Why Garmin and Strava calculates differently



novetan

New Member
Sep 1, 2012
102
0
0
I realized there’s abt 300 Cal difference when calories are calculated between Garmin and Strava. For eg, when Garmin registered 700 Cal during my cycle, and after uploaded to Strava, it becomes 1,000 Cal. Why is that so and which one shld I trust. 300 Cal is a huge difference.
 
novetan said:
I realized there’s abt 300 Cal difference when calories are calculated between Garmin and Strava. For eg, when Garmin registered 700 Cal during my cycle, and after uploaded to Strava, it becomes 1,000 Cal. Why is that so and which one shld I trust. 300 Cal is a huge difference.
Apparently, they each use different methods. As for which one to trust, I'd trust neither. If you want to get a lot more accurate measure of how much energy you've used, get a power meter. If you want to get the most accurate measurements, get some metabolic testing done to characterize your body's use of energy and couple that knowledge with the use of a power meter.
 
Are you getting altitude variations between Strava and your Garmin?
 
My guess is that Garmin is using only heart rate to compute energy expenditure whereas Strava is using your weight coupled with speed and elevation from the GPS data to compute power and energy expenditure.
 
Strava also takes into account the weight of the bicycle. If you change the bicycle used in the ride the energy output and calories change too.
 
Originally Posted by maydog .

My guess is that Garmin is using only heart rate to compute energy expenditure whereas Strava is using your weight coupled with speed and elevation from the GPS data to compute power and energy expenditure.
I'm using Garmin Edge 800. It has my weight and bike input, as well it can record elev, speed and cadence. And Strava is extracting figures from Garmin when I download.
 
Strava posts this: https://strava.zendesk.com/entries/20959327-Calorie-Calculation

"Calorie calculations are only estimates. You may notice differences between Strava's calculations and those from other tools, even given the same activity data. This is most likely due to differences in Calorie calculation algorithms, and the data factored into the calculations. We believe our estimation is as accurate as possible given the limitations in data."
 
There is a nice whitepaper on the firstbeat algorithm in the DC Rainmaker article. I would point out, however, their accuracy claims in the white paper use respiratory rate coupled with heart rate. Heart rate alone has twice the error.

I would also note that the algorithm uses a trained neural network. That may be fine if you are within the group of people it is trained to. I am 3 standard deviations outside the weight and height range so I don't expect highly accurate results.

I wonder how this compares to the Cycleops powercal algorithm. I have both a Garmin 405cx (firstbeat 1st generation) and a Powercal. During a normal relatively intense hour workout, the 405cx would report 800-900 calories burned, the Timex Global trainer with powercal would be in the 700-800 range. I am pretty certian that the powercal significantly underestimates my power for intense workloads.

Perhaps I should wear both for a direct comparison.
 
It's not like Calories even mean a whole lot. Food labels are inaccurate in their Calorie counts, using an antiquated burn-measurement system that doesn't necessarily correlate to how the body actually digests and absorbs the food. For instance, included on nutrition labels is the Calorie count of fiber--included non-soluble fiber that passes through the digestive system untouched.
 
"Perhaps I should wear both for a direct comparison."

Look at the bright side...carrying the additional weight will burn more calories!

Rainmaker posts some very interesting data. MHO is that any data is a good comparative baseline.