MFP Overestimates Value of Google Fit Steps Calories

jfmmxiv

Users have been reporting this issue for several years, but I just started using MFP seriously and had to learn the hard way that MyFitnessPal still has a bug that makes dramatically overestimates of Google Fit Steps calories. In fact, the higher the GF step count, the bigger the error will be.

MFP will tell you that you have earned 4-5X the calories than you really have.

I've reported this to technical support several times and always the blame is placed on Google. However, all you have to do is create a spreadsheet and do some math---or download the free version of MyPlate---and you will get much more realistic estimates. MyPlate, which does not have as many features and costs just as much for the premium version, nonetheless properly strips out the BMR Google Fit calculates based on your age, height, and weight, and gives you a reasonable value for steps.

For instance, with a daily BMR of 1,776, at 6:47pm (with 78.3% of the day complete) the approximate GF BMR up to that point was about 1,391. In this example, GF reported a steps count of 4,586 and 1,497 calories total. The net calories for the steps then is 106 after subtracting out the BMR used by GF. (NOTE: There usually seems to be a 12-14 data lag with the GF feed, so if you're making a spreadsheet for a current day, you will need to account for that.)

MyPlate connected to the same GF account does this same calculation instantly and there is NO REASON MyFitnessPal can't also do it.

So, at the same time I got 106 calories from my spreadsheet and using MyPlate, MFP was telling me those same 4,586 steps were worth 522 calories (!) and added that to my food budget.

I basically have my phone with me all of the time, so the step count is fairly good.

But until MyFitnessPal finally fixes this on the Android, I will have to disable the much more convenient automatic step counting links to Google Fit and make manual adjustments each day based on my spreadsheet or MyPlate. Of course, I may end up dumping MFP completely, but I am hesitant to do that b/c of the time invested so far and its other good features.

In anticipation of the predictable responses that MFP calculates calories better or at least differently than GF, etc., etc., that's all fine and maybe MFP does have better algorithms. I don't know. But I do know that this feature is broken and it is on the MFP side. And it is consequential.

Fix it, guys!

PDKL45

I have been noticing this error as well. Does anyone know if there is an exercising tracking app like Google fit that reflects calories more accurately? I really like Google Fit for its automatic nature; it just starts recording movement no matter what.

PAV8888

Well my n=1 says that google fit doesn't know the first thing about caloric expenditure and under-estimates my expenditure based on RESULTS by >20% ... by contrast in the same time span Fitbit over-estimated by less than 3%... I'll take % vs >20% error any-time

Having said that, MFP assigns more than 1x BMR to each minute till midnight depending on the activity setting you've selected. So you should not be adjusting at <100% of your day without taking this into account.

Your step calories are NOT an adjustment = to the calories that came from your steps. They are a TDEE adjustment comparing the expected value of your external source (google fit) to what MFP is expecting you to spend till end of day.

The current work-around for iphone "step" problems is to use an intermediary app that is liked by both sides (such as Pacer, only mentioned because I am aware that it works). You may want to try something similar with Google Fit... esp since Pacer would also be running on the same phone anyway...

jfmmxiv

I use Polar Beat for running, cycling, etc. I mainly just use Google Fit to get a rough indicator of activity and automatic walking tracking. I know it's not perfect, but neither is MFP.

PAV8888: 1) Thanks for the Pacer idea. I'll look into it. 2) I know that the GF steps calories is supposed to be an adjustment, but the values MFP assigns make no sense, are consistently very high, and result in an "Exercise" tally that is very inflated. This, in turn, says you can eat way more than is good. If the overall caloric budget that resulted made some sense, I'd be less concerned about what they call it or how it is derived, but this is just broken.

Lastly, there has been lots of back and forth and many opinions offered about the merits of setting the MFP activity level to Sedentary to minimize this adjustment problem. What do you all think?

jsparks4000

To be honest, Ive found MFP over estimates most exercise. Although to be fair so do most gym machines and other trackers. Ive researched my most common exercises and figured out most accurate formula and use that and manually edit the entry.

08_GreenEyedMomma

I don't have Fit linked for steps; I only use it to track intentional exercise but MFP reports higher cal burn than Fit shows. It's a weird glitch, and it's also relatively new. MFP used to show me exactly the #s that Fit did. Hopefully it gets resolved, but until then I just correct the cal burn to what Fit shows.

jhanleybrown

I've found MFP estimates for my main exercises (cycling, walking, running) to be off substantially (50-100%). But there are accurate formulas out there.

GBO323

I am running into this issue as of Sept 5, 2020. I've submitted a Support ticket as well and noted I'm a Premium member if that might help.

I created a food called 'Step Adjustment' and enter in half of the calories that the Google Fit Calorie Adjustment shows into my exercise diary. Not ideal, but it works. I still think it's a tad high, but better than keeping ALL those calories.

Hope MFP can fix this, but at least we can workaround it.