I use a few apps and devices to track my activity and calorie in take. It helped me lose weight last year and as I’m starting to prepare for sportives this year I’m using them again. However I’m confused, when I’m using the apps / devices to calculate calories burned during exercise they give different estimates. For example this morning on a 16 mile ride Strava estimated I used 453 calories, while my Garmin estimated 1019. Both have had the same personal data (age, weight, height) entered and updated at the same time.

I know they are only estimating the calories used but it’s quite important that I’m using the estimate that is most realistic.

So, which is likely to be closest to reality.

the lower 😉

I’ve never tried Garmin but I used a Fitbit HR for a few months, like OP Strava would usually given a figure about half what Fitbit would – this seemed like great news someone a terminally hungry as me, and you’d bet on it being the more accurate – after all Strava just uses gps / map data plus weight – no idea of how hard or easy the surface is, roots rock and other technical features you have to deal with or indeed how hard you’re blowing coming down – whereas the Fitbit knows heart rate so it knows how hard your body is working.

But despite under eating by supposedly 1000-3000 calories a day, I bit a bit of weight on…

stick to the lower one.

I have been going by the Strava estimate. But while I’m losing weight I’m limiting myself to net 1000 calories per day so if it’s under estimating by half then it might cause problems for me.

Strava estimated I used 453 calories, while my Garmin estimated 1019

Around 500 calories for an hours cycling seems about right so your Garmin is way out. My garmin is reasonably ‘accurate’ so not sure why yours is so far out.

But despite under eating by supposedly 1000-3000 calories a day

You must be burning some amount of calories if you under eat by 3000 calories a day!

do your devices measure heartrate as well?
big factor

the measurement will always be a rough estimate

At the moment I’m not using a HRM or powermeters. So the Garmin is just using the GPS to track my effort.

I know that to get an accurate reading I need to use a HRM and I have ordered one.

Thing that annoys me about all these devices apps is that, as far as I can tell, they make no distinction between road riding and mountain bike riding. You can get e-bike, indoor bike and gawd only knows what other kind of bike options, but the on/off road distinction seems to escape them.

If I’m out mud plugging on my singlespeed for four hours, I’#m definitely burning more calories than if I was pootling along on the tarmac at 9 miles an hour, as it seems to assume I am if I tick the only bike or ride option on their list.

Thing that annoys me about all these devices apps is that, as far as I can tell, they make no distinction between road riding and mountain bike riding

That’s why you need to use a heart rate monitor. There’s no point in the device making a distinction between different types of cycling as without heart rate data it’s meaningless. It’s work rate that matters, not what you do the work on.

do your devices measure heartrate as well?
big factor

It is not as simple as using heart rate data = accurate. Have a read here:

https://www.dcrainmaker.com/2010/11/how-calorie-measurement-works-on-garmin.html

If you CBA to read and process all that, I choose the lower number and round down.

Calories are a measurement of energy, HR isn’t a measure of energy expended, power is.

Some reading here.

I’d always wondered but not cared enough to look it up. Thanks.

5) Speed/Distance Algorithm: This is the most basic method of determining calories, as it is only used when a heart rate strap is not enabled/used (default). Given the lack of heart rate data, the unit will simply use speed/distance, as well as the weight you entered in the device setup. The reason this is less accurate (65-80% accurate) is that it can’t differentiate how much effort you’re expending to travel a given distance – which while less important for running, is quite important for cycling. For example, if you’re coasting down a 7 mile descent, you’ll burn virtually no calories compared to ascending the same mountain. This speed/distance algorithm does not consider or evaluate the impact of elevation change – primarily due to concerns the team had about relying on GPS-based elevation to determine calories.

I’d always assumed that for cycling it took account of elevation at the very least. So my assumptions would be it’s vaguely useful for flattish road rides. Pretty much useless if you’re climbing hills and putting some effort in, vastly underestimates mountain biking. Isn’t bad for running.

That’s why you need to use a heart rate monitor.

IMO even with HR data it’s still a estimate. The calculation knows nothing about your metabolism or other factors which will be affecting your HR, it almost certainly doesn’t even know accurate HR zones unless you’ve had them measured & entered into the calculation.

I’d always assumed that for cycling it took account of elevation at the very least.

The question of whether elevation is being measured accurately is a whole other kettle of fish!

Personally I can’t be bothered with that and just adjust accordingly, i.e. if weight is creeping up eat a bit less or cleaner, if I’m suddenly doing more exercise for a period of time just eat a bit more. Being fairly routine in what/when you eat and how you exercise is helpful if you do it this way!

Personally I can’t be bothered with that and just adjust accordingly,

I’m not someone who needs to track calories but it’s always interesting how few calories you burn when exercising. It’s easy to see how people trying to get fit/lose weight overcompensate by eating more and put it on. 14 minutes hard work on a rowing machines yesterday morning (which has a power meter so fairly useful) was about 300 calories from memory.

IMO even with HR data it’s still a estimate.

Yeh of course but it’s going to be a lot closer with an hr than without.

it’s always interesting how few calories you burn when exercising

…and how many are in certain foods.

I watched some nonsense on the telly about that the other day, a group of ladies did 20 odd mins of brisk walking on the treadmill and felt that they had “earned” a Starbucks cappuccino and a lump of cake on the way home. Despite the exercise, they were actually in the red for about 500 calories.

Gary_M – Member

Thing that annoys me about all these devices apps is that, as far as I can tell, they make no distinction between road riding and mountain bike riding

That’s why you need to use a heart rate monitor. There’s no point in the device making a distinction between different types of cycling as without heart rate data it’s meaningless. It’s work rate that matters, not what you do the work on.

I do, though.

Strava, Garmin and myfitnesspal all still seem to think I’m pootling along a flat stretch of tarmac on my raod bike in terms of the number of calories it thinks I’ve burned.

Seems to be a very blunt estimation based solely on your average speed.

It’s something they all need to sort out.

I watched some nonsense on the telly about that the other day, a group of ladies did 20 odd mins of brisk walking on the treadmill and felt that they had “earned” a Starbucks cappuccino and a lump of cake on the way home. Despite the exercise, they were actually in the red for about 500 calories.

It’s quite disturbing how many people think 20 mins ‘brisk’ walking actually counts as exercise.

Edit: For the purposes of calculating calories, use the estimated average power on Strava, multiply by 4 then multiply by hours. Not perfect, but as estimates go its about as good as you are going to get.

Gary_M – Member
it’s going to be a lot closer with an hr than without.

It’s probably more accurate the other way around.

If I ride up a 100m hill, slow enough that drag isn’t a factor, and me and my bike weigh* 1000N then

e=fd = 1000 x 100m
= 100 kJ
We know the body is baout 21-25% efficient on a bike, and there are roughly 4 J in a calorie, so I’ve burnt off 100 kcal.

I’m a fat unfit chubster, my HR was 185 the whole way to the top.
Chris Froome rode next to me, with panniers filled with sand to match by weight and barely got above 100bpm. My HRM tells me I’ve burnt 200 calories, his says he’s burnt 3.

In reality we both burnt 100 kcal +/- a small degree of accuracy.

So when Strava knows the height gained, and your weight, and has an approximation built in for drag, it’s probably not that far off. The only thing it can’t see is mud and knobly tyres, although I suspect it makes a fair guess based on the bikes weight (30lb = MTB, 25lb = XC bike, 20lb = CX bike, 16lb = road bike) and it probably has a fair correlation to work with given the amount of data it has access too (thousands of people, on thousands of bikes with thousands of power meters).

And then there’s the concept of “metabolically appropriate heart rate”. Riding down a hill at a trail centre takes very little pedalling and therefore burns few calories (relative to say, going up the hill). But your HR could be sky high due to all the adrenaline, the HRM doesn’t know that.

*yes I said weigh, not mass

But despite under eating by supposedly 1000-3000 calories a day

You must be burning some amount of calories if you under eat by 3000 calories a day!

I wasn’t the fitbit was wildly inacurate, I was getting an extra 1500-2000 for a lap of Cwmcarn it’s less than half that.

I get through 2800 a day just doing my normal day to day stuff according to the Doc, I eat (at the moment) 1800 a day.

So when Strava knows the height gained, and your weight, and has an approximation built in for drag, it’s probably not that far off. The only thing it can’t see is mud and knobly tyres, although I suspect it makes a fair guess based on the bikes weight (30lb = MTB, 25lb = XC bike, 20lb = CX bike, 16lb = road bike) and it probably has a fair correlation to work with given the amount of data it has access too (thousands of people, on thousands of bikes with thousands of power meters).

Not sure that’s going to work though – on most downhills on a road bike you’re pedalling, a lot more than you would on a mountain bike. (Which is why pro mtb racers also use a road bike, it’s a lot better for base fitness). I think you really need to add HR in too.

Chris Froome rode next to me, with panniers filled with sand to match by weight and barely got above 100bpm. My HRM tells me I’ve burnt 200 calories, his says he’s burnt 3.

The algorithm for Garmin HR to EE estimation has a max error of 10% so your ‘example’ is pretty meaningless IMO.

Calories are a measurement of energy, HR isn’t a measure of energy expended, power is.

Find measurements based on PM data work really well for me. So if I take the cals estimated by TrainingPeaks or Strava based on power data, add about 2400 cal to that, that’s roughly where I need to be daily to maintain a stable weight. Averages out to somewhere between 3800 and 4000 a day over a week.

HR or GPS based estimates are utter bobbins.

Not sure that’s going to work though – on most downhills on a road bike you’re pedalling, a lot more than you would on a mountain bike. (Which is why pro mtb racers also use a road bike, it’s a lot better for base fitness). I think you really need to add HR in too.

Yes, bu then Strava still knows the gradient, and the drag, and your speed, so knows how many watts of pealing you’re doing.

e.g. im doing 20mph down a hill, Strava knows it takes 300W to do 20mph on a 16lb bike (because it assumes a 16lb bike is a road bike and has a fair idea of the CdA), it knows the gradient is giving me 50W, therefore I’m pedaling 250W. Chris Froome has dissapeared but Lance Armstrong is next to me, his HR has now dropped so low that the team medic is telling him to hurry up so his heart doesn’t stop due to all that extra thick blood.

HR is a bit like using revs to judge the speed of the car. You can’t, because the gearbox is in the way (unfit geezer is 1st gear, Froome is in 6th).

The algorithm for Garmin HR to EE estimation has a max error of 10% so your ‘example’ is pretty meaningless IMO.

Nope, i might go to a spin class later, i’ll put my garmin in my back pocket if you like, want a sweepstake on how many calories I’ll burn in 30min? The reality is probably 250 or so, I bet it will read 500+.

Yes, bu then Strava still knows the gradient, and the drag, and your speed, so knows how many watts of pealing you’re doing.

Strava doesn’t differentiate between road and mtb though…

e.g. im doing 20mph down a hill, Strava knows it takes 300W to do 20mph on a 16lb bike (because it assumes a 16lb bike is a road bike and has a fair idea of the CdA),

It needs to estimate CdA and Crr, both of which have can huge variance from rider to rider and road to road. Makes assumptions about air pressure and windspeed. Your power for that scenario in summer (light winds, thinner air, lighter more aero kit) will vary significantly to your power in winter, but the strava estimate will remain the same.

Off road, CdA less of an issue but Crr can be massive.

Strava doesn’t differentiate between road and mtb though…

Neither can* your HRM differentiate between you and Chris Froome which is the far more important factor.

Similarly it doesn’t know if you’ve got a cold. Get a cold and your HR change, but the amount of work done doesn’t, and if you can hit the same power outputs then neither does the calories.

*it might, i think mine lets you guestimate your fitness from 1-10, so it depends how accurately you can guess/lie about your fitness.

It needs to estimate CdA and Crr, both of which have can huge variance. Makes assumptions about air pressure and windspeed. Your power for that scenario in summer (light winds, thinner air, lighter more aero kit) will vary significantly to your power in winter, but the strava estimate will remain the same.

Off road, CdA less of an issue but Crr can be massive.

Yup, but still less of a difference than you’d get plugging me and Froomey into the same HRM’s and expecting the same calorie numbers.

And as I said, I suspect it comes up with an overall C to account for CdA and Cr based on the fact it has a lot of data for people on heavy bikes with high overall C values, and 15lb bikes with low ones.

I’m not saying STRAVA is accurate, I’m saying if I had two numbers and one was based on heart rate, I’d take the other one.

Strava doesn’t differentiate between road and mtb though..

Not 100% sure, but I think it actually does.

Nope, i might go to a spin class later, i’ll put my garmin in my back pocket if you like, want a sweepstake on how many calories I’ll burn in 30min? The reality is probably 250 or so, I bet it will read 500+.

Does your Garmin use the Firstbeat algorithm? White papers (admittedly their own) claim 7-10% accuracy. What are you using to get the ‘reality? bearing in mind crank or hub based meters will measure energy expenditure through the bike only, not the whole body.

Not 100% sure, but I think it actually does.

Online it doesn’t, there’s cycling and running.

Online it doesn’t, there’s cycling and running

Why is it then that when I ride ~12-13mph on forest tracks on any of my MTB’s then Strava comes up with the same average power as ~15-16mph on purely road routes? I’ve checked bike weights are entered correctly, so its not that.

Edit
Strava themselves say they do: https://support.strava.com/hc/en-us/articles/216917107-How-Strava-measures-Power

Strava themselves say they do: https://support.strava.com/hc/en-us/articles/216917107-How-Strava-measures-Power

My mistake, I stand corrected – it’s been so long since I last got a new bike that I’d completely forgotten you had to define its type. Obviously a reason to get another 🙂

Genuine question.

Am I not going to burn more calories than Chris Froome even if we are doing the same exercise? My heart going at 190bpm must require additional calories to power. My lungs and shoulders will be heaving and my body generating sweat. All this must burn extra calories even if the power going through the cranks is the same?

Am I not going to burn more calories than Chris Froome even if we are doing the same exercise?

Yes, I’m assuming he’s considerably more efficient than you (that is, of the amount of energy he’s expending, a considerably greater portion is going to the pedals.)

Have a look at this article…

http://home.trainingpeaks.com/blog/article/how-accurate-is-that-calorie-reading

So my heart rate is a more useful indicator of the calories I am burning than a power meter?

What are you using to get the ‘reality?

Assuming an FTP around 200W, the max that person could burn is 360 kcal if they rode on the redline for the full 30 minutes. So knock someoff for the warmup and cool down, and reduce the average of the middle bit a bit as short intervals with rest between them still have to average less than FTP, and 250 kcal seems about right.

The idea that an average cyclist is burning 1000 calories or more an hour is just ludicrous.

Am I not going to burn more calories than Chris Froome even if we are doing the same exercise?

Heart beats aren’t burning many calories, it beats 70’ish just sitting still when the whole body is only burning double digits per hour.

Arms, shoulders, etc. Yes if you flail around all over the place they’ll burn calories that a power meter or strava can’t see. But your legs are moving 70-100 times a minute too, and even a weakling could push their bodyweight with one leg, comparatively the strength of your shoulders and arms (and thus work done) is negligible. Even rowing which is about as full body as you can get, is 6x more legs than arms.

So compared to legs everything else is within the margins of error for HRM based calorie calulations, even if you take those garmin numbers as gospel (and you’ve not lied to it about fitness, and it’s the latest and greatest algorithm, and it’s had enough use to learn how your heart responds to efforts).

So my heart rate is a more useful indicator of the calories I am burning than a power meter?

Think of it this way. The heart is a pump. It’s pumping blood with oxygen in it.

My blood will have less oxygen than Froome has because my haemocrit level is lower.

My heart will pump less blood than Froome because it’s smaller.

The heart only (normally) pumps as fast as it needs to to supply the energy needs of the body.

So if my HRM says 160bpm, and Frooms say’s 160bpm, they’re going to come out with a similar calorie numer over an hour. Yet Froome will have pumped a lot more energy (and ridden far faster) than me, and thus burnt more calories.

But if we both sat on a Wattbike and did 250W, within a percent or two we’d be burning the same calories per hour.

Assuming an FTP around 200W…….250 kcal seems about right.

So basically guessing then, but more than that what you are guessing is your cycling power output, you aren’t taking into account the other energy that the body is ‘using’ during the exercise – you get quite hot spinning, where does that heat come from? Average calorie burn for an adult male is ~2500 calories per day, assuming an even distribution that’s 100 an hour that isn’t included in your FTP based guess but would be included in the HR based calculation.

So if my HRM says 160bpm, and Frooms say’s 160bpm, they’re going to come out with a similar calorie numer over an hour.

Not necessarily, the algorithm uses more than just the HR value to calculate calories burned – if you have the same mass, max HR, min HR, VO2 max etc. then yes the 2 will be similar.

Yet Froome will have pumped a lot more energy (and ridden far faster) than me, and thus burnt more calories.

Again, you are just looking at the cycling output power – not the system as a whole.

But if we both sat on a Wattbike and did 250W, within a percent or two we’d be burning the same calories per hour.

I think Froome would burn a lot less. Think of it another way, I go out for a run with Mo Farah, we do an half an hour at 6min/miles, I’d be knackered but Mo would be barely out of breath – my calorie burn would be way higher than Mo’s because he is much more efficient than me and isn’t operating anywhere near his limit, whereas I’d be on mine for the whole time. If we both went out and ran at 95% of our max HR then I would expect our calorie burns to be similar but I wouldn’t see him after a few hundred metres.

If we both went out and ran at 95% of our max HR then I would expect our calorie burns to be similar but I wouldn’t see him after a few hundred metres.

I don’t believe that’s correct. Going back to a cycling example, two riders, one with a high FTP and another with a lower value. If both are putting out the same amount of power the perceived effort will be very different, but the amount of calories burnt will be roughly the same. If both are riding at their FTP, hence similar PE, the one with the higher FTP will be burning more calories.

I don’t have a huge amount of data but as a comparison; for my last marathon I burned ~22 calories per minute, Ryan Hall (US Olympian, I’m pretty sure his VO2 max is much higher than mine) burns ~20 calories per minute. I weigh about 8% more than him, so if the numbers are normalised for weight they will be fairly similar.

for my last marathon I burned ~22 calories per minute

Says your device that makes up fairytale numbers.