Originally Posted by merlinextraligh

One problem is that cycle computers grossly overestimate calories expended.

My ride today was 80 miles, 4'17" I did it on a bike that doesn't have a power meter.

The Garmin 1000 says I expended 6337 calories, which is patently absurd.

Based on similar rides with a power meter I'd ball park it at 2500-2600 calories less than 40% of the Garmin number.

Strava guesses it at 2748, which I think is pretty close, but likely a bit high.
https://www.strava.com/activities/447318455

Garmin's calories calculations, especially without HR data, are worse than worthless.

Originally Posted by caloso

I also noticed Strava calories to be high: 2832kcal v. 2542kj in my ride yesterday.

https://www.strava.com/activities/449125142

Originally Posted by caloso

I also noticed Strava calories to be high: 2832kcal v. 2542kj in my ride yesterday.

https://www.strava.com/activities/449125142

Originally Posted by gregf83

I believe Strava uses 1.1 X kJ which assumes your efficiency is lower than 25%.

Originally Posted by Heathpack

You know that a kilojoule is not equal to a kilocalorie?

One kilojoule equals 4.18 kilocalories. From kilojoules (measured directly from powermeter), all software will use a formula that estimates metabolic efficiency. Which happens to be something around 25% (range I think is 23-28% or something like that) for most people. That makes it seem like a kjoule is the same as a kcal, because most formulas take direct power output in kj, multiply by 4.18 to convert to kcal of work, but then because humans are only about 25% metabolically efficient, the kcal is then multiplied by ~0.25 (the guesstimated typical metabolic efficiency) to yield the final calorie burn.

Of all the formulae, for me personally Strava has it the most correct. (I know this because I count calories, track exercise calories from Strava and my weight pretty much makes sense.).

Any calorie burn estimation without power data is pretty much a wild guess (for your average person who never had power data). Now I can pretty decently guesstimate calories burned on a ride, having paid such close attention for a while now. But even so, Strava is way better than me.

Ymmv and the next guy too. Why? Because we all vary in our metabolic efficiency. Read Faster by Michael Hutchinson. Fabulous time trialist, gifted with an ungodly V02max. But had a relatively poor metabolic efficiency. The poor metabolic efficiency offset the extraordinary V02max. That's how the cookie crumbles.

Originally Posted by Titanflexr

From the Cyclops site: https://www.saris.com/CalorieCalculator.aspx

If I do 1000 Kjoules on a ride, can I calculate how many Kcals my body has burned?
One Kcal is equal to 4.186 Kjoules. Based on this conversion, if a rider does 1000 Kjoules on a ride the Kcal equivalent would be 239 Kcals (1000 Kjoules / 4.186). Remember, however, that the energy measured by the Power Tap in Kjoules only represents the mechanical energy delivered to the rear hub by the body, not the total energy expended by the body. If our bodies were 100% efficient then all of the metabolic energy produced by the body would be converted to mechanical work and a 1000 Kjoule ride would cost the body 239 Kcals. Unfortunately, we are not 100% efficient. On average a pedaling cyclist is only about 18 to 24% mechanically efficient. For every 100 Kcals our body burns, only about 18 to 24 Kcals actually goes into moving the bicycle. The other 76 to 82 Kcals gets wasted as heat or is used for the maintenance of other bodily functions. If we were to assume that a person was about 22% efficient, then to do a 1000 Kjoule ride, he would actually have to burn 1086 Kcals (239 / 22% or 0.22). Because 1 Kcal is about 4 Kjoules, and because only 1 Kcal out of every 4 Kcals burned by our body actually goes into moving the bicycle, 1 Kjoule of work performed on the bicycle is about equal to 1 Kcal burned by the body.
In the Applied Exercise Science Laboratory we actually measure mechanically efficiency and calculate personal conversion factors that allow athletes to convert mechanical Kjoules to metabolic Kcals. This conversion ranges from about 1.05 Kcals per Kjoule for the most efficient athletes to about 1.15 for our least efficient athletes. It’s important to note, that a number of factors like training status, temperature, and biomechanics may change a person’s efficiency. When it comes to calculating energy expenditure, however, the potential errors introduced by changes in an athlete’s efficiency are small compared to the potential errors that may occur when heart rate is used to calculate energy expenditure. While there is no perfect way to calculate the total Kcals used during a ride, measuring power output is currently our best estimate in real world conditions.
It’s important to note that food labels generally refer to energy as a Calorie and that a Calorie with a capitol C is the same as a Kcal. So a 250 Calorie Power Bar is actually a 250 Kcal Power Bar.
1 Kcal = 4.186 Kjoules

For every Kcal burned by the body only about 22% is used to move the bicycle.

1 Kjoule on Power Tap ~ 1.1 Kcals burned by the body Can I use my power meter to estimate how much food I should or should not be eating during a training ride or race?
Although our bodies can use carbohydrate, fat, and protein as a fuel source during exercise, once carbohydrate or glycogen stores are depleted an athlete’s ability to maintain a given pace becomes compromised. To maintain a high intensity for a prolonged period of time the ingestion of carbohydrate becomes critical. In order to determine how much carbohydrate should be ingested a rider needs to determine the following; 1) The Kcals burned, 2) The percentage of energy coming from carbohydrate, and 3) The total amount of carbohydrate stored in the body.
To estimate the number of Kcals being burned simply multiply the Kjoules of work done by 1.1 or just add 10 to every 100 Kjoules of work done. For example, if a person rides 2.5 hours at an average power output of 222 Watts, they would do about 2000 Kjoule of work and burn about 2200 Kcals.
If the pace is easy for a person you can assume that about 50% of the Kcals burned will come from carbohydrate. If the pace is somewhat hard about 80% of the Kcals might come from carbohydrate. And finally, for a hard to very hard pace carbohydrate would supply 100% of the Kcals. If we assume that the rider in the example above felt that 222 Watts was a somewhat hard pace, then about 1760 Kcals out of 2200 Kcals would come from carbohydrate (80% of 2200 = 1760). Because carbohydrate can be stored in the body as liver and muscle glycogen, the actual amount of carbohydrate that would need to be eaten would be equal to the difference between what was needed (1760 Kcal of carbohydrate) and what might be stored as glycogen.
The amount of glycogen available for energy during exercise is dependent upon a number of factors including an athlete’s training state, the type of activity, and most importantly a person’s diet. On a normal diet (50 to 60% of calories from carbohydrate), the amount of carbohydrate stored in the body is approximately 18 Kcals per kilogram of body weight. On a high carbohydrate diet (70 to 80% of calories from carbohydrate) the total is approximately 30 Kcals per kg. For a 70 kg (154 lbs) rider, the total energy from stored carbohydrate might range from 1260 Kcals (70 Kg x 18 Kcals per Kg) to 2100 Kcals (70 Kg x 30 Kcals per Kg), while the range for a 60 kg (132 lbs) rider would be 1080 to 1800 Kcals (Table 2).
Based on these values a 70 kg cyclist on a normal diet and riding at what he felt was a somewhat hard pace (222 W) for 2.5 hours (2000 Kjoule total work) would have to consume approximately 500 Kcals (1760 –1260 = 500) of carbohydrate to maintain his pace. This would be equal to about 4 bottles of Gatorade or 2 Power Bars. If the rider wanted to maintain body weight, they would eventually have to replace the additional 1700 Kcals (2200 Kcals – 500 Kcals = 1700) through their normal diet in addition to the Kcals needed to maintain their basic energy needs.
In contrast, if a recreational cyclist trying to lose weight does a 1hour ride at an easy pace of 100 watts he would only do about 360 Kjoules of work and only burn about 400 Kcals. Since this ride was at an easy pace only about 200 Kcals would come from carbohydrate. At 70 Kg and a normal diet, they would have about 1260 Kcals of stored carbohydrate. Theoretically, this person could do the whole ride without eating.
Estimating Carbohydrate Need
1) Figure out how many Kjoules of work you expect to do:
Kjoules = Average Wattage x Hours x 3600 / 1000

E.g., For a 200 Watt average for 4.0 hours the total Kjoules would equal:

200 watts x 3 hours x 3600  1000 = 2,880 Kjoules
2) Figure out how many Kcals you expect to burn:
High efficiency = Kcals = 1.05 x Kjoules
Normal efficiency = Kcals = 1.1 x Kjoules
Low efficiency = Kcals = 1.15 x Kjoules


E.g. For a person of low efficiency 2,880 Kjoules would be equal to:
1.15 x 2880 = 3312 Kcals
3) Figure out how much of the Kcals will come from carbohydrate:
Very Easy Ride = 30% of total Kcals from carbohydrate
Easy Ride = 50% of total Kcals from carbohydrate
Somewhat Hard = 80% of total Kcals from Carbohydrate
Hard to Max = 100% of total Kcals from Carbohydrate


E.g., For a somewhat hard ride, the total Kcals from carbohydrate would be equal to:
3312 Kcals x 80% = 2650 Kcals
Figure out how much is on board:
Normal Diet = 18 Kcals per Kg x Weight (Kg)

High Carbohydrate Diet = 30 Kcals per Kg x Weight (Kg)

1 Kg = 2.2 lbs

E.g., 70 kg male on a high carbohydrate diet 165 lb rider / 2.2 = 75 Kg 75 kg x 30 Kcal per kg = 2250 Kcals


Take the difference between the Kcals from carbohydrate and the carbohydrate on board:
E.g., 2650 Kcals – 2250 Kcals = 400 Kcals

400 Kcals = 2 Power Bars or 3 water bottles of Gatorade
Use the calculator at the upper left of this page to find your own calorie needs.

Originally Posted by Heathpack

If that's their formula, they are assuming 26.3% metabolic efficiency, aren't they?

Example:
1000 kj work performed (as measured by power meter) x 4.18 = 4180 kj energy
4180 kj expended x 0.263 metabolic efficiency = 1100 kcal burn

If Strava would take that 1000 kj and multiply by 1.1 to yield 1100 kcal burn, the assumption seems like it must be 26.3%. Unless I've done some bit of that math wrong. Which is possible because my brain has to think hard when it comes to math. I have low brain mathematical efficiency it seems.

Originally Posted by merlinextraligh

This post on slow twitch, which is quoting a now dead Saris link sets out the math:



So 1.1 kJ to 1 calorie splits the difference in the range of tested efficiency and is close as you're going to get short of having your own metabolic testing done.

Originally Posted by Heathpack

Any calorie burn estimation without power data is pretty much a wild guess (for your average person who never had power data).

Originally Posted by Seattle Forrest

Cyclists historically distrust heart monitors.

I do a lot of running lately. I have a few loops but it's almost always the same time, distance, and elevation gain (which means about the same pace ergo the same power output). If calories from HR is really just a wild guess it should give me 1,500 kCal for one mile, then 65 for the next, and then ... but it varies maybe 10 to 15 % between runs. I haven't looked very close but my hunch is the amount of uphill (always minimal) should explain the difference.

Experiment: I did two big climbs this fall, one on my road bike with power, the next day on a rented mountain bike with just my chest strap. Both rides were about 3 hours (2:54 road, 3:17 mtn), and very similar effort levels. Average and max HR were within a few beats of each other on both rides. Based on perceived effort, I'd guess power output was very similar. Garmin said 1,650 kCal for one ride and 1,682 for the other. The HRM did remarkably well.

Originally Posted by Heathpack

What I can or any cyclist can do depends very much on their trained state, their physiology, their relative ability to burn fat vs carb, their body size etc. This includes calories consumed on a ride and things like pace.

.

Originally Posted by Heathpack

You posting your original ride as "proof" that cyclists should universally eat less on the bike is similiar to you posting a specific route as proof that every cyclist should be able to ride it at the same pace as you. <Or that a 40 hr/month cycling goal is worthwhile.>

.

Originally Posted by merlinextraligh

But for the majority of the rides most of us do there's no need to eat at all.

Originally Posted by Heathpack

When I said HR results in a wild guess, I was referring to across a population- because people are very different in the intensity at which they are working at any given HR depending on their trained state and what kind of effort they are putting out and also because people vary greatly in size.

Originally Posted by Heathpack

A lot of the online calculators don't take body size, trained state or effort into account. When you look at two different rides that you do within days of each other, you could probably get a pretty good estimation of calories burned for each given ride from HR, provided you are well-rested, hydrated and temps are the same during both efforts. But if you try to compare you and I, or me today to me a year ago when I had poorer aerobic conditioning, or me today to me riding in triple digit temps, or me today to me in a fatigued state, then the HR to calorie correlation will start to break down.

Originally Posted by Heathpack

I've argued before that even though power is my key training metric, I still think HR is hugely valuable. You just have to really understand the factors that come in to play to affect HR and where it can lead you astray.

Originally Posted by merlinextraligh

My link was merely an example to illustrate a point.

Originally Posted by colnago62

It does take two to tango.