Bike Forums > Calories burned while biking
 Register All Albums Elite Membership Forum Rules Calendar Search Today's Posts Mark Forums Read

 Commuting Bicycle commuting is easier than you think, before you know it, you'll be hooked. Learn the tips, hints, equipment, safety requirements for safely riding your bike to work.

07-19-05, 07:32 PM   #26
Roody
Sophomoric Member

Join Date: Jan 2005
Location: Dancing in Lansing
Bikes:
Posts: 24,146
Mentioned: 6 Post(s)
Quoted: 613 Post(s)
Quote:
 Originally Posted by noisebeam Which is why it is a good tool, if it wanted to be more accurate, it would be too complex and require to much data input in regard to the engine (the rider, metabolisms, effciencies, wind print size), the tool (the bike, tires, etc.) and the environment (terrain, gradient profiles, weather, wind, temp, etc.) that the average (and often above average) user would not have readily available. Al
Now we're talking about something I actually understand--statistics. All of the factors Al mentions here are classified as error. They are random variances with a mean of zero, so they can be disregarded. For instance, on a long ride, you will go uphill, and then downhill, then uphill, then downhill. It all averages out in the end, so you don't even have to think about it. (In fact--if you start riding at home, and finish at home, you must ascend and descend equal distances. It is impossible to do otherwise.)

Other variables are similar. Sometimes your tires are soft, sometimes they're overinflated, but it all evens out so you don't have to measure it. Same with wind, temperature, your general physical condition, etc. The only factors you have to worry about are those which have a consistent effect on whatever you are trying to change or measure. These factors are called independent variables.

This means that Al's point is correct and the program mentioned by Yontman is unnecessarily complex, since it confuses error terms for independent variables. The simple Internet calculators mentioned by Al and simco are probably better tools.

07-19-05, 10:41 PM   #27
CdCf
Videre non videri

Join Date: Sep 2004
Location: Gothenburg, Sweden
Bikes: 1 road bike (simple, light), 1 TT bike (could be more aero, could be lighter), 1 all-weather commuter and winter bike, 1 Monark 828E ergometer indoor bike
Posts: 3,208
Mentioned: 0 Post(s)
Quoted: 0 Post(s)
Quote:
 Originally Posted by Roody Now we're talking about something I actually understand--statistics. All of the factors Al mentions here are classified as error. They are random variances with a mean of zero, so they can be disregarded. For instance, on a long ride, you will go uphill, and then downhill, then uphill, then downhill. It all averages out in the end, so you don't even have to think about it. (In fact--if you start riding at home, and finish at home, you must ascend and descend equal distances. It is impossible to do otherwise.)
True, you ascend and descend equal distances, but not for an equal amount of time.

Your fastest time will always be for a flat course given that you begin and end at the same location.
What you lose climbing can not be gained fully descending.
The higher the hill (or mountain!) you pass on the way, the longer you will take to complete, and the slower your average speed will be.

Think of it this way:
A flat course will take you the shortest time.
Now imagine the middle of the course rising to form a hill.
If you think about it, the following is true.
As the height of the rising hill approaches infinity, so does the time to climb to the top.
But you can't gain back a near-infinite amount of time lost by going down a near-infinite distance, even falling straight down!

07-19-05, 11:21 PM   #28
Roody
Sophomoric Member

Join Date: Jan 2005
Location: Dancing in Lansing
Bikes:
Posts: 24,146
Mentioned: 6 Post(s)
Quoted: 613 Post(s)
Quote:
 Originally Posted by CdCf True, you ascend and descend equal distances, but not for an equal amount of time. Your fastest time will always be for a flat course given that you begin and end at the same location. What you lose climbing can not be gained fully descending. The higher the hill (or mountain!) you pass on the way, the longer you will take to complete, and the slower your average speed will be. Think of it this way: A flat course will take you the shortest time. Now imagine the middle of the course rising to form a hill. If you think about it, the following is true. As the height of the rising hill approaches infinity, so does the time to climb to the top. But you can't gain back a near-infinite amount of time lost by going down a near-infinite distance, even falling straight down!
Yes but....In this application, the question is how many calories does one burn on average in one hour at a given speed, not calories expended in one particular hour of riding up and down one particular hill. As Al suggested, it is much too cumbersome (and probably impossible, according to other posters) to enter all the data on hills, wind, temperature, metabolism, etc. for each hour of riding. Furthermore, this figure would not be particularly useful for real world cyclists. Finally, it is mathematically pointless to compute this figure, since these random differences will sum to zero in a short time anyway.

CDCF--I love your posts. You're so smart, and you are almost always correct, and I am almost always wrong. But this time--you have not convinced me that I am wrong. I just hope I am explaining clearly.

07-20-05, 04:53 AM   #29
dfw
Stercus accidit

Join Date: Mar 2005
Location: Dallas/Fort Worth
Bikes: Trek Pilot 2.1
Posts: 686
Mentioned: 0 Post(s)
Quoted: 0 Post(s)
Quote:
 Originally Posted by joejack951 Speed and calories burned have very little to do with each other. If all you did was ride downhill (and not pedal) you wouldn't get much of a workout. But that calculator would say you are. I weigh around 160 and can burn 1000 calories per hour at my [almost] max effort (according to my Polar HRM). This is when I average over 160 bpm. When that BPM number drops, my calories burned also drops. At around 140 bpm, I'm burning 800 calories per hour. Speed can vary so much based on terrain/bike configuration that it's really meaningless if all you care about is burning calories.
HR and calories probably have less to do with one another than speed. Someone who is out of shape would have a much higher HR than someone who is in shape given the same wattage produced by both. Watts and calories burned have a very close relationship. Average speed would be closer because most people start and eventually end at the same point, so the final difference in elevation is 0.

 07-20-05, 06:08 AM #30 CdCf Videre non videri   Join Date: Sep 2004 Location: Gothenburg, Sweden Bikes: 1 road bike (simple, light), 1 TT bike (could be more aero, could be lighter), 1 all-weather commuter and winter bike, 1 Monark 828E ergometer indoor bike Posts: 3,208 Mentioned: 0 Post(s) Tagged: 0 Thread(s) Quoted: 0 Post(s) I'll give you an example. Take a rider+bike weighing 100 kg. Let's say this fictional rider is capable of putting out a maximum of 200 W. His total CdA is 0.4. His bike's Cfr is 0.006. Let's say he rides 100 km over three different courses. Course #1 is a completely flat road. No climbs or descents at all. His average speed is 31 km/h. The total energy spent is 2640 kcal (or around 820 kcal/hour). Course #2 is a course with 25 km flat road, 25 km with a 5% climb, 25 km with a 5% descent (one hell of a ride!) and finally 25 km of flat road. Here, the average speed drops to 24.7 km/h, but the energy spent goes up to 3320 kcal! Course #3 is a 10% climb for 50 km, followed by a 10% descent for the remaining 50 km (wohooo!!!). Average speed is now down to a snail-like 12.7 km/h, but the energy required is a massive 6490 kcal!!! He begins and ends at the same elevation in all three rides, yet the differences in energy and averages are staggering! Convinced?
 07-20-05, 06:30 AM #31 Cyclaholic CRIKEY!!!!!!!     Join Date: May 2005 Location: Sydney, Australia Bikes: several Posts: 4,269 Mentioned: 29 Post(s) Tagged: 0 Thread(s) Quoted: 375 Post(s) At this rate you guys had better start factoring in how many calories you burn while typing about riding
07-20-05, 06:49 AM   #32
dfw
Stercus accidit

Join Date: Mar 2005
Location: Dallas/Fort Worth
Bikes: Trek Pilot 2.1
Posts: 686
Mentioned: 0 Post(s)
Quoted: 0 Post(s)
Quote:
 Originally Posted by CdCf I'll give you an example. Take a rider+bike weighing 100 kg. Let's say this fictional rider is capable of putting out a maximum of 200 W. His total CdA is 0.4. His bike's Cfr is 0.006. Let's say he rides 100 km over three different courses. Course #1 is a completely flat road. No climbs or descents at all. His average speed is 31 km/h. The total energy spent is 2640 kcal (or around 820 kcal/hour). Course #2 is a course with 25 km flat road, 25 km with a 5% climb, 25 km with a 5% descent (one hell of a ride!) and finally 25 km of flat road. Here, the average speed drops to 24.7 km/h, but the energy spent goes up to 3320 kcal! Course #3 is a 10% climb for 50 km, followed by a 10% descent for the remaining 50 km (wohooo!!!). Average speed is now down to a snail-like 12.7 km/h, but the energy required is a massive 6490 kcal!!! He begins and ends at the same elevation in all three rides, yet the differences in energy and averages are staggering! Convinced?
You can't really compare a person riding like this with a commuter. For higher speeds, wind resistance becomes a signficant factor. As wind resistance increases, wattage will increase exponentially for small increases in speed. I would guess 95% of commuters are going to average in the 10-20 mph range with nothing even approaching a 10% climb or descent.

 07-20-05, 08:00 AM #33 CdCf Videre non videri   Join Date: Sep 2004 Location: Gothenburg, Sweden Bikes: 1 road bike (simple, light), 1 TT bike (could be more aero, could be lighter), 1 all-weather commuter and winter bike, 1 Monark 828E ergometer indoor bike Posts: 3,208 Mentioned: 0 Post(s) Tagged: 0 Thread(s) Quoted: 0 Post(s) Man oh man... Don't you think I've factored all that in? I've got a serious Excel spreadsheet going, you know...
 07-20-05, 07:40 PM #34 simco Member   Join Date: Jul 2005 Bikes: Posts: 25 Mentioned: 0 Post(s) Tagged: 0 Thread(s) Quoted: 0 Post(s) Whatev. I just don't want anyone looking at these things and thinking they're exactly right. I know some people do.
 07-22-05, 07:30 PM #35 dfw Stercus accidit     Join Date: Mar 2005 Location: Dallas/Fort Worth Bikes: Trek Pilot 2.1 Posts: 686 Mentioned: 0 Post(s) Tagged: 0 Thread(s) Quoted: 0 Post(s)
07-22-05, 08:42 PM   #36
Roody
Sophomoric Member

Join Date: Jan 2005
Location: Dancing in Lansing
Bikes:
Posts: 24,146
Mentioned: 6 Post(s)
Quoted: 613 Post(s)
Quote:
 Originally Posted by CdCf I'll give you an example. Take a rider+bike weighing 100 kg. Let's say this fictional rider is capable of putting out a maximum of 200 W. His total CdA is 0.4. His bike's Cfr is 0.006. Let's say he rides 100 km over three different courses. Course #1 is a completely flat road. No climbs or descents at all. His average speed is 31 km/h. The total energy spent is 2640 kcal (or around 820 kcal/hour). Course #2 is a course with 25 km flat road, 25 km with a 5% climb, 25 km with a 5% descent (one hell of a ride!) and finally 25 km of flat road. Here, the average speed drops to 24.7 km/h, but the energy spent goes up to 3320 kcal! Course #3 is a 10% climb for 50 km, followed by a 10% descent for the remaining 50 km (wohooo!!!). Average speed is now down to a snail-like 12.7 km/h, but the energy required is a massive 6490 kcal!!! He begins and ends at the same elevation in all three rides, yet the differences in energy and averages are staggering! Convinced?
No. It seems that the cyclist is riding slower in each successive course; therefore exercizing for a longer time. Does that not account for most of the difference in calories expended?

07-23-05, 03:23 AM   #37
CdCf
Videre non videri

Join Date: Sep 2004
Location: Gothenburg, Sweden
Bikes: 1 road bike (simple, light), 1 TT bike (could be more aero, could be lighter), 1 all-weather commuter and winter bike, 1 Monark 828E ergometer indoor bike
Posts: 3,208
Mentioned: 0 Post(s)
Quoted: 0 Post(s)
Yes, of course. I set the sustained power output to 200 W, so the total energy used will be a result of the total time and the sustained power.

But the issue here is that it doesn't even out!
With your reasoning, the rides should've used the same amount of energy and/or taken the same amount of time.

Here's the bit I'm referring to.

Quote:
 Now we're talking about something I actually understand--statistics. All of the factors Al mentions here are classified as error. They are random variances with a mean of zero, so they can be disregarded. For instance, on a long ride, you will go uphill, and then downhill, then uphill, then downhill. It all averages out in the end, so you don't even have to think about it. (In fact--if you start riding at home, and finish at home, you must ascend and descend equal distances. It is impossible to do otherwise.)
Contrary to what you stated, the average speed drops significantly, while the energy used goes up, for the same distance. That's why you can't exclude terrain from any reasonable energy calculator.
If you put the speed from the third example ride into one of those calorie calculators, you'll get a very different answer! In the case of the calculator previously linked to, it gives an estimate of 2900 kcal for ride #3. That's less than half of actual.

It overestimates the first example, giving 4700 kcal, while actual is 2640...

07-23-05, 03:08 PM   #38
dfw
Stercus accidit

Join Date: Mar 2005
Location: Dallas/Fort Worth
Bikes: Trek Pilot 2.1
Posts: 686
Mentioned: 0 Post(s)
Quoted: 0 Post(s)
Quote:
 Originally Posted by CdCf Contrary to what you stated, the average speed drops significantly, while the energy used goes up, for the same distance. That's why you can't exclude terrain from any reasonable energy calculator. If you put the speed from the third example ride into one of those calorie calculators, you'll get a very different answer! In the case of the calculator previously linked to, it gives an estimate of 2900 kcal for ride #3. That's less than half of actual. It overestimates the first example, giving 4700 kcal, while actual is 2640...
First example:
100 kg rider + bike = ~190lbs for rider
31 Km/h = ~19 Mph
100 Km @ 31 Km/h = 194 minutes

Plugging this into the web site yields 2989 which is pretty close to what you've calculated. Come up with whatever formula you want, and there's going to be situations where it won't work well. For any statistical analysis you should figure out what your subject is, and how much accuracy you need. As I said before, estimating calories as a function of speed is close enough for almost all commuters. How many commuters do you think are going to do a 10% grade climb for half of their commute?

If you want to start spitting hairs, you'd also need to consider variables like wind speed and direction, how many stop signs/lights you stopped at, metabolic rate, accelleration rates, temperature, humidity, road surface conditions, altitude relative to sea level, dew point, barometric pressure, visibility, moon phase, local cat population density, the dow industrial average, etc...

 07-23-05, 03:16 PM #39 CdCf Videre non videri   Join Date: Sep 2004 Location: Gothenburg, Sweden Bikes: 1 road bike (simple, light), 1 TT bike (could be more aero, could be lighter), 1 all-weather commuter and winter bike, 1 Monark 828E ergometer indoor bike Posts: 3,208 Mentioned: 0 Post(s) Tagged: 0 Thread(s) Quoted: 0 Post(s) Don't know what site you're talking about, but I get quite different numbers using this site: http://www.caloriesperhour.com/index_burn.html Yes, my examples were extreme, but I needed them to be to clearly make the point that climbs and descents don't average out in the end. They don't. Not by a long shot. As soon as you introduce climbs, however small, the average speed will drop but the total energy required for the distance will go up.
07-23-05, 04:31 PM   #40
dfw
Stercus accidit

Join Date: Mar 2005
Location: Dallas/Fort Worth
Bikes: Trek Pilot 2.1
Posts: 686
Mentioned: 0 Post(s)
Quoted: 0 Post(s)
Quote:
 Originally Posted by CdCf Don't know what site you're talking about, but I get quite different numbers using this site: http://www.caloriesperhour.com/index_burn.html Yes, my examples were extreme, but I needed them to be to clearly make the point that climbs and descents don't average out in the end. They don't. Not by a long shot. As soon as you introduce climbs, however small, the average speed will drop but the total energy required for the distance will go up.
The site I speak of is this one: Calories Burned Calculator

Yes, and when you go down the other side (or come back home), your average speed will increase and energy required will go down. The biggest reason it doesn't even out is because wind resistance eats up some of the advantage on decents. So for small hills, you'll still be in the ballpark. I would agree that if you are climbing a mountain, the results will be off.

07-23-05, 04:38 PM   #41
Roody
Sophomoric Member

Join Date: Jan 2005
Location: Dancing in Lansing
Bikes:
Posts: 24,146
Mentioned: 6 Post(s)
Quoted: 613 Post(s)
Quote:
 Originally Posted by CdCf Don't know what site you're talking about, but I get quite different numbers using this site: http://www.caloriesperhour.com/index_burn.html Yes, my examples were extreme, but I needed them to be to clearly make the point that climbs and descents don't average out in the end. They don't. Not by a long shot. As soon as you introduce climbs, however small, the average speed will drop but the total energy required for the distance will go up.
But still, cyclists (me at least) don't care to know how many calories we would burn riding up and down a given hill at a given speed. We want to know how many calories we burn in one typical hour at a given speed. The desired measurement is calories per hour, not calories per ride, not even calories per kilometer.

If I ride up a 10 % hill today, I will ride up a 5 % hill tomorrow, and on the flats the day after that. It all evens out over days, weeks or months of riding. Besides, how would I even compute the grade or length of every hill I climb up and down? No way Jose. Same with wind--yes it makes a difference, probably significant, in calories expended on a certain ride. But today the wind blows from the east, tomorrow it comes from the west, sometimes it blows all over the place. It all averages out. The only consistent variables, that can be measured with little error, are body weight and speed. You can forget about everything else and still get a "good enough" answer

07-23-05, 05:05 PM   #42
dfw
Stercus accidit

Join Date: Mar 2005
Location: Dallas/Fort Worth
Bikes: Trek Pilot 2.1
Posts: 686
Mentioned: 0 Post(s)
Quoted: 0 Post(s)
Quote:
 Originally Posted by Roody If I ride up a 10 % hill today, I will ride up a 5 % hill tomorrow, and on the flats the day after that. It all evens out over days, weeks or months of riding. Besides, how would I even compute the grade or length of every hill I climb up and down? No way Jose. Same with wind--yes it makes a difference, probably significant, in calories expended on a certain ride. But today the wind blows from the east, tomorrow it comes from the west, sometimes it blows all over the place. It all averages out. The only consistent variables, that can be measured with little error, are body weight and speed. You can forget about everything else and still get a "good enough" answer
If you were that concerned about calorie expenditures and had a need for that degree of accuracy, you would just get a power monitoring hub.

 07-23-05, 05:43 PM #43 CdCf Videre non videri   Join Date: Sep 2004 Location: Gothenburg, Sweden Bikes: 1 road bike (simple, light), 1 TT bike (could be more aero, could be lighter), 1 all-weather commuter and winter bike, 1 Monark 828E ergometer indoor bike Posts: 3,208 Mentioned: 0 Post(s) Tagged: 0 Thread(s) Quoted: 0 Post(s) You still don't get what I'm saying... This whole discussion was about the usefulness of online calorie calculators. I showed you that using speed alone as a measure of effort, as the online calculators do, is close to worthless. You could easily be off by 50% on a typical ride! I also question whether the variables are as variable as you make them out to be. Where I live, wind almost always blows from the southwestern quadrant, and wind speeds are almost always between 3-5 m/s. The hills are fixed, and the number of roads around here is low, so I regularly pass the same climbs and descents on most rides, regardless of where and how far I ride. Given that none of these variables are taken into account by the calculators, the error they end up with almost becomes an constant, and in this case, it's an unknown constant. For most people, the constant isn't even a known unknown, but an unknown unknown (sorry Rumsfeld... ). If I were to take the calculated result as a fact, that's accurate "because it all averages out in the end", I'd most likely experience a constant error offset in the energy expenditure estimates.
 07-23-05, 06:06 PM #44 Roody Sophomoric Member     Join Date: Jan 2005 Location: Dancing in Lansing Bikes: Posts: 24,146 Mentioned: 6 Post(s) Tagged: 0 Thread(s) Quoted: 613 Post(s) The problem with entering a lot of independent variables is that for each additional variable, you introduce a new source of error. If this error is consistent (reliable), it will throw off all of your calculations. For instance, if my measurement of wind speed is always 5 kph too high, my calorie calculations will always be too high. Soon I will be wondering why I have lost so little weight. On the other hand, if I do not measure wind speed at all, because I assume its effects are inconsistent or random, the mean my calorie calculations will be accurate, even if each single calculation is not. Counterintuitive, perhaps, but actually a simple measurement, with fewer (but easily defined) variables, is likely to be more accurate than a complex measurement with many (poorly defined) variables. Maybe the wind almost always comes from the southwest, but you do not really know it's direction or speed at every point of your ride. Since wind speed is, therefore, a poorly defined and poorly measured variable, best to just ignore it. Last edited by Roody; 07-23-05 at 06:14 PM.
07-23-05, 06:11 PM   #45
dfw
Stercus accidit

Join Date: Mar 2005
Location: Dallas/Fort Worth
Bikes: Trek Pilot 2.1
Posts: 686
Mentioned: 0 Post(s)
Quoted: 0 Post(s)
Quote:
 Originally Posted by CdCf You still don't get what I'm saying... This whole discussion was about the usefulness of online calorie calculators. I showed you that using speed alone as a measure of effort, as the online calculators do, is close to worthless. You could easily be off by 50% on a typical ride!
I plugged in your first example in the calculator link I provided and came up with almost exactly what you did. The first example you gave was the only example I saw that I would classify as a typical ride. Even on the 2nd example, which is definitely atypical for a commuter, the results weren't that far off.

It is worth saying that if someone regularly climbs a mountain on their way to work, they probably can't count on a calorie calculator being all that accurate.

 07-23-05, 06:22 PM #46 CdCf Videre non videri   Join Date: Sep 2004 Location: Gothenburg, Sweden Bikes: 1 road bike (simple, light), 1 TT bike (could be more aero, could be lighter), 1 all-weather commuter and winter bike, 1 Monark 828E ergometer indoor bike Posts: 3,208 Mentioned: 0 Post(s) Tagged: 0 Thread(s) Quoted: 0 Post(s) It's not poorly defined or measured. I live right at the coast, and six out of seven days every week, the wind comes from the SW quadrant... It's quite annoying. In my younger years, I had a telescope, and I could almost never use it, because the wind would blow, making my eyes water too much to see properly. We've had 2-3 days of virtually no wind in the past couple of months. That's all. But hey, if you don't agree with me, I won't force my opinion on you. I just feel that an online calculator assumes far too much, and ignores even more, to be of any practical use. Personally, I'm going to find a day when there's no wind, and ride full speed along the water near here. The road has virtually zero elevation. After timing myself, and knowing the distance, I get my average speed for maximum effort. I'll check my pulse at the same time. I'll then use that to estimate my power output using a few methods I've come up with. That power value can then be used later to pretty accurately estimate energy spent for other rides.
07-23-05, 06:46 PM   #47
CdCf
Videre non videri

Join Date: Sep 2004
Location: Gothenburg, Sweden
Bikes: 1 road bike (simple, light), 1 TT bike (could be more aero, could be lighter), 1 all-weather commuter and winter bike, 1 Monark 828E ergometer indoor bike
Posts: 3,208
Mentioned: 0 Post(s)
Quoted: 0 Post(s)
Quote:
 Originally Posted by CdCf If you put the speed from the third example ride into one of those calorie calculators, you'll get a very different answer! In the case of the calculator previously linked to, it gives an estimate of 2900 kcal for ride #3. That's less than half of actual. It overestimates the first example, giving 4700 kcal, while actual is 2640...
Quoting myself here in response to dfw's post.
Would you say those numbers are "almost exactly" or "not that far off"?

07-23-05, 09:43 PM   #48
dfw
Stercus accidit

Join Date: Mar 2005
Location: Dallas/Fort Worth
Bikes: Trek Pilot 2.1
Posts: 686
Mentioned: 0 Post(s)
Quoted: 0 Post(s)
Quote:
 Originally Posted by CdCf Quoting myself here in response to dfw's post. Would you say those numbers are "almost exactly" or "not that far off"?
What I would say is the 3rd example is nowhere near a typical commute.

You said, "You could easily be off by 50% on a typical ride!"

Are you trying to say a commuter will climb a 10% grade on a "typical" commute?

 07-24-05, 12:56 AM #49 jaubele1 Newbie   Join Date: Jul 2005 Bikes: Posts: 1 Mentioned: 0 Post(s) Tagged: 0 Thread(s) Quoted: 0 Post(s) My experience has been the same -- even when I ride with my kids (well below 13mph) my HRM shows 150+ calories burned per hour.
07-24-05, 03:09 AM   #50
CdCf
Videre non videri

Join Date: Sep 2004
Location: Gothenburg, Sweden
Bikes: 1 road bike (simple, light), 1 TT bike (could be more aero, could be lighter), 1 all-weather commuter and winter bike, 1 Monark 828E ergometer indoor bike
Posts: 3,208
Mentioned: 0 Post(s)
Quoted: 0 Post(s)
Quote:
 Originally Posted by dfw What I would say is the 3rd example is nowhere near a typical commute. You said, "You could easily be off by 50% on a typical ride!" Are you trying to say a commuter will climb a 10% grade on a "typical" commute?
You still don't get it.
My examples were extreme to clearly show how unrealiable average trip speed is for this purpose.
I could've used wind instead. In fact, here's a new example...

Same rider as in previous examples.

Ride #1:
10 km on a flat road with no wind, and then back the same way.
Average speed 31 km/h. Total time 39 min. Total energy 530 kcal.
(Online calculator suggests 1040 kcal for this ride...)

Ride #2:
10 km on flat road with a 5 m/s headwind followed by riding the same distance back, with a 5 m/s tailwind instead.
Average speed 22 km/h. Total time 54.5 min. Total energy 740 kcal.
(Online calculator suggests 730 kcal for this ride...)

So, average speeds go down while energy spent goes up.
But the online calculator tells you that the first ride used more energy than the second...

And 5 m/s of wind is not at all uncommon!