How well does Fitbit track cycling?

[mikeptc9107]

How well or accurately does Fitbit Force track activities such as Spin Class or Cycling? Has anyone compared the caloric numbers to a device such as a Garmin? I would love to hear of others experiences in this regard. Thanks, Mike

[SteveH]

Short answer: not very well. Your fitbit is basically a pedometer so whilst it's great at tracking step based activities, it is not so great at tracking other activities, such as cycling.

By wearing your device on your foot you may get some steps recorded but the recorded calories may be questionable.

However, you can manually log activities that are not step based to give reliable results. Here's how:

https://help.fitbit.com/customer/portal/articles/413311-how-do-i-log-or-record-an-activity-

[FitDiitz]

I have worn my Fitibt several times when on the bike trainer at home for 45 minutes or so.  I also use a Garmin to measure wattage, as that is a more accurate approach for cycling.  That said, Fitibit records about 1000 steps for the time I'm on the bike.  I also enter my cycling workout in the activities section.  I usually wear the Fitbit at the waist level when I'm on the trainer.  I'm a 60 year old woman, so depending on your age and sex and how long you stay on the trainer, you may get a very different number.  I'd be curious to hear what folks who take a spin class get for a step count.

[gerrymcd]

I use a bluetooth heart rate monitor and the app digifit on my iphone to record calories I burn cycling.  It works really well. 

[Mary]

Trying to estimate cycling calories with a Fitbit would be like trying to take an xray with a camera.  You might get something but it's the wrong tool so why bother.  

[JeffWunder]

My fitbit is great, but for cycling I use a Garmin Device to track.  I then enter it manually on the dashboard when I get home.  It is the most accurate way to do it.

[Baltoscott]

I agree with everyone else -- it doesn't track cycling very well.  That said, I wear my One in my pocket anyway.  I get about one quarter the steps/mile as when I walk and it tracks floors when I do hills. Since I am motivated by floors I seek out hillier rides and generally get a better workout on my commutes.

---

@mikeptc9107 wrote:

How well or accurately does Fitbit Force track activities such as Spin Class or Cycling? Has anyone compared the caloric numbers to a device such as a Garmin? I would love to hear of others experiences in this regard. Thanks, Mike

---

If you want to get a better estimate of the calories you are burning a garmin with a HRM would be much more accurate.

[au4all]

You need to use a watt-meter on your bicycle to measure calories burned. Almost everything else vastly over-counts calories burned.

100 watts for an hour is 364 calories.

It's explained right here: http://www.bikeforums.net/archive/index.php/t-153865.html

[Srfleury21]

I just went for a ride today, i wear my fitbit on the bottom of my biking shorts and it actually tracks pretty well.

18.9 miles, cadence (per garmin) was approximately 85 to 90 for most of the ride and I rode for 1 hour and 6 minutes.  I ended up with just over 5000 steps which coorelates pretty closely to my RPM x total time.

Just my experience though.

Shawn

[Srfleury21]

Oh and once i get home i manually log the activity in the log and it seems to do a good job of not double counting the activity.

[Baltoscott]

That is similar to my experience, Shawn.  I don't know what my cadence is, but I'm riding slower than you are so it makes sense that I would get more "steps/mile" on the bike than you do because it takes me longer to cover the mile. I would probably get around 8000 - 9000 steps for a 19 mile ride at my normal commuting speeds of 12-13 mph.

[Mary]

---

@Srfleury21 wrote:

Oh and once i get home i manually log the activity in the log and it seems to do a good job of not double counting the activity.

---

Why do you bother wearing it if you overwrite the data by manually logging the activity when you get home? 

[Srfleury21]

Because it keeps the step count from the device but increases the calorie
count on the website.

[emtboss]

Thank you- that was very helpful!

-emtboss

[gsansoucie]

For spinning, I use Digifit which is the best app for spinning and fitbit integration IMHO.

I wear my Fitbit One on my right bike shoe (Shimano SPDs with velcro straps) and let the fitbit One track my steps.  I have a HR monitor that I use with Digifit and it gives an accurate Calorie burn count that auto syncs with Fitbit after my workout.

With all that said, it is spring here and I am no longer spinning as I prefer to ride my roadbike.  Synchronization of my road rides is a lot more tedious.  I've not found an app that offers decent fitbit integration yet...

[Heybales]

---

@gsansoucie wrote:

For spinning, I use Digifit which is the best app for spinning and fitbit integration IMHO.

I wear my Fitbit One on my right bike shoe (Shimano SPDs with velcro straps) and let the fitbit One track my steps.  I have a HR monitor that I use with Digifit and it gives an accurate Calorie burn count that auto syncs with Fitbit after my workout.

With all that said, it is spring here and I am no longer spinning as I prefer to ride my roadbike.  Synchronization of my road rides is a lot more tedious.  I've not found an app that offers decent fitbit integration yet...

---

Doesn't Digifit have a GPS component?

Missing cadence perhaps.

[CORiverRat]

If you use a garmin or other app you might want to check out sites like this.  They allow you to download your cycling activities into Fitbit directly.  http://www.fitdatasync.com/

[2labz]

The logging program (activities tab on fitibit.com) does a lousy job of estimating calorie burn while cycing. The program uses a discrete function to calculate your calorie burn, based on that chart that says that 12-13.9 mph is 8 METs, 14-15.9 is 10 METS, etc.

You can test this assertion yourself by entering a number of distances and holding the time constant. FOr simplicity's sake, I used a constant one hour time and tested a number of distances between 12 and 14 miles (in other words, exactly 12 to exactly 14 MPH). The calculated calories burned stayed exactly the same for 12, 12.1, 12.2, 12.3, 12.4... up to 13.95 miles. At 13.96 miles traveled (by definition, 13.96 MPH) I supposedly burned 110 more calories in my hour than I did at 13.95 MPH.

I reported this to support, but they don't get it. What bothers me is that the site's estimation of calories burned on other activities, including walking/running might also be programmed wrong. 

[Heybales]

---

@2labz wrote:

The logging program (activities tab on fitibit.com) does a lousy job of estimating calorie burn while cycing. The program uses a discrete function to calculate your calorie burn, based on that chart that says that 12-13.9 mph is 8 METs, 14-15.9 is 10 METS, etc.

You can test this assertion yourself by entering a number of distances and holding the time constant. FOr simplicity's sake, I used a constant one hour time and tested a number of distances between 12 and 14 miles (in other words, exactly 12 to exactly 14 MPH). The calculated calories burned stayed exactly the same for 12, 12.1, 12.2, 12.3, 12.4... up to 13.95 miles. At 13.96 miles traveled (by definition, 13.96 MPH) I supposedly burned 110 more calories in my hour than I did at 13.95 MPH.

I reported this to support, but they don't get it. What bothers me is that the site's estimation of calories burned on other activities, including walking/running might also be programmed wrong. 

---

All the databases do this, and most use the same MET foundational database.

Any descriptions that actually do get specific, are actually a block amount for that time, there is no curve on straight line function to the calorie burn even though we know that happens in real life.

But considering a lot of exercise have no specific description of pace or speed or intensity, it actually improves the possible accuracy compared to the alternative - Slow, Medium, or Fast descriptions and blocks of calories.

Here's a better biking calc.

http://www.bikecalculator.com/index.html

It's not a matter of wrong, they are merely going for simplicity and giving you more options than needed to be given to improve accuracy.

They could have done like running though, skip the ranges and just use whatever speed it was that is best match for the burn. Then let the user decide to round up or down.

[2labz]

Sorry, not going to bother learning the ins and outs of yet another crowd-sourcing forum, so bear with me on the "quoting" function..

"Any descriptions that actually do get specific, are actually a block amount for that time, there is no curve on straight line function to the calorie burn even though we know that happens in real life"

I said that already. That's what "discrete function" means. 

"But considering a lot of exercise have no specific description of pace or speed or intensity, it actually improves the possible accuracy compared to the alternative - Slow, Medium, or Fast descriptions and blocks of calories."

"Improves"? I doubt it when it comes to cycling, which is what I'm specifically addressing. It's just lazy programming or, more likely, poorly thought-out specifications to the code monkeys. It would have been quite easy to use that UNC compendium to estimate a linear function comparing average velocity to METS per minute: assuming that the posted values in the table are for the midpoint of the speed range, the formula is something like X = 0.7093Y - 0.7575. Converting that to total calories burned is merely an exercise in eighth-grade math. My guess is that the function isn't really linear, but it looks it through the 10-16 MPH range.

For what it's worth, that's what user support at MapMyRide.com says they do (they actually understood the question, which Fitbit support did not). That being said, their calculator seems to me to overstate calorie usage. 

None of that is important. What is important is that Fitbit claims to calculate a user's "total burn" due to steps (and for some units, "flights") while wearing their family of pedometers using some sort of corporate special sauce -- irrespective of logged activities.  Their formula obviously uses steps per minute and user-supplied stride length(s) to calculate speed, presumably on a five-minute granularity. If they use the same flawed methodology to calculate calorie burn from the pedometer function, then the calculated calorie count is - statistically speaking - never right.

It's a question of precision vs. accuracy: the dashboard displays a nice, precise-looking number four-digit number - but is it accurate? Based on somewhat slipshod programming in other areas, this is a valid question.