smurray

I use Strava on my iPhone to log my daily bike commute to and from work. I then use tapiriik.com to sync my Strava account to all the various sites I use (Runkeeper, Endomondo, Garmin Connect, etc...). I noticed today that there is a huge difference in estimated calories burned between Strava and the other sites. For example, this morning I rode for just under 50 minutes and went 11.2 miles. Strava calculates that I burned 281 calories during the ride, Runkeeper estimates 526 calories, and Endomondo estimates 555 calories. All three sites have the same info in terms of my height/weight/age and they are all reporting about the same distance/time/average speed. I know that all of these sites can't give anything more than a rough estimate without HR data, but why would Strava be estimating so much lower than the other two, and which is most likely to be correct?

kingsqueak

With HR data it's a wild estimate...without it's more of a guess.

Power + HR is closer, but without metabolic data for the individual still just an estimate.

I look at calories burned as a coarse estimate of effort/output but don't consider it further.

Looigi

Of course it depends on many factors but IMO, Cal burn for 50 min and 11.2 miles on flat/undulating terrain is more in line with the low Strava estimate, most definitely not near 500 Cal.

pdlamb

Best calorie calculator is your body, but it lacks an LCD readout. Everything else is just a guess.

That said, I second Looigi's opinion. The Strava estimate seems reasonable, the others seem optimistic unless you're 500 pounds riding a rust bucket with massive MTB tires.

smurray

Only other factor I forgot to mention was I ride a fixed gear bike so there is no coasting. Gear ratio is 48/19 w/ 25c tires. Neither Strava or any of the others have any of that data though so I'm still not sure why there's such a huge discrepancy.

smurray

Here's the ride in question.

zacster

Having just joined Strava 2 weeks ago, and having just bought an HRM/GPS watch last week, I'm somewhat mystified myself by the Strava and Garmin numbers I get. A quick 20 mile ride in the park gave me 950 calories, a strenuous 57 mile ride with climbing gave me ~2000. Something isn't right about that.

cccorlew

How can having a power meter be anything but near perfect? I mean, it's actually measuring the power you use. That maps to energy pretty darn closely doesn't it?

My PowerTap reports pretty close to what Strava guesses on most rides. Strava can't tell about wind, or my pannier laden commuter bike, but it's great for most road rides.
All others -- Garmin, Ascent, MyFitnessPal -- seem wildly optimistic about how many calories get burned.

kingsqueak

Power gets your power out more correctly, still calories burned while producing that power will vary person to person. Some will burn less, some more.

NeilMyers

I have always wondered how this all works. Would I be correct to assume the power measured on a power meter (assuming it is working correctly) is the minimum amount of power your body would need to expend to produce that power? Or are the two measuring completely different things? Trying to understand how this works ...

njkayaker

Yes.

Power meters measure power output. Not the power input into the system.

You would actually have to produce more power than is output because of losses between you and the wheels.

NeilMyers

Understood, but unless we've discovered a perpetual motion machine, doesn't the power input have to be the same or more than the power output (the difference being lost to friction, and other inefficiencies)?

In other words, one would expect the calories burned to be related to the power measured by your power meter.

As I said, this is interesting to me so I did a quick Google search. Here is an excellent article on the subject. In this article the author explains that the body is actually horribly inefficient at converting food to energy ... about 76% of the energy is lost to things like heat. He explains that if you convert the power meter readings to calories, and then account for these inefficiencies, you get a very good approximation of calories burned to produce that power. The example he gives is a rider who maintains 300 watts for an hour has burned 1,000 calories to generate that power.

I love forums ... you learn something every day!

njkayaker

Yes (that's what I was implying). I updated my post (before you replied) to make it clear.

Calories burned > power output.

Or:

Calories burned = power output + power loss.

Yes. It's related. Note that the efficiency (calories burned compard to power output) isn't going to be constant for a single individual (and, certainly, different people will likely have different efficiencies). Note that you are also burning calories for purposes other than driving the bike (eg, keeping warm).

The effiiciency isn't constant in one person (76% is some sort of average). And some people are more efficient than others.

Power losses:

metabolic inefficiencies.
technique inefficiencies.
mechanical losses (friction/flexing).

NeilMyers

You would think so. However, the article says the efficiency rate is remarkably constant and always close to 24%, regardless of the individual. I found that very interesting.

True, but what most people are interested in is how many extra calories did I burn on this or that ride. That is what the power meter tells you (indirectly).

I don't believe those come into play in this discussion. Yes, those may affect how much of your power is converted into speed, but whether the power was effective or ineffective from a cycling perspective, it is still power you generated and that came from calories you burned.

Looigi

""Converting watts to Calories

A 1 watt work rate
= 1 joule / second
= 3.6kJ/hour

There are 4.184 joules in 1 calorie
Therefore:

a 1 watt work rate
= 0.239 calories / sec
= 860.4 calories / hr
= 0.8604 kilocalories / hr
(kilocalories are usually expressed as capital 'C' Calories when talking about human nutrition and energy expenditure)

therefore, an energy expenditure rate of 1 Calorie/hr
= 1/0.8604 watts
= a work rate of 1.16 watts

The human body is around 20-25% efficient at converting food energy into mechanical energy. Digestion and absorption are energy intensive processes. The conversion of absorbed fats, carbs, and amino acids into the muscle's final energy molecule, adenosine triphosphate (ATP), is also energy intensive.

Based on this, the energy expended to create n watts at the pedal is thus::

100 watts = 0.864 * 100 / .24 = 358.5 Cals / hr = 1500 kJ /hr
150 watts = 537.8 Cals / hr
200 watts = 717 Cals / hr""

From: https://brisbanebike.blogspot.com/201...d-cycling.html

lsberrios1 

Strava uses your power meter data once it is set up to indicate calories (maybe only in the paid version).

For me an hour stroll at about 13-14mph on a 30'/mile terrain is about 300 cal (this is zone 1 recovery pace of less that 140watts). The same ride at 19+mph solo (z4 230+watts). is about 750 cal. Intensity definitely varies but calories can only be accurately measured with a Power meter. I had reached a plateau in weight loss until I got the power meter. Now I know exactly what I shed everytime I work out and pounds just come off since I can measure it accurately against my food intake.

Also, I've found out that comparing HR data and power data for calories burner HR data usually comes at 20% high. So if you go by "heart" a good guess would be to multiply the estimate given by 80% +/-. Of course that may vary depending on your cardiovascular condition and might be only specific to me.

When it comes to weightloss it is always better to underestimate calories burned and overestimate calories consumed if you are not 100% certain.

lsberrios1 

Yup. A better summary of what I meant to say. so take it that for a regular person at a leisurely pace it is 350 calories per hour. At an intense pace it is around 650 to 700 per hour.

njkayaker

It does make that claim (though we don't really know what "remarkably constance and always close" means. We don't know if it's correct. Note that the article is talking about efficiency with respect to calories burned. That's not that interesting compared to actually moving the bike.

That is, supposing the variation is +-2%, that different is negligiable with respect to calories burned.

But it is much more significant as a component of power production (2% of the total calories burned is 8% of the power being used to drive the wheels).

According to this, shivering consumes 425 watts and cycling at 13–18 km/h consumes 400 watts (larger than what the power meter would show). If those numbers are correct, then the variation of calories consumed and power produced can be very different than 24% (it's going to be much smaller if you are cold).

https://cnx.org/content/m42153/latest...ol11406/latest

No, the power meter tells you how many calories/power you used to move the wheel. It doesn't include the calories/power you produced metabolically for other purposes (including to overcome friction/flexing).

Metabolism -> power to pedals -> power to wheel (measured typically) -> power to road -> speed.

The technique inefficiencies and the mechanical losses occur before the power to the wheel is measured (those losses of power don't get to the wheel). Just like the metabolic inefficiencies.

NeilMyers

Looigi ...

You lost me on one part. At the top you say 1 watt work rate = 3.6kJ/hour. Later you say 100 watts = 1500 kJ/hour. If the top statement is true, then why wouldn't 100 W be 360kJ/hour? What am I missing? I am sure there is something obvious I am missing here ...

NeilMyers

Do you see this as being any different for the Garmin Vector, which measures power at the pedals?

njkayaker

I don't see it as very different. But the Vector power > wheel power. Keep in mind that some of the power applied to the pedals is used to move your body weight off of the seat (more or less).

NeilMyers

Interesting discussion. So, a better way to evaluate this is the power measured on your power meter gives you a conservative estimate of the calories you burned. If it says you burned 1,000 calories, then you burned at least 1,000, and perhaps 5-10% more.

For the purposes most people use that information for, I would say that is plenty accurate.

Now if I can figure out how many calories are burned when I walk 18 holes carrying my golf bag. That one is more complex, and the fellows on the golf forums are far less analytical!

njkayaker

Yes. Yes.

("Calories burned" is one of the less useful purposes of a power meter.)

Looigi

I didn't write it, just cut and paste from the site I cited. Anyway, the author is being loose with his units. In the last equation, it should be

100 "Watts mechanical" = 360 "kJ mechanical"/hr = 358.5 "Cal burned"/hr = "1500 "kJ burned"/hr (due to the .24 metabolic efficiency)