is helped by swapping out the plasticky watchband for an elastic one bought form Amazon helps a lot.
I normally use a Polar H10 chest strap HR sensor for recording my training sessions. This is widely accepted as a good reference device, and that's how I'm using it here. My Polar H10 is attached to the Polar Flow phone app, but isn't linked to the Vantage V3. I've seen comments that if a sensor in attached to Polar Flow, it automatically links to the Vantage V3. But this doesn't seem to be the case for my tests: firstly the optical sensor is active throughout each workout and secondly there are always some differences between the HR data recorded by the Vantage V3 and via the H10 - and some of these data are very different.
I've used DC Rainmaker's Analyzer tool to generate these graphs. I haven't shown GPS traces in this report because my focus has been on HR data quality. On the occasions I've recorded GPS data on the Edge 840, Polar Vantage V3 and Vivoactive V3 the tracks have been virtually identical. It has to be said that I wasn't testing GPS in difficult environments. In each case, you can click the graph to view a larger size.
Summary
Prior to this series of tests, I'd always found optical HR measurements with a wristwatch mounted sensor to be woefully inaccurate. I've now revised that view and reckon these sensor issues may depend on (a) how much 'flesh' you have beneath the sensor and (b) how snugly the watch sensor fits.
For me and my thin bony wrists, particularly when wearing the watch on my left wrist, accurate HR readings are really difficult to get. There are two things I have changed which greatly improve this. Firstly, don't wear the watch actually adjacent to the wrist bone. For me at least, my wrists are too thin and bony! Moving the watch band up my forearm by a centimetre or two greatly improves the data. The second factor is replacing the supplied wristband with an elastic nylon fabric band. This makes pushing the watch up the arme easier, and makes for a good, snug, fit.
Optical HR data always seems to experience a lag at the start of a workout before accurate data is acquired. This may be a biological thing - maybe it just takes a minute or so for the blood flow to rise and be detected. Thereafter, the data seem to be pretty good quality to me.
I think these comments apply to optical HR detection in general.
22/7/25 Indoor bike training
The main data set was recorded via Zwift (purple trace below) - the HR data were derived from a Polar H10 chest strap sensor. I wore the Vantage V3 on my left wrist using the supplied watch band. I think the chart below needs a little explanation. I noticed that the watch wasn't tracking my HR terribly well at around 15 minutes in to this workout, and at around 20 minutes in, when I reached the end of the first interval, I loosened the watch band slightly and moved it up my forearm by a cm or two, whereupon the Vantage V3 and Polar H10 HR data semed to align pretty well.
29/7/25 Bike ride, West Yorkshire.
Gear: Principia with Powertap hub; Garmin Edge 840 paired to Powertap Hub and H10 HR sensor (ANT+); Polar Vantage V3 Optical HR, left wrist, regular strap; Garmin Vivoactive 5 Optical HR, right wrist, elastic strap.
Many YouTube reviewers reckon that outdoor rides are more challenging than indoor rides for optical HR devices mounted to the wrist. In my case my wrists are pretty thin and bony. For this ride, I wore the Vantage V3 on my left wrist using the supplied silicone plastic band, and the Garmin Vivoactive 5 with an elastic fabric band on my right wrist. The power meter used was an Powertap hub, with only ANT+ communication, so I couldn't connect that to either watch,
The Vivoactive 5 HR tracked the H10 pretty well, but the best I can say for the Vantage V3 was that it was completely woeful on this ride.
A more detailed review sees some drops in HR on the Vivoactive 5 (for example a couple at around 1h55m), but on the whole it’s not bad. Maybe the elastic strap helps keep the optical sensor tight.
GPS
On this ride, the GPS traces for the three devices were essentially identical. This was a countryside bike ride, and I don’t think there was any difficult terrain (such as tall buildings or dense forest) for GPS problems.
2/8/25 Indoor bike training
These data were recorded using the same setup as for 22/7/25, with the exception that I started with the Vantage V3 a few cm up my forearm from my wrist, and I used an elastic watch band. Here, the Vantage V3 optical HR data are in red. As usual, there's a bit of a delay before the HR data pick up, thereafter there's a pretty good match to the Polar H10.
A second indoor training workout on the same day shows the same initial lag in the optical HR measurement.
6/8/25 10 mile Time Trial Bedfordshire
I rode the club time trial on a local hilly course, and recorded data on my Garmin Edge 840, using Polar H10 sensor for heart rate and Favero Assioma power meter pedals. I recorded the same ride using the Vantage V3 optical HR and linked the pedals via BT.
Heart Rate
Both the Polar H10 data recorded via ANT+ on the Garmin Edge 840 and the Polar Vantage V3 OHR data match pretty well, certainly good enough for use. Aside from a few minor deviations, I notice that the Vantage's optical HR sensor has a bit of a lag in registering elevated HR, and this is something I'd noticed before with OHR from a variety of devices.
Power
I think the Favero Assioma pedals are linked by ANT+ to the Edge 840 and by BT to the Vantage V3. At first sight there's a pretty good correlation between the two devices.
There seems to be something of a series of dropoutsin the Vantage V3 data at around 14 minutes. I don't know if this is a data sampling error or BT dropouts. Closer inspection shows each is for one or two seconds. I also notice that the Vantage V3 doesn't record L-R data or other pedalling metrics.
GPS tracking
Comparing the Edge 840 and Vantage V3 GPS tracking again showed near complete concurrence. The tracks are so close that I had to really zoom in to see any differences between them.