Mithrandir

Or, at least I did in whatever la-la land that MapMyRide lives in.

https://www.mapmyride.com/workout/168165417


This is utterly ridiculous. In fact I would say it's downright irresponsible. Had I not known any better, I probably would have used this information to justify a "reward" meal of some sort, thus completely obliterating any progress I'd already made.

I know that most sites exaggerate the number of calories burned, but this is by far the most egregious example I've ever seen. Sigh.

FrenchFit 

Funny, unless you're a stone cold pro I suspect it's pretty unusual outcome to burn more than 1000 calories an hour on any sustained basis. I've been doing 40 minute runs, I bet I'm lucky if I burn 400 calories...exaggerating calorie burns is good fun regardless, I approve. So...congratulations, you are a cycling stud.

chefisaac

if you want a better way of knowing what you burn, get a HRM.

chasm54

That won't do it for you. It will provide a guide, but conditions - temperature, fatigue and so on - make HR an unreliable method of assessing your calorie consumption. You'll find that different HRMs give widely different estimates of calorie consumption - even different models from the same manufacturer. My old Garmin 305, for example, typically gave a figure about 30% igher than the Garmin 500 I replaced it with, and even the 500 tends to be on the high side.

At my lactate threshold I burn about 15 kcal per minute. So 900kcal per hour is about as hard as I can go - and then only for one hour.

chefisaac

Notice I said "better way". HRM versus computer based no info type of estimation. For me, it is HRM.

OiS

Really? What can give you a more accurate calorie burn than your HRM? I thought that was the most reliable.

Incidentally, my ride from yesterday was about the same time, 3hrs 14mins with 96.3km traveled. My HRM showed cals burned as 2391.

krobinson103

https://www.endomondo.com/workouts/lVF25LmgVRA

According to endomondo I burned over 5000 calories on yesterdays ride. Yes, it was hot, yes I was pushing hard for most of it, and yes no matter how much I ate/drank on the ride I still felt utterly drained on the last 20km. In fact my legs are still telling me about it this morning. But, I still can't credit a ride of 5 1/2 hours would burn that much energy.

Mithrandir

HRM-based calculations won't work for me. I'm on blood pressure meds that dramatically lower my heart rate (I'm trying to convince my doctor that I don't need them. My blood pressure at home is almost always near 120/70, but spikes up to 160+/90+ at the clinic. I have no idea), so it usually says I'm burning 200-300 calories an hour, which is much lower than what I deem realistic (500-800/hr).


... I need a power meter.

Mithrandir

Power meters are generally considered the most accurate. Heart rate can be high even if you're not expending any energy (ie: an exciting downhill, a hot day, dehydration, dog chasing you (happened on todays ride! (hey, a parenthetical inside a parenthetical!!)), etc). In terms of physics there's a direct linear correlation between power expended and calories burned; heart rate is much more variable and thus has much higher error numbers.

Generally speaking the human body is 20-25% efficient in terms of calories burned to joules reaching the pedals. One calorie (not calorie burned, but unit of measurement) is equal to 4.184 joules, which roughly cancels out the 25% efficiency.

So you can ballpark calories burned by calculating Kilojoules expended during a workout, which is Average Watts * Time. I did a 40 minute spin last week at 208 watts average, which works out to 499.2 kiloJoules reaching the pedals. Say we have 25% efficiency, that means my body actually produced 1,996.8 kJ of energy. Multiply that by 1/4.184 to get 477 kiloCalories (or "nutritional" Calories as we refer to them on food packages) burned, which puts me about 715 calories per hour. I don't actually believe my body is 25% efficient, but for me it makes more sense to assume I'm more efficient. For example if I assumed I was only 20% efficient, then that number turns out to be 596 calories, or 894 calories per hour. For the sake of weight loss calculations, assuming I burned less calories (ie: 477 bs 596) will help me in the long run. If I really am that efficient then I don't have to worry about overeating. If I'm not that efficient, then I'll lose weight faster. In theory. For professional athletes though they'll probably need the accuracy and can't afford to under-estimate calories burned, lest they bonk during a 3-week stage race when they run out of energy.


The sources of error are going to be:

1) the power meter itself, typically up to 5% (some claim 1%, but those are marketing materials that should be taken with large grains of salt)
2) Any drivetrain parts that come before the strain gauges in the power meter. If it's a hub-based PM then the bottom bracket, pedal spindles, derailleurs, and chains can all affect accuracy if they're not well maintained, because they will suck power out of the system before you even get to the hub.
3) This is the big one: not knowing how efficiently your body converts energy into motion. 20-25% is actually a fairly large range, and as far as I'm aware the only way to calculate this is within a laboratory environment, which can be quite expensive. But if you need extreme accuracy this might be worth it.


So there you have it.

CommuteCommando

I am happy with the results I now get with a Garmin Edge 500, and HRM. If you are expecting 0.00% error with any kind of measurement, dream on. Mapmyride, RidewithGPS and others are great for route planning, and I use RWGPS almost exclusively. I started tracking on a smart phone app. The Garmin seems to be a lot more accurate with altitude, which leads me to trust it's calorie calculator a little more.

goldfinch

Some ammo: https://www.slate.com/articles/health...th.single.html

TrojanHorse

I just rode a century yesterday with 4,100 feet of climbing and averaged about 15.6 mph. 6:27 moving time, 8:00 overall and I do have a HR monitor. Actually it's interesting, I spent 32% of my time in zone 4 or higher.

Garmin says 4,895 burned, strava says 3,726 burned.

I don't rely on either of them.

Mithrandir

I don't think anyone is naive enough to expect 100% accuracy. I simply wanted to express my outragous disdain for MapMyRide's 2,552 calories/hour calculation.

I do believe that a power meter will be a lot more accurate than a heart rate monitor, especially in my case. That being said, it's not the primary reason for getting one, as I'm trying to transition into doing power-based intervals on my bike instead of just the trainer now.

Neil_B

Wired Magazine on fitness trackers and why calorie counts vary:

https://www.wired.com/playbook/2012/08/fitness-trackers/

Helmet tip to Instapundit for the link.

Mithrandir

I did an interesting ride on Thursday, my 2nd day back on the bike since my accident. My calculations showed:

Zone 1: 0 min
Zone 2: 1 min
Zone 3: 2 min
Zone 4: 20 min
Zone 5: 51 min

Granted with my meds it's hard to calculate what my zones should be, but I'm using the maximum HR I've obtained while on them and my resting HR as the two boundaries.

That's 96% in Z4+. Not sure if that's dangerous though. I was busting my ass trying to keep up with a much fitter group than I. (I succeeded!)

Anyhoo:

MapMyRide: 3668 calories, 2,770 calories/hour
RideWithGPS: 1480 calories, 1,118 calories/hour, 311 watts
Garmin: 914 calories, 690 calories/hour
Endomondo: 914 calories, 690 calories/hour
Strava: 910 calories, 687 calories/hour, 187 watts


So I have no idea. It is interesting that Garmin, Endomondo, and Strava are so close together, but as others have said they get wildly varying numbers from those sites. My first reaction was that they use the same algorithm, but that can't really be the case if other people have varying numbers.

Either way I believe the ~690 cal/h numbers are low. Above I calculated with a reliable wattage number 715 cal/h for a spinning session wherein my heartrate only entered Zone 5 for a total of 3 minutes and stayed almost entirely in Zone 3. This I believe is a consequence of the beta blockers; my heartrate was much lower than it normally should have been and therefore Garmin etc think I burned less calories. There is no way I burned less calories per hour in that workout than my earlier spin session. I could hardly walk the next day and had to take a break.

E.S.

None of these numbers matter. Just ride.

Mithrandir

Oh I am. First week back and I hit 112 miles and I still have 2 more days to go before a full 7.

But the numbers do matter to me. I'm looking for a way to get a better handle on my diet/exercise requirements. At the moment I'm losing about 0.5-1 pounds a week and at this rate it's going to take bloody forever to drop the 180+ pounds I need to lose. 3.5 years in the best case, 7 in the worst... assuming no plateaus or setbacks. Plus, winter is coming. I have a very bad habit of losing track of my fitness over winter. Also I'm a math geek. Numbers always matter

Myosmith

Heart rate is only indirectly related to calorie burn. It can give a decent approximation when the calculations account for an adequate number of variables, but it is far from a precise measurement. There are many factors that affect heart rate that have nothing to do with calories burned. Are HR based calculations useful, sure, but recognize them for what they are, an educated guess.

Mithrandir

Interesting article. I wish he would have explored more about why so many calorie counters grossly overestimate the numbers. I figure there's two reasons, probably both contributing to it (but of course I have no actual evidence backing up my hypothesis):

1) calorie burn algorithms are based on calculations done with professional athletes, where they try to come up with a formula that allows the user to input some key variables (gender, weight, hr) and arrive at a value that roughly matches what is seen in the lab. Given that most people exercising are not professional athletes, we have wildly varying figures and the formula's don't apply to the common man. Additionally, us clydes probably have the worst end of the stick, since we're extreme outliers in athletic data by our very definition, and thus are so far off the formula that it's just not realistic to use.
2) marketing schmucks figured out that people like devices that tell them what they want to hear, and thus pump up the burn numbers artificially in order to get people to like the devices they use more.


#2 is very cynical, I know. But I'm not going to lie, when I switched from my Polar HRM to a Garmin HRM last year, I was very disappointed that it said I was burning about 1/3rd less calories, and briefly considered switching back to the Polar. Of course the GPS feature of the Garmin sold me and I stuck with it, but that temptation to fudge the numbers existed just for a brief moment before I realised how ludicrous it all was. Seeing larger burn numbers made me immensely happy, and I imagine there's at least some people out there who would have made the change. Again though, I literally have no proof that either of those scenarios is even remotely accurate.

IthaDan 

Seriously. Eat less, ride more. Losing weight isn't a skill I'd like to master- Ideally, it's a once in a lifetime task. Counting calories isn't a solution, you're just white knuckling along waiting until you can reward yourself again.

vesteroid

I had a long post written up, and decided to shorten it to this. Training with power involves a lot of factors you may not have considered.

Its not just some number that shows up on your garmin...it can be an evil master that turns fun rides into tedious torture sessions...and from my experience is not group friendly.

chasm54

The way it assesses altitude has nothing to do with how it calculates calories burned. And it isnt very accurate at either.

Mith is right, if you want accuracy within even 5%, (and the variation is about your body's efficiency, the bike's drivetrain is typically very efficient) then a powermeter is the way to go. But speaking for myself, to spend the amount of money a powermeter costs just to get accurate calorie data would be ludicrous. If one wants a powermeter for training, fine: see the calorie data as a by-product. Relying on your Garmin is almost certainly introducing big errors, maybe 25%.

None of it matters, of course, unless you are using the data to inform decisions about how much you eat. In that case, big errors do matter and most people would be better off assuming they burn about 30kcal per mile. Slightly less if you're light, more if you're heavy, but in the ballpark and at least it avoids gross overstatements.

chasm54

These figures are all over the place. If I were you I'd forget about HR training until you are off the beta blockers. Whichever system of zones you are using, it is always the case that Z5 is above your FTP or lactate threshold. So by definition it is a level of effort you cannot sustain for more than a brief period of time. Were the zones set up correctly it would be impossible for you to spend most of a ride in Z5 unless that ride lasted only a few minutes.

As for getting off the meds, ask the doctor to give you a 24-hour monitor that you can wear. S/he'll then be able to see that your BP drops to normal levels when you're just going about your business.

Drew Eckhardt 

1000 Calories an hour is 277W which is enough for 24 MPH on flat ground assuming a 200 pound rider atop a 20 pound bike with .4 m^2 Sd and .760 Cd (as riding solo on road bike hoods) with .004 Crr .

krobinson103

Given a road bike is in this calculation I'd guess that 190 pounds of me on a 30 pound mtb with the shocks locked out and 1.95 dual purpose tires would require more energy to maintain the same pace. Thus using more calories per mile.