LG Watch Sport vs. Microsoft Band 2: GPS Accuracy

I recently switched from a Microsoft Band 2 fitness tracker to an LG Watch Sport. I wasn’t sure that it was a step up for me, honestly, so I decided to run a test.

I am a fair-weather runner. I run in the Spring, Summer, and Fall, and I run for myself. I don’t race, I don’t push myself for any reason beyond general fitness. I’m not, by any definition, hardcore.

I also am (used to be?) a big Microsoft fan. I’m the guy who adopted Windows Phone 7 the moment it came out and stuck with it through the Lumia 950. I’m typing this on my Surface Pro 3. I bought the Microsoft Band when it launched and the Band 2 shortly after its release. The Band was good. The Band 2 was even better.

Always-on heart rate monitor? Table-stakes. Stand-alone GPS? Check. Notifications? You got it. Ability to reply to texts by typing? That was there. Microphone for talking with Cortana? Yup, it had that, too. Then you add in the extra stuff: barometer, UV detector (yes, my Band 2 told me when to apply and reapply sunscreen), sleep tracking, and a few helpful mini-apps and it had everything I needed.

Better still, it had training regimens. The Band 2 taught me how to run, working me up through three different training programs, mixing up distance runs, wind sprints, fast-slow runs, runs with body weight exercises thrown in. The Band 2 took me from unfit to being able to run 5 miles consistently. It motivated me to get up in the morning and go.

The Band 2 was nearly perfect. One of the few things it lacked was a reliable band (ironic). I went through three Band 2s. The rubber/plastic strap on each cracked and split.

My third Band 2 broke back in March of this year. Microsoft had discontinued the hardware and there was no free replacement to be found. Electrical tape and super-glue held it together while I looking for a new solution.

You see, I’d (foolishly?) committed to running the Washington DC Ragnar race with a team this September. I needed to transform my casual running into actual training. Being a geek, I’m motivated by technology and running without a tracker that I was excited to use just wasn’t going to work for me.

Fitbit’s lineup was uninspiring and didn’t have the combination of semi-smart watch, fitness band, and training platform that I was after. As a newly-minted Android user the Apple Watch 2 wasn’t going to work for me. The LG Watch Sport seemed to be right up my alley.

My wife picked one up for my birthday and despite the rainy day I took it out for a run (waterproof!). I came back, curious about the results.

My run’s map made me look like a drunk squirrel.

Believe it or not, I was running on the Bluemont Junction Trail, not the road. Or the grass. Or anywhere else I was showing up.

Another run and a walk in downtown DC on equally dreary and drizzly days produced similar GPS results.

I was disappointed. How was I supposed to train well if I didn’t have data that I could trust? My first run with the LG Watch Sport transformed a 5 mile run into a 5.8 mile run. We’re not talking about a few hundred feet here or there.

A colleague pointed out that GPS and rain don’t mix very well, which may have been the primary issue. Again, being a geek, I decided to do a test. This past weekend I headed down to the W&OD trail with my old Band 2 on one wrist and my LG Watch Sport on the other.

At the end of the run I loaded each run’s GPS details into Strava and took representative screen shots of the output. A quick note before we dive in: due to user error I accidentally ended the run on the Watch Sport at 5.2 miles instead of the 6.0 miles for the full run.

So, what did we see? First, my old favorite, the Microsoft Band 2:

Forgive the stats: my fitness is not the focus. But look at dat map!

And now, the LG Watch Sport:

A bit rough-around-the-edges.

At a glance the LG Watch Sport is much more squiggly. The lines from the Band 2 adhere much more closely to the trails and roads I ran. The LG Watch Sport also has me running faster, in general. Of note: the squiggly-ness of the LG Watch Sport is much reduced from the rainy days. Clear skies greatly improved its measurements.

Let’s look at some details, though:

Zooming in emphasizes the smooth lines of the Band 2, but you can start to see that this is betraying some level of accuracy. The Watch Sport tracks my cross under Rt. 66 accurately: two nearly 90-degree turns. The Band 2 misses this.

The run on the Custis Trail along Rt. 66 is tricky for GPS: your shoulder is up against a metal noise-wall separating you from the highway. Neither tracker performed particularly well here (I did not take a detour up to Fairfax Dr., per the Watch Sport, nor did I run through an apartment building, per the Band 2).

There are weaknesses, again, for both trackers as I detour around a building to dodge some construction. The lines from the Band 2 are more visually appealing, though. And this is never more clear than in this next shot:

I went “out” along the W&OD trail, circled around and came “down” the Bluemont Junction Trail, and then went “out” the W&OD once more only to come back along Four Mile Run Trail. That is more clear with the Watch Sport, but the Band 2 is crisp and holds closely to the trail.

So what is going on? Another co-worker caught it right away: the Watch Sport is pinging my GPS location much more frequently. The Band 2 takes fewer measurements and fills in the gaps. This is very evident around curves and along tight turns where it doesn’t update its location often enough to capture the details.

The Watch Sport gets the location more often, which is why it caught the 90-degree turns under Rt. 66, but because of the variability in GPS location these more frequent readings lead to more errors.

So which is better? I don’t know. The pace graph for the LG Watch Sport has higher variability:

LG Watch Sport

The pace graph for the Band 2 (on a different scale, remember), is more smooth:

Microsoft Band 2

I don’t know which is more accurate. I don’t know which is more helpful for training, either. Maybe too I have a bad Watch Sport and other people are not seeing similar results.

It would be incredibly valuable to see how these devices stack up against a real runner’s watch from the likes of Garmin as well.

Any thoughts, tips, or supporting information as I try to answer these questions would be incredibly helpful. Enjoy the analysis!

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.