Happiness Alert: Google Fit Can Be Wildly Inaccurate

Happiness alert: Google Fit can be wildly inaccurate.

I like a lot of what Google has done with Android.

But, my first job is to tell the truth. As Happiness Is a Serious Problem author Dennis Prager says, “Tell the truth, then give your opinion.”

The truth is that Google Fit on an Android phone is a poor substitute for a good pedometer and a good GPS watch.

Relying on Google Fit now to help you with your training and racing will lead to unhappiness.

To prove how wildly inaccurate Google Fit can be, here are results from two runs.

My group training program had mapped each run in MapMyRun, and my GPS watch’s distance measurements closely matched the MapMyRun distances.

Google Fit was running on a Moto X cellphone, which I had in a pocket of my hydration belt.

I also had in my hydration belt — in two other pockets — an Omron pedometer and a Garmin 310XT Forerunner watch (because the wrist-strap was broken).

Here are the results:

 
I trust the Garmin distance and duration measurements and the Omron step-counts. Garmin dominates the GPS world, and Omron dominates the pedometer world.

So, what can account for the discrepancies?

• I was careful to set aside my hydration belt as soon as each run ended. So, I can’t blame Google Fit’s higher distances, durations, and steps on walking a lot more with the cellphone than with the pedometer and GPS watch.
• The Omron’s measurement of ~3K extra steps for an extra 0.11 miles in run #1 vs. run #2 looks odd, especially to anyone who has read Weight Loss in 10,000 Steps a Day. However, I used the 3:1 method for the first 5.5 miles of run #1 and then switched to the 5:1 method for the remainder. In contrast, I used the 3:1 method for all of run #2. You might expect that more-frequent walking in run #2 than in similar-distance run #1 would yield more steps in run #2. But, as I showed in “Can Lowering Your Run:Walk Ratio Make You Faster?”, the opposite can occur. A lower run:walk ratio can, paradoxically, lead to fewer steps for the same distance.
• Distance, duration, and steps counted across both runs are all higher with Google Fit than with Garmin + Omron. But, look at the drastic errors with Google Fit in run #2 that are not as drastic in run #1.

So, not only can Google Fit be quite inaccurate, it also can be unpredictable in its inaccuracy.

Conclusion and Recommendation

Conclusion:
[Tweet “Google Fit can be wildly inaccurate when compared to a GPS watch + a pedometer.”]

My wife now coaches half-marathon walkers. A newcomer to the sport often will tell her that a cellphone app — on Android or iPhone — reports a walking pace that my wife knows is impossible — as in faster than reality. My wife then has to advise the newcomer to take what such an app says as only a very crude estimate.

The smartphone industry has excelled at selling consumers on the notion that cellphones can replace nearly any dedicated device. While they do many things well, my wife’s experience and my experience show that cellphones cannot do everything well.

Recommendation:

1. Measure your run (or walk) duration with a stopwatch or GPS watch.
2. Measure your distance with a GPS watch.
3. Measure your steps taken with a pedometeter.

Yes, as I advise in 123 Cellphone Tips for Runners and Walkers, train and race with a cellphone. It’s the safe thing to do. But, stick to dedicated devices, not cellphone apps, for accurate measurements of distance, duration, and steps taken. You’ll be a happeir runner, walker, or run:walk athlete.