Garmin 955 Accuracy Return to the Garmin Forerunner 955 Solar Review In a nutshell: With caveats, the Garmin Forerunner 955 is an accurate sports watch. I've been testing the FR955 in the maximum accuracy mode (Multi-Band) and can contrast what I found with the Garmin Epix 2 which had Garmin's most accurate GPS ever. I did also the review the Garmin Fenix 7 (non-Sapphire) but without the maximum accuracy Multi-Band option. Multi-band definitely improved accuracy on the Epix 2 but there are other important sensors on Garmin sports watches. Arguably, the Elevate optical HR sensor is the most important Garmin sensor as the data it collects is a crucial input to all of Garmin's Firstbeat physiology algorithms. Q: Why might the FR955 be more accurate than Epix 2 even though they share the exact same sensors? A: The lightness of the FR955 will mean there is less movement on your wrist and that should increase the accuracy of oHR and maybe GPS too. A: The plastic shell may impact differently on the GPS antenna...could be better...could be worse. A: FR955 has arrived later in the year. Thus, I've tested the FR955 in warm weather and, in my case, that tends to give more accurate oHR results. Historically I'd put that down to more blood near to the surface of my skinny wrists. Here are the results of tests on oHR, nightly HRV, altitude, GPS position, instant pace and battery consumption. oHR Accuracy I tend to compare oHR to a chest strap like the HRM-PRO and a bicep strap like the Polar Verity Sense both of which tend to be accurate. I'll then sometimes add in optical HR sensors like WHOOP or another sports watch. Generally, I'm a bad candidate for oHR accuracy on the wrist. Here's what I found in Summer with the FR955. Trail run - Apart from the start, the 955 performed well when soft underfoot Road Ride Apart from the start, the 955 performed well on an often bumpy route give or take the odd spike. OWS - Swim Every single one of my three OWS tests has gone belly-up with the other devices. Each time the 955 seems to have recorded a plausible track, unlike the Apple Watch and HRM-PRO. This is one of the better results! but the second effort period looks to have values that are too high but I have no comparators. Trail Ride FR955 handled this bumpy trail ride very well. FTP Test This ride effectively amounted to something close to an FTP test and the FR955 did a great job on the HR. Running Efforts A great performance here with the FR955 doing a few short intervals interspersed with a combination of jog and walk recoveries. Moderated Effort Ride over Bumpy Hills A surprisingly good oHR performance over the Surrey Hills Tempo Bike on TT Bars Another great performance on the road, this time on my TT bike. Note a slight omission of data at 38 minutes. Steady-State Run Good, once more. Elevation Accuracy There's nothing to see here. It's accurate enough for me. [gallery size="medium" columns="2" ids="73437,73438"] GNSS Accuracy I'm looking here at the accuracy of the track recorded in the workout file compared to where exactly I was and also in comparison to other devices. I have one standard 10-mile test that I repeat for every GPS watch but the other tests are part of my training, although I find that some of the other routes I do are particularly challenging in some aspects of GPS so I often repeat those for a new watch. There's little point in my running up and down a river bank or through my local park over slightly different routes if the sky is clear each time. I've found a good place with dense tree cover not too far from home so now the only significant GPS reception conditions I omit are 1) City centres with skyscrapers (I have a nearby large town with 6/7/8 story buildings) 2) Mountainous valleys (I have the Surrey Hills but it's not as challenging for GPS) The GPS firmware has been updated to v2.33 compared to the v2.27 on my earlier Epix tests. So something has changed on the software side as well. Generally Dense, medium-height tree cover You can look at the interactive map of these files here on the DC Rainmaker Analyzer. It's hard to see where the real tracks are as the tree cover is dense. There's also no point in looking at pictorial map views as there is no guarantee that they have the trail in the exactly correct place. Anyway, I know where I ran and the FR955 produces a consistently better track than the Apple Watch 6 and the Coros Vertix 2 (Multi-Band). [gallery size="medium" ids="73441,73439,73440"] Easy GPS - Windsor Road Loop There's nothing to see here except the comparison devices making small errors. FR955 is great. [gallery columns="5" size="medium" ids="73446,73442,73443,73444,73445"] OWS Yep, this looks fine. The Apple Watch had a little blip on the smaller loop (top right side) [caption id="attachment_73447" align="aligncenter" width="470"] Garmin 955 (Brown)[/caption] 5K Run Around The Lake There's nothing to see here except the comparison devices making small errors. FR955 is great. [gallery size="medium" ids="73450,73448,73449"] Canalside Ride Under Some Serious Bridges Nothing to report here. A slight wobble when the FR955 goes under a bridge. Better than the other devices on test. [gallery columns="5" ids="73455,73451,73452,73453,73454"] Shop Test - Run Running up and back on the same side of the road right next to a parade of shops which open up onto the pavement. Coros Vertix 2 is better in places here than the FR955 but then will have another section where it is inaccurate. So the FR955 is most consistent overall. The FR955 should NOT have its position shifted away from shops as the Multi-Band GNSS should reduce multi-path errors. However, it does do that a bit although better than what I'd seen on the Epix 2 in this instance. Grid Test - Run These are excellent results for the FR955. Not perfect, but not far off. [gallery size="medium" columns="4" ids="73460,73458,73459,73457"] Urban Centre Test Another good performance here. The second image shows the section near Sigma Sports and this is where watches always seem to show multipath errors. The Epix did so. But the Fr955 nails it here. Maybe the algorithm has been improved? The town centre GPS performance is generally very good however there is one section where the FR955 and other devices are all wrong but the FR955 is not too bad even n that section. [gallery columns="5" size="medium" ids="73465,73464,73461,73462,73463"] 10-Mile GPS Test You can inspect the full route and comparison to the Epix 2, V800 and Ambit 3 here plus see the full methodology for the test here. The FR955 was not quite as good over the test route as the Epix 2 and Ambit 3 but did beat the Apple Watch 6, so it was a good overall performance with not much to separate all these excellent watches which have a similarly high level of overall accuracy. I did the 10-mile test at the end of all the other testing and, if I'm honest, I expected the FR955 to give 'best ever' results based on what I'd already seen. Instant Pace Accuracy This involves looking at two wrists simultaneously when running! It's somewhat subjective. Instant pace accuracy felt improved compared to what I'd seen with previous generations of Garmin running watches. Generally, the instant pace seemed within 5-10 seconds/km off the figure that Stryd showed. However, FR955 was sometimes incorrect by over 25 seconds/km. That's not too bad compared to other watches even though it sounds inaccurate. Personally, I wouldn't use it but if you used a current lap pace it would probably be accurate. Nightly HRV accuracy This looks at the HRV assessment of sleep. Garmin choose to provide an average nightly HRV value but that could (should) still correlate with single morning readings from other tools. I use HRV4Training and a Polar H10 which are scientifically validated. The Garmin HRV figure seems quite a bit different to the Oura nightly average which is generally accepted to be good and which has the best correlation for me (below) [caption id="attachment_73471" align="aligncenter" width="1263"] 30-day rMSSD via HRV4Training (my data)[/caption] After a few more weeks of data, Garmin's baseline correlation remains poor but the day-to-day (delta) correlation is improved. Garmin's new 'readiness to train' metrics cannot sensibly be compared across brands as too many fudge factors are added to the algorithms, Garmin included. Aside: the above table shows that there is a far superior correlation between my own subjective assessment of sleep quality and HRV than anything else! Who needs software tools! (Of course, I omit the other 9 subjective assessments which show poor correlations ;-) #CheatingAtScience) Battery Claims Test I have the solar model and so battery results vary according to sunlight. These battery consumptions all use Multi band and optical HR. 6%/hour gives 16.67 hours capacity - on a very sunny day! 4.8%/hour gives 20.83 hours capacity - on a sunny evening with the sun low in the sky 3.6%/hour gives 27.71 hours capacity on a long sunny bike ride 6%/hour gives 16.67 hours capacity under fairly dense tree cover 3.69%/hour gives 27 hours capacity on a late evening run with no sun. So I don't quite know what's going on here. My feeling is that the watch does seem to burn through battery quickly. I'm never concerned about it during 1-3 hour workouts as it's always sufficiently charged beforehand, however in between workouts I feel obliged to charge it up. It doesn't feel like I would get 20 hours on Multi-Band. Conclusions This is one of Garmin's most accurate ever sports devices. Heart Rate - These were the best and most consistent results I've had from an Elevate sensor for quite some time. oHR is usually poor on me but not this time. I put the good results down to warm weather as I had fewer good results with Elevate 4 on other Garmin watches. GNSS - these were market-leading results, though let down slightly when viewed from the perspective of instant pace Altitude is fine when automatically calibrated to the start point over relatively short rides of up to 3 hours with little change to the atmospheric pressure. The increased power requirements required for GNSS accuracy noticeably hit the battery life. Moving forwards with the FR955 as my main running watch for 2022 I will add an HRM-PRO and STRYD footpod whilst at the same time using GPS-only mode. Return to the Garmin Forerunner 955 Solar Review
