Cycling Calories Burned

coryrood

Ok, I have a question. I get HUGLY differing calorie burns for bike rides.

For Example:

Sunday I did 40.1 Miles
Avg. Pace was 13.8 MPH
Moving Time was 2:54:16
Elevation Gain was 1208ft.


Strava says I burned 1213 cal(30cal/mile)
MFP says 1806 cal(45cal/mile)
My Polar M400 with HR strap says 2202 cal(55cal/mile)

Multiple calculator's online vary anywhere in between.

My question is what would you go with? I'm not trying to lose weight, so an accurate assumption would be great to know how much to eat.

coryrood

Anyone?

Machka9

I calculate cycling calories burned like this ...

100 cal for every 5 kilometres (about 32 cal/mile)


If you do a search in google using words like calories burned 40 calories per mile, you'll find a few websites that use that calculation. 40 calories per mile works out to about 125 cal for every 5 kilometres, so slightly more than the calculation I use, but if you're not trying to lose weight, you might go with that. However, there is some thinking that 40 cal/mile might be just a bit high.

Of the three choices you've listed, I'd probably lean toward what Strava says.

czymom123

I just use my Garmin gps app that tracks my distance and average speed. It is pretty even with what MFP says. Sometimes within 1 calorie!

heybales

http://www.bikecalculator.com/

Strava has had terrible reviews for it's calculations for watts and calorie burn - from people that had watts measured on the bike and then compared what Strava estimated when the watts data was not supplied.
I know it's estimation of calories for my VO2max test using indirect calorimetry (gas exchange) for calorie burn was way low too. Running not as bad as biking though.
It may be because folks list their bike weight as if a contest to see how light theirs is compared to others, not what the real bike weight is, along with clothes and bottles and tubes, ect. (though I do include real weight being moved).

MFP has rather wide chunks of speed, as any public database would have - because the studies that provide the formula's don't have finer divisions. So outside the middle of the range, accuracy potential drops.
Plus MFP doesn't use resting metabolic burn as multiplier like the database is supposed to use, but converts first to use weight. And a 200 lb 20 yr male with 5% bodyfat does not have the same metabolism as a 200 lb 60 yr female with 50% BF.
And was that avg 15 mph because of terrible headwind and hard uphills?
Or was that avg 20 mph because you drafted in a group the entire time barely spinning?
Over long distance solo MFP can start being more accurate estimate.

The problem with simple formulas is the fact they don't take in to account air resistance at all - and once that kicks in as THE major factor about 14-15 mph, every 1 mph increase in speed is 4 x increase in energy expenditure to obtain it.
And tail wind does NOT balance out head wind, same way downhill does not cancel out uphill energy expended.

Running is at least a tad closer with those things balancing out, with wind not being the major thing anyway. And downhill being 55% of energy burn of equal pace going up same incline, so better chance to balance out.

Does your Polar have accurate Athlete Profile answer?
Does it have accurate restingHR stat.
You adjusted the HRmax if you ever saw one near start of workout?
Doesn't it have a VO2max self-test?
With those corrected stats - that can be very decent test - and very decent calorie burn estimate.

Or for your effort - 9 METS.
So take the MFP Mifflin BMR results / 1440 = calories per min burned basic metabolism.
9 x min (174) x that BMR level burn = total burn.

And that's not even taking in to consideration extra wind blowing, just air resistance for your speed.

https://docs.google.com/viewer?a=v&pid=sites&srcid=ZGVmYXVsdGRvbWFpbnxjb21wZW5kaXVtb2ZwaHlzaWNhbGFjdGl2aXRpZXN8Z3g6MjgyY2EyMzQzNWFlN2Q3OA

robertw486

heybales wrote: »

The problem with simple formulas is the fact they don't take in to account air resistance at all - and once that kicks in as THE major factor about 14-15 mph, every 1 mph increase in speed is 4 x increase in energy expenditure to obtain it.

I would have to think that the entire aero and resistance package would play a huge part in where a bike and rider really hits the wall as far as energy expenditure. Certainly a road bike with thin tires and lower bars would hold an advantage over a fat tire mountain bike, but as speeds increase the energy required would differ even more greatly. Even riding an old, heavy, fat tired mountain bike with my close to 200 pounds, I think the energy expenditure ramps up slower than that. If I could do a 16 MPH 5 miles and expend the same energy as riding a 13-14 MPH 20 miles, I could shred calories at a crazy rate doing shorter quicker rides.

I do agree with your points about variables, especially in regards to winds and/or drafting. Though it's fun riding in a group, there is huge energy to be saved doing so.



I think a major problem with any calculation is it is just that... a calculation with too many unknown variables. As with the bike calculator above, it may be great for steady state riding, but would not factor in all the acceleration events even if all the weights, rolling resistance, aero, etc was correct.

A GPS based tracker can account not only for averages, but acceleration events, top speeds, interval speeds, etc. For that reason I think GPS based would always have more potential to be more accurate. But in reality, short of a rolling weather station on your bike, neither of the above knows how the wind just shifted, the velocity of the wind that shifted, etc. When factored in with all the mechanical differences in equipment, the variation could be huge.


I've used Endomondo and just accept that nothing is 100% accurate. If it's off by X % for the specifics of my bike and riding conditions, it should always be roughly off that X %.

Mr_Knight

I use 2/3 of whatever Strava dumps out. That gets it pretty close to bikecalculator.com, most of the time.

heybales

robertw486 wrote: »

heybales wrote: »

The problem with simple formulas is the fact they don't take in to account air resistance at all - and once that kicks in as THE major factor about 14-15 mph, every 1 mph increase in speed is 4 x increase in energy expenditure to obtain it.

I would have to think that the entire aero and resistance package would play a huge part in where a bike and rider really hits the wall as far as energy expenditure. Certainly a road bike with thin tires and lower bars would hold an advantage over a fat tire mountain bike, but as speeds increase the energy required would differ even more greatly. Even riding an old, heavy, fat tired mountain bike with my close to 200 pounds, I think the energy expenditure ramps up slower than that. If I could do a 16 MPH 5 miles and expend the same energy as riding a 13-14 MPH 20 miles, I could shred calories at a crazy rate doing shorter quicker rides.

Well, it's not total energy expended that increases that fast, but the amount of increase for that 1 extra mph, it's just not linear. The air resistance increases with the square of speed, so that needs even higher power output compared to speed, power increasing with the cube of speed. But that increase gets added on to the expended amount due to all the other, so as you mention, doesn't ramp up that fast.
But it does explain why the increase from 17-18 is much harder than the increase from 15-16 say, other variables being equal.

And true about the bikes being used, some have enough of the other resistance that getting much above 15 just isn't going to happen much, and those knobby tires remain the major resistance say, or weight in general.
Or if on road bike you start tucking down anyway as speed increases, or going in to the wind, so the increase is automatically offset.

I think the well calibrated and confirmed HRM has best chance of accuracy, along with knowledge of when your body could be fooling it. Like dehydrated on 2nd half of ride say, and HR is elevated because of that.


Edit to add:
Was looking for graphs that I remember seeing long ago, and forgot the formulas.
Should have known they'd be in one place probably.

https://en.wikipedia.org/wiki/Bicycle_performance#Air_drag
https://en.wikipedia.org/wiki/File:CyclingPowerComponents.svg

Mr_Knight

The only way to really know is with a power meter.

heybales

Ya, that's better than HRM, but I view that as a big jump in expenditure and usefulness to average rider, so keep forgetting it is an option of really desired.

I'm waiting on Watteam so I can keep pedals and new crankset. Some of these recent companies are sure delayed. Can't ship the desired glue, good grief.

KittensMaster

50 calories per mile when 195 lb male goes > 17mph

That is the max ml number I ever use

Speed is a factor once you get beyond 15.

My garmin gave me the highest ever reading of 59 calories and hour when I did 20 miles at 20 mph when
I was doing my fast 20 ride.

MFP is always ridiculous on the high side.

sijomial

The most accurate numbers I get are from my Polar FT60 HRM which is customised to my tested VO2 max and max HR. When I "calibrated" it against an expensive power meter equipped indoor trainer it's almost identical.

The downside of using HR on the road is you get the issue that if you hit stop/start traffic the numbers will become inflated unless you pause the counter.

Runkeeper generally is reasonable for me, Strava is pretty useless (far too low), Garmin is low by about 10 - 15%.
Maybe Strava and Garmin are low as I'm not very aerodynamic - high riding position and not a typical cyclist physique? Who knows really.....

Mostly I just use my Garmin (Edge with HR strap) for convenience and add 10%.

Mr_Knight

heybales wrote: »

Ya, that's better than HRM, but I view that as a big jump in expenditure and usefulness to average rider, so keep forgetting it is an option of really desired.

I'm waiting on Watteam so I can keep pedals and new crankset. Some of these recent companies are sure delayed. Can't ship the desired glue, good grief.

Ah, yeah, the 4iiii situation....wow.

heybales

sijomial wrote: »

The most accurate numbers I get are from my Polar FT60 HRM which is customised to my tested VO2 max and max HR. When I "calibrated" it against an expensive power meter equipped indoor trainer it's almost identical.

The downside of using HR on the road is you get the issue that if you hit stop/start traffic the numbers will become inflated unless you pause the counter.

Runkeeper generally is reasonable for me, Strava is pretty useless (far too low), Garmin is low by about 10 - 15%.
Maybe Strava and Garmin are low as I'm not very aerodynamic - high riding position and not a typical cyclist physique? Who knows really.....

Mostly I just use my Garmin (Edge with HR strap) for convenience and add 10%.

I always use the Garmin auto-pause setting at I think 4-5 mph. Still some negative effect on avg speed and such, but that's what's happening. I don't have much problem getting out of city traffic and not having a stop except a mile or two along, then less.

If you received your VO2max HR data, or at least several stop points in there with HR, CO2/O2 usage, you can actually create your own Garmin metabolic profile and upload it to the device, just for the cycling aspect.
Doesn't add much except better calorie burn estimate, wish I'd left mine original for compare, since I manually calculate anyway based on VO2 test formula.

I really wish they'd just allow you to input known stat of VO2max, rather than calculate off amount of workouts and resting & HRmax values. Not sure what Edge uses though, if the same as 310XT.

heybales

Just got a reminder of why the Strava seems low calorie burn compared to database, calculation, HRM, ect.

When it looks at the hill profile and speed and mass and runs the calc's for watts, converted with time to Kjoules, it is ONLY for the work done on the bike, the energy expended to make the mass move the way you did.

Compared to every other method is the total calorie burn for that chunk of time - resting metabolism included.

So there's an extra 80 cal / hr to add on to their figure for me. Still a low value though having to do with wind presence or tires.

Makes me wonder now about the syncing of Strava with Fitbit - where Fitbit is expecting a synced workout to contain a figure of Gross calories burned for duration of time given to replace what was already there (which includes BMR) - and unless Strava adds that in - they are providing Net calories burned.

andrikosDE

heybales wrote: »

...

Does your Polar have accurate Athlete Profile answer?
Does it have accurate restingHR stat.
You adjusted the HRmax if you ever saw one near start of workout?
Doesn't it have a VO2max self-test?
With those corrected stats - that can be very decent test - and very decent calorie burn estimate.

This.
If your profile is correct, Polar has a very good algorithm for estimation of caloric expenditure.

robertw486

Lots of interesting and informative info to be had here, but also showing how complex the entire picture is.

Taking a quick look at some of my rides, the calorie burn seems way too high for what the formulas and statements here suggest. I've got rides showing calorie burns of upper 60s to lower 70s calories per mile on some rides. This seems very unlikely, but it would be fun to do an in depth analysis of how real ride conditions vs steady state conditions would work out. In my case the acceleration events are quite a few and there is very limited opportunity for anything resembling steady state riding.

When combined with traffic issues that include major slow downs and/or stops, I often have to click off several miles at a higher (16-18 MPH) pace if I want a ride to average even in the high 14s. I would think that the acceleration events take almost as much energy if not more at times than the steady state riding does. But then again I'm still 195 lbs, the bike is old and heavy, and the rotating mass with big knobby tires is high.

Either way, it sounds like my calorie burn shown by Endomondo is probably on the high side. As I've said before, I just accept error as error, and know that riding X miles helps drop some weight. I'm not weighing any food yet, so I'm sure I've got error on both the exercise and intake sides. Over time I'll probably try to nail it down to a smaller window of error.

robertw486

heybales wrote: »

Just got a reminder of why the Strava seems low calorie burn compared to database, calculation, HRM, ect.

When it looks at the hill profile and speed and mass and runs the calc's for watts, converted with time to Kjoules, it is ONLY for the work done on the bike, the energy expended to make the mass move the way you did.

Compared to every other method is the total calorie burn for that chunk of time - resting metabolism included.

So there's an extra 80 cal / hr to add on to their figure for me. Still a low value though having to do with wind presence or tires.

Makes me wonder now about the syncing of Strava with Fitbit - where Fitbit is expecting a synced workout to contain a figure of Gross calories burned for duration of time given to replace what was already there (which includes BMR) - and unless Strava adds that in - they are providing Net calories burned.

Being I'm not using a Fitbit I'd actually prefer something that reports only net calorie burn. My BMR is already accounted for here in my calorie goals.

It's actually surprising that of the apps available, nobody seems to have done much to allow for different variables. Using Endomondo I couldn't even get a straight answer on why various modes vary calorie burn... is commuting cycling somehow different from sport cycling? I don't know, and I don't think they do either.


With GPS, it would seem they could have much more accurate calculations and allow a user to customize for their equipment. Until then, I'm just going to have to watch the numbers and adjust the best I can. I just hate the added effort to find accuracy.

KittensMaster

I'm thankful I have a paved rail road track 100 miles to ride

I got in 115 miles in Friday thru Sunday. Averaged 17.5 with maybe 5 stop or slows to cross a road.

I called it 50 calories mile such may be slightly low.

MFP wants to give nearly 80. Garmin is around 59.

So with occasional drafting on a good racing bike, 50 seems right.

robertw486

KittensMaster wrote: »

I'm thankful I have a paved rail road track 100 miles to ride

I got in 115 miles in Friday thru Sunday. Averaged 17.5 with maybe 5 stop or slows to cross a road.

I called it 50 calories mile such may be slightly low.

MFP wants to give nearly 80. Garmin is around 59.

So with occasional drafting on a good racing bike, 50 seems right.

I'd kill to have a solid 8 or 10 miles without major traffic issues. Around here, it doesn't exist unless I want to put the bike on a rack and drive there.



I did an interesting experiment today. My intention was to run a live bike using two phones, one with Strava and the other with Endomondo. Though the ride was cut short due to a puncture, I used that to further check out the stops and acceleration events. I actually walked the bike a short way at one point, then the last 3/8 mile or so. So really the event logged 3/4 or maybe a mile tops of me walking or jogging the bike. At one point I stopped to add air and paused the apps, the second air stop I let the apps keep running. Combined with the traffic that is close to home, the pace is nothing, but there were several minutes of dead stopped. I hope next time I can get to where I'm not dealing with as much traffic and set a few quicker miles to check the differences out.


Strava distance logged 9.7 miles
Endomondo distance logged 9.65 miles
Cycling computer distance logged 9.92 miles (the computer is cheap but roll out was measured and set)


Strava average 10.9 mph
Endo average 9.33 mph
computer average 11.1 mph (auto pause by default, apps did not pause)


Strava max 21.5 mph
Endo max 21.68 mph
computer max 22.1 mph


Strava calories 536
Endo calories 626
If you factor my BMR (I'm 193 lbs) things end up really close between the two apps.



Observations:

Strava is much more accurate on altitude, very close to precise I would even say. Endo frequently shows elevations that don't exist in this city.

Both apps show different things in the free versions. Endo shows pace and speed for various distances, Strava does not.

Strava does estimate watts, and it seemed very in line with the curve on hard acceleration. Though I doubt my max was actually 1,146 watts that could have been influenced by grade. Strava allows you to select a bike type, but I'm not sure if this factors tire sizes or weights at all. Maybe at some point I'll run two instances of Strava and set one bike up as a lightweight road bike to see how it calculates calories and/or watts.


I also did some calculations based on acceleration energy used. Though rough guesstimate numbers for wheels weights and such, accelerating me and my bike from 0 to 18 MPH takes roughly the energy it takes to cruise at 20+ mph for 12 seconds. Even a slow acceleration to 12 mph (0-12 in 12 seconds) takes as much energy as riding at 15 MPH steady state. So for anyone riding in urban areas or anywhere that has more stop and go, the energy expenditure can go up fairly quickly. In less than 10 miles I had at least 15 acceleration events that were showing slow downs to below or well below the average pace. Which, considering the logged stopped time, was slowing down to about 1/2 or less of my normal pace even on a fat tired mountain bike.

Mr_Knight

robertw486 wrote: »

Strava allows you to select a bike type, but I'm not sure if this factors tire sizes or weights at all.

Yep, it does.

robertw486

Ok, so I did another test today Strava vs Endomondo. Today I picked out a stretch away from the major traffic slow downs, though I still had 1 light that created a complete stop, and several instances below 7-8 mph. The idea was to see if the variation in the apps was more due to steady state or acceleration variables. My intention this ride was to ride closer to steady state if possible, and minimize major acceleration events.

On the way out to the turn around point, I tried to keep pace between 15=16 mph. On the way back I tried to keep pace at 17-19 mph, with a quick blast of 20+ towards the end. I had some cross roads and one light that altered the average pace, but when moving kept things in a fairly small window.

As before, the same trends in distance, speed, and calories existed. Keep in mind that with two different types of phones, this might be a matter of how efficiently the phone pings the GPS birds, so later on I'll test that theory out. It does seem that resting metabolism is included in the Endo app, but not in the Strava app. At least in two tests, subtracting that brings the two apps fairly close in calorie count.


But, even more interesting to me personally, after looking at the app numbers I punched the numbers into the bikecalculator.com link that Heybales had provided upthread. I then took a look at the acceleration events I couldn't avoid, and factored that in using another calculator I found. As the numbers got closer, I even broke down the lap times per mile, since the stops/slows had altered things farther from steady state riding than I had intended.

And when all was said and done, I found that either of these two apps actually gave some fairly accurate numbers for the ride I did when compared to that biking calculator. The difference in reported calories does in fact seem to be related to one app adding in BMR, while the other does not. Looking back at some longer and varied rides, this also explains why Endomondo gives such a great calorie burn shown for long, slower rides.

heybales

I don't know if you've ever heard of SportTracks - great Windows program for uploading GPS/HR data files and going over the data every which way.

Recently broke down and bought ($5?) a Powerwatts app add-on for it, along with weather for getting wind data from wunderground stations.
http://www.sporttracks-plugins.com/SportTracksPlugins3/index.html?content=http://www.sporttracks-plugins.com/SportTracksPlugins3/MyPlugins/Weather%20Plugin/manual.html#Settings_Page_Weather

Rider weight, bike weight (with all accessories), tire weight for the accelerations, and that wind data and humidity since it effects things too.

For Saturday's ride with little to know wind - matched up with calc you found, and decently close to Strava - they both do NET calories.
But then again - these are well known formulas that any site like that will use. It's a matter of how accurate do they attempt to get.
http://www.analyticcycling.com/AnalyticCycling.html
Even this app has you select a ride profile (hands on hoods, drops, tops), and that really doesn't apply all the time.
But they had several people test it against their Powermeters, and it was incredible how dead on it was over the course of the ride.
Sometimes you could see one line above the other for a period - but the profile almost exactly matched. So that was likely when either weather data was incorrect or rider profile didn't match selection.

Like today's ride had gusts up to 25 mph that were not accounted for on the map data, but I sure felt them, mainly as head wind messing up rhythm, but also tailwind that helped my avg 25 mph for a 7 mile stretch.
But it estimated an avg watts of 340, and while I was a whole lot deader than normal because of heat and effort, I know not that good, but gusts of head and tail would probably cancel out a tad more than that huge increase over normal.

I think the reason is because I got a lot of traffic assist drafting going in to the wind, all traffic was positive when it was there, so my average speed for 16-25 mph mainly headwind was not entirely my own effort.

And there I'd thought I'd found a nice solution and some extra good data. But fooled the formula's again!

Oh, and Strava with no consideration of the wind was WAY off. Not sure if their premier version does or not, I've seen stats on other's rides that have premier, and it has some weather data.
I'll have to go back and compare a ride we did together, though I've got probably 20 lbs on them.

edit - yep, just looked at stats for the mainly headwind for first 25 mile, avg 15.9 mph, where I was getting varying levels of traffic assist.
Avg 468 watts. I'm sure if I did that speed with the 16 mph headwinds it has down, up to 25 mph gusts, and riding on hoods, that would be true.
But I'll bet it's correct when I had tailwind on 7 mile stretch, avg 25 mph - 173 watts.

robertw486

Did a quick yet telling test today with Strava vs Endomondo.

Turned the apps on, set the two phones side by side on the desk, and didn't touch them. In just shy of 70 minutes of nothing happening...

Endomondo showed a calorie burn of 159 calories

Strava showed a burn of 5 calories


For some reason when I first hit the "record" button on Strava, it showed a speed of like 20mph. Maybe I bumped the phone a little and the GPS picked up on it.

At any rate, this test seems to show me that Strava is not recording resting metabolism at all. Good or bad depending on how you track your calorie burn.

heybales

I forgot to comment on some testing I did regarding their elevation changes/gain that is used in their computations of watts, then converted to KJ and then calories.

I had running route over some rolling hills, net change of about 5 ft down, but up and down a whole lot more in the 1 mile.
I used their site to look at the top of the hill and the bottom for elevation and subtracted, and added up all those changes and came up with total elevation gained. So what I had to move my mass up.

They were about 1/2 that amount.

Spot checked some rolling bike hill segments too - incorrect total gain.
Shoot, starting looking at some segments at my PR's, and noticed that the elevation change as shown moving the mouse from bottom of hill to top doesn't match when they have as the stat of elevation change.

So I really don't trust them now, don't know where they get those figures from.

Besides the fact their other issue is not knowing about wind, or lack of wind resistance when in pace line drafting.
I noticed some huge average watts on some team riders on there, and then of course realized that they all stick together and drafting - they don't have nearly those watts.

And for my interest, obviously not the calorie burn either derived from those watts.

But I've found several sites that deal with watts, and by extension calories.
They all do what you found with Strava, leave the RMR calories out, merely looking at extra energy expended moving the bike, which makes sense.

robertw486

Around here we have no real terrain, so any testing dealing with altitude would be tough. In the side by side actual rides though, Stava was much closer to Endomondo in dealing with total altitude changes. Endomondo was coming up with altitudes not even possible in terrain this flat.

For now I'm going to stick with Strava for my biking.

As for the details of wind and such... well really short of having a computer with a weather station that's a lot to expect from an app. Though we do get some wind around here, hopefully in the end it's usually a wash or calories not counted.

The test was really shocking. Anything that logged almost 160 calories in 70 minutes for a stationary device is rather scary. I might test Endomondo in some other exercise modes to see what it does.

heybales

Ya, that is well above RMR level - very strange.

rbakedq

One way to check the accuracy, or really the lack thereof, of any of the apps is to do an out and back on some hilly terrain. I use Runkeeper, Endomundo, and Wahoo fitness rarely does any of them show the same elevation on the return as going out.