Strava over-estimates Calories

Does anyone else notice that the calorie info on Strava is a huge overestimate from the amount of calories you actually burn?  Sometimes the estimated calories burned on Strava after I upload my ride is 2 to even over 3 times what I actually burned.  What algorithm is used and why is there such a discrepancy?  

Today alone I did a ride where I only burned 1,500 kcal according to my Garmin computer (which is actually really about 800 additional calories burned over my basal rate.)  Strava states I burned 3,100 kcal.  My Garmin 810 with all of my personal info, HR data, VO2 Max, FTP, etc I know is much more accurate.  

How could Strava possibly calculate anyone who rides for only 45 miles stayed in Z2 70% of the time, Z3 15%, and Z4 for only 1% could possibly burn over 3,000 calories?  I know it is not accurate, but there are people who do depend on this as their only means of calculating calories burned.  Without a HR monitor it exaggerates even more.  It would have indicated that I burned well over 4,000 calories, when I actually burned less than 1,000. 

BTW, when the ride is uploaded on GarminConnect, the calories info is the same as that on my bike computer.

16

Comments

72 comments
  • Garmin 936 calories versus Strava 281 calories (today)

    Garmin 1038 C versus strava 230 (2 days ago)

    Garmin 1053 C versus strava 271 C (4 days ago)

     

    any idea why strava is showing allot less C based on the import from Garmin.

  • I don't know which is more accurate, the Garmin 500 (with HR monitor) or the Strava entry for the same ride after it was auto-synced. 

    Just did a ride today that was 32.4 miles, avg. speed 16.7 mph (moving speed, not counting the time spent waiting for traffic lights). Avg. HR was 121 bpm (elapsed time was 2:03:34, moving time was 1:56:37, so 7 minutes wasted sitting at lights where my HR fell dramatically).

    Strava calories: 1083
    Garmin calories: 704

    Bottom line is that I'm more inclined to believe that the Strava number is more accurate.

    1) My HR data was skewed downward quite a lot by the non-moving time where I was stuck at stoplights. My HR during the ride was usually in the upper 120s to 130s every time I looked at it. There were around 7 minutes of non-moving time in a roughly 2-hour ride.

    2) Garmin Connect reports my average speed as 15.7, because it doesn't calculate moving speed, just distance divided by the total time. It therefore uses an incorrectly lower average speed to calculate the effort. 

    I believe the Garmin data is skewed downward by this combination of lower heartrate average and lower average speed to yield a lower effort. 

    This isn't scientific, but in the past I've done a lot of exercise on a stairmaster at the gym. On this stairmaster one inputs one's age, sex, and weight, and it calculates calories burned, and stats like rate of calorie burn per hour. In the past I did a lot of stairmaster 30-minute blocks at various difficulty levels, and paid attention to my heart rate (the machine had those metal pads you hold to measure heart rate). Given similar heart rate levels I would assume roughly similar levels of effort, and hence calorie burn. The ~530 cal/hr calculated by Strava for today's effort seems fairly in line with calorie burn rates I saw on that stairmaster at the gym for stairmaster walks at roughly the same heart rate.

    I have no idea what the real calorie burn is. It seems all of the calculators can be all over the place. I've used some online web-based calculators where you input sex, age, weight, distance ridden, and speed ridden, and gotten numbers like 2500 calories for the same ride as this one today. I'm pretty sure that's wildly inaccurate.

    I guess the only way to get a truly (roughly) accurate calorie burn calculate is to put a power meter on the bike. I may end up with a Stages left-only power meter for Christmas or something. I'd really like to know what I really burned.

  • Strava versus Amazfit:

    522 kcal vs 219 kcal (238%)

    275 kcal vs 140 kcal (196%)

    437 kcal vs 197 kcal (221%)

    586 kcal vs 280 kcal (209%)

    391 kcal vs 203 kcal (193%)

    662 kcal vs 353 kcal (188%)

    234 kcal vs 155 kcal (151%)

    358 kcal vs 200 kcal (179 %)

    I mostly do quick walks for fat burning, some short jogging and walk-jog mixes. The difference between Strava and Amazfit are huge - especially with walks (about 6 kmph). I'm not saying that the Amazfit app is perfect; it probably underestimates energy consumption but come on, 200% difference with two professionals is not normal.

     

  • Hi, I realize this is a 'Zombie' and likely won't get any attention. Apparently I am burning 10's of thousands of calories per long run. For instance:

    That was a short ramble up a hill and around some trails.
    This happens on every activity. It is like it's multiplying the actual number by a factor of 100.

    Here's some slow HR training:

    10,000 Calories while maintaining a 135-140 HR. Shouldn't I be dead?
    Thanks....

  • Jade, your weight on your Strava profile was set to over 600 kg. I think that's very unlikely? I have changed your weight and Strava now calculates a touch over 1,000 calories. 

  • Haha! @ellie. Damn, OOPS! Thanks for that. :)

  • It's still drastically different for my activities as well. <1000 calories on Garmin vs >1400 on Strava for an 8 mile hike.

  • I have the same issue as most mentioned above. I've recently started using LoseIt and hoped to be able to connect my runs directly through Strava (as LoseIt does not link to Garmin) but the calories estimate from Strava are on average 20% higher than Garmin and LoseIt itself! 

    If we can't trust the app to get the calorie estimate within reasonable range it makes the feature effectively useless. When will Strava do something about this?!

  • For my LoseIt! solution, I wound up going with RunGap. It talks to a lot of platforms. I route it to Apple Health and LoseIt! picks it up from there. Kind of nuts to have to do that, and pay for an app to do what Strava should be able to handle, but for now it's a workaround. I trust Garmin's calorie burn algorithms far more than Strava's.

  • Whilst I'd like to see the calorie estimation improve in accuracy I doubt it could ever satisfy everyone simply due to wide ranging efficiency differences between people. I have determined my own ratio to Strava numbers (e.g. 60%) based on analyzing weight & food intake etc over time and now just go with that. Just watch different people running or cycling and you can see how efficient/inefficient some people are, and a calorie calculator will never account for that - It will always be for the "average" person.

  • For most runners of a given sex, height, and weight there are only minor differences in efficiency. So accurately calculating calories based on an average athlete would be fine for most of us. At least it would be better than the current numbers which are a bad joke.

  • I am new to Strava, this still an issue. I’ve been liking the app but the over-estimate calories are a turn off... I am used to other apps like garmin, Nike+  and Iwatch activity tracker... they are all consistent, but Strava is blows the ball out of the park. A 45 min run normally burns 450calories but with Strava I burn 600... 

Please sign in to leave a comment.

Didn't find what you were looking for?

New post