How accurate is the garmin gadgets calories counter?

chrisw12

Following on from Poppettes' thread how accurate is the forerunners and edge at working out calories used?

I managed a wooping 6135cal over 5.5 hours yesterday and I take it from reading the other thread that may be on the high side.

Does anyone know how they calculate the calories?

Assuming that there was some accuracy there (doudt it) and that energy and power are interlinked, could I use calories used as a measure of ride intensity?

forgotrafe

I can't answer your question, but am in the same predicament.

I entered my age, weight & bike weight accurately (well, without camelbak!) into my Edge 205 and went out for my first ride with it today.

My mate and I did 35 miles in 3 hrs (we're unfit ). My Edge said 3000 calories, my mates much more basic cycle computer (no stats, just based on time/speed) said 750.

I'd love to know which was the more accurate.

Rafe

Alex_Simmons/RST

Neither

The only way to know how many calories you expended on a ride is to know:

i. the total energy produced propelling yourself along (= average watts x number of seconds riding) expressed in joules (J), which of course requires a power meter.

ii the gross efficiency of your body during the ride (i.e. the ratio of energy expended propelling yourself forward vs the total amount of energy expended by your body, such as heat mostly, energy propelling yourself forward, noise etc)

Since it is almost impossible to measure gross efficiency on a ride (can be done in a lab though) and it varies, then even with a power meter it is still an approximation.

However, the gross efficiency, while it varies, doesn't vary that much, to the extent that the reading from a power meter can be used to estimate calorie consumption with reasonable accuracy (certainly with enough accuracy to manage dietary intake).

By an interesting quirk - the typical gross efficiency of a cyclist is roughly 23-25%, which neatly cancels out the 4.184 joules per calorie, or 4.184 kJ per Calorie.

Hence if your power meter tells you the ride energy output was 1,000 kJ then you can, with reasonable accuracy, know that your total energy expenditure was 1,000 Calories.

Alex_Simmons/RST

100 watts for 1 hour (3,600 seconds) = 360 kJ = 360 Calories expended in total

200W for 1 hour = 720 kJ = 720 Calories expended

300W for 1 hour = 1,080 kJ = 1,080 Calories expended


So 3,000 Calories in 3hrs = 1,000 Calories per hour ~= 1,000 kJ per hour propelling bike ~= 280 watts average.

I somehow doubt unfit cyclists could average 280W for 3 hours.

chrisw12

My figures are even more silly or I should be an olympic athlete or something else.

My 6135 divided by 5.5 gives 1115 cals expended wich would give me a power figure over 300w 8)

Which is funny because when I did have a power meter my highest 10 mile power was 290 and during a race 100 my av power was showing something like 240 (before it packed up.)

So we can safely conclude that either a) I've improved so much that I should be a pro or b) the garmin energy calculator over-estimates.

My money's on b

forgotrafe

Does sound like option B applies to us both. Shame.

Rafe

richa

I have Edge 205 and a Polar CS200 and tend to take the figures from the Polar (which also factors in my heart rate).

I think that the KCs and KC/hr is a nice stat - not in its own right - but by looking at intensity of a ride compared to other rides.

cervelorider

forgotrafe wrote:

I entered my age, weight & bike weight accurately (well, without camelbak!) into my Edge 205 and went out for my first ride with it today.


Rafe

It might be worth experimenting with the input values in order to come up with something more realistic.

Alex_Simmons/RST

RichA wrote:

I have Edge 205 and a Polar CS200 and tend to take the figures from the Polar (which also factors in my heart rate).

I think that the KCs and KC/hr is a nice stat - not in its own right - but by looking at intensity of a ride compared to other rides.

Using energy expenditure as a measure of ride intensity is only valid where the rides you are comparing are very similar in type.

Firstly, intensity is not a linear function of the rate of energy output (power).

You could not, for example, use kJ/Cal to measure the relative intensity of say a workout focussed on developing lactate tolerance (you know the all out 30sec intervals work) vs an endurance level ride.

Apart from absolute power levels, variability in power output is also a key factor when determining overall intensity. Variability is not measured by the energy expenditure.

Here is an example of what I mean when using a measure which does take into account ride intensity and variability (in this case the Training Stress Score):

http://alex-cycle.blogspot.com/2006/12/tss-vs-kj.html

chrisw12

So Alex if I read this right, there is a correlation between 'a power intensity measure' and calories burned as long we use only one rider performing the same ride type?

For example, for my ride on Saturday I 'scored' 6135 points over 5.5 hours giving me an intensity score of 1115 per hour.

Now I use this ride as my stock endurance ride and I'll ride it a few times a month under the same conditions (weather excluded) Do you think there's any point in me continuing to compare scores for this ride? Or is the correlation and accuracy so bad that it's not worth bothering?


Thanks for your help by the way 8)

Alex_Simmons/RST

I think there are too many variables in order to answer the question. I don't know how these devices can measure energy output without measuring power or doing a lab gas in/out analysis. They must make some huge assumptions.

IOW - I don't think it would be any more accurate than saying your PE was a 6.5/10 for 5.5hrs.

If you were taking a calories measure from a power meter then I'd say, that's probably OK (as long as you ride the ride in the same way each time). But even a day where you feel like punching it up the rises and recovering down the other side can produce quite a different intensity score than the same course where you keep a steadier more isopower like effort going. But then if you used a PM then you'd be able to measure that.

Ric_Stern/RST

Chris, sell your Garmin and get a Power Tap ;-). that way you'll know how you're improving (and if you get one from me, you know i'll take care of you should anything go wrong with it -- and to be honest i think i've had only something like one or two returns on PT over the last 18 months).

ric

ut_och_cykla

My impression has been that many HRM calorie use calculators use an algorythm based on running.The results I get out of mine fit reasonbly well with energy use calculators on the web. (all of which are based on moving a given weight over a distance in a given time).

HRMs I've used seem to consistently over estimate kcal usage on bikes presumably because the algorythms don't know your doing the work using gears, wheels etc!
The tacx flow I use seems to give W/kcal results that tie in with 'real life' performance when compared on sites like kreutzzotter.de etc. they are only indications but they're good enough for me! :-)

There are calculations on the net which indicate theoretical kcal usage per minute on bikes - assuming you are riding on the flat with no wind etc. but they are just that - theoretical.

hambones

The Garmin only calculates calories burned based on time riding/speed//distance etc. IT cannot take into account what type of riding you are doing because as has been said it does not measure power.

However the HR seems to have little bearing - for instance my last 2 rides were...

50km in 4hrs 15mins with average HR of 161, calories burned 1474

109km in 4hrs 32mins with average HR of 149, calories burned 3840

Trust me, the first ride was much much harder with it all being off-road through thick mud, as shown by the HR.

phil s

Ric_Stern/RST wrote:

Chris, sell your Garmin and get a Power Tap ;-). that way you'll know how you're improving (and if you get one from me, you know i'll take care of you should anything go wrong with it -- and to be honest i think i've had only something like one or two returns on PT over the last 18 months).

ric

Forget this blatant effort to sell you some very expensive equipment and just enjoy your Garmin. Let's face it, the calorie function is so unimportant as it has many other benefits.

Alex_Simmons/RST

phil s wrote:

Forget this blatant effort to sell you some very expensive equipment and just enjoy your Garmin. Let's face it, the calorie function is so unimportant as it has many other benefits.

How much are the Garmins?

phil s

About six times less than a PT

Ste_S

RST will be advising using a Power Tap to solve problems in the Gaza Strip next.

Bought direct from RST, natch

Alex_Simmons/RST

phil s wrote:

About six times less than a PT

So they pay you to use one?

I only ask because I actually don't know, never seen anyone here use them. None of my clients have/do either.

I just looked up one on the web after your helpful reply and it was about $450, so about 60-70% of the price of a PT Pro (although you can pick up good used PTs for closer to $500).

Maybe I'm looking at the wrong model. What so good about them?

I am more familiar with Polars.

phil s

Alex, they're about £130 now. No idea what that is in upisde down dollars
The great thing about them is that they give you all the functions of a Polar plus you can program routes off web-based maps into the unit and thereby not get totally lost or have to stop to read maps on long rides/unfamiliar rides. This has totally rekindled my enjoyment of getting out and training again. Also if you calibrate correctly each time you can use them on the turbo as it takes speed off the rear wheel. Very decent piece of kit...
(And for all you power perverts I hear the 750 model is SRM and PT compatible and with colour mapping. Would love to buy it with said set-up but I'd need to rob a bank/win the lottery/sell my body to do so)...

...oh, and the one drawback is the calorie thing is way-off. I'm sure it could be made a tad more accurate by playing with the settings but this function is totally irrelevant to me

popette

phil s wrote:

Alex, they're about £130 now. No idea what that is in upisde down dollars
The great thing about them is that they give you all the functions of a Polar plus you can program routes off web-based maps into the unit and thereby not get totally lost or have to stop to read maps on long rides/unfamiliar rides. This has totally rekindled my enjoyment of getting out and training again.

same here - it's great not having to stop and read maps every two minutes when going on a new route for the first time. I plan my routes in memory map, use this conversion program to load them onto the Garmin and away I go. I've been on so many new routes in the last 4 weeks, exploring places I've never been to before. Then you get home and download all the data onto the pc and you can plot various variables against each other.It's really cool.

Ric_Stern/RST

what's wrong with getting lost? I've had great enjoyment over the years getting completely lost on my bike and then 'discovering' new roads.

ric

Ric_Stern/RST

Besides, if i recall correctly, the original question was not how useful is a Garmin (or other system) at map reading, but calculating energy expenditure, which is what i was replying about.

Ric

Alex_Simmons/RST

phil s wrote:

Alex, they're about £130 now. No idea what that is in upisde down dollars
The great thing about them is that they give you all the functions of a Polar plus you can program routes off web-based maps into the unit and thereby not get totally lost or have to stop to read maps on long rides/unfamiliar rides. This has totally rekindled my enjoyment of getting out and training again. Also if you calibrate correctly each time you can use them on the turbo as it takes speed off the rear wheel. Very decent piece of kit...
(And for all you power perverts I hear the 750 model is SRM and PT compatible and with colour mapping. Would love to buy it with said set-up but I'd need to rob a bank/win the lottery/sell my body to do so)...

...oh, and the one drawback is the calorie thing is way-off. I'm sure it could be made a tad more accurate by playing with the settings but this function is totally irrelevant to me

That's about A$300.

I heard about the 750 on power forums. I'm not sure the 750 is either SRM or PT* compatible, so I'd say that's a misrepresentation. I understand the 750 is ANT+ compatible, which IIRC neither SRM or PT are. The yet to be released Quarq power meter is supposedly ANT+ compatible.

* Saris confirmed this directly to the wattage forum.

Anyway - back on topic, on bike calorie counters on anything other than a decent power meter are bunkum.

chrisw12

Ric_Stern/RST wrote:

what's wrong with getting lost? I've had great enjoyment over the years getting completely lost on my bike and then 'discovering' new roads.

ric

I've seen you get lost one day in the Neath valley. a sunny day about three years ago. One minute you were there, the next you were gone. Can you remember?

phil s

Going off piste slightly, Alex and Ric, do you see the price of power-measuring devices coming down any time soon?

Alex_Simmons/RST

In my opinion, no, not soon anyway. That's based on the limited market size, coupled with limited manufacturers of reliable, proven products. Where not talking iPods here.

New players will help to some extent but even the new (yet to be released) Quarq is supposed to be priced in between PT SL 2.4 and SRM. ergomo I doubt can afford to drop their price.

Polar have their PM but by the time you set up a new polar and power module, well you may as well have a PT. and the Polars simply don't work on home trainers with any consistency/reliability.

The best value (reliability, performance, price) are the PT Pro models.

My clients do or have used all of these meters.

IBike has struggled with credibility, although they have a dedicated fan base and seem to have kept working on their product.

Several other products have been mooted but none have really surfaced into the world of commercial reality.

Ric_Stern/RST

chrisw12 wrote:

Ric_Stern/RST wrote:

what's wrong with getting lost? I've had great enjoyment over the years getting completely lost on my bike and then 'discovering' new roads.

ric

I've seen you get lost one day in the Neath valley. a sunny day about three years ago. One minute you were there, the next you were gone. Can you remember?

Do you mean around the circuit where they hold the Presidents H'Cap road race? In which case i shredded my rear tyre on something (can't recall what) and had to go home by car.

ric

Jeff Jones

I did a very rough comparison today over one of my 45km routes.

The Garmin said I'd burned 1670 calories for the distance, which involved 500m of climbing and took me 1hr35 at an average of 69% of my max HR.

I dug up an Ergomo file from a ride I did last year on the same route, but on my TT bike (which is of course faster than my training bike). It said 206W average (256W NP) for 1hr25 = 206*85*60/1000 = 1050kJ, or roughly 1050 Calories burned after cancelling out conversion factors. For the record, my avg HR was roughly the same (70% of max) as this morning's ride.

Even taking into account the relative efficiency of the TT bike, I can't have burned more than 1200 Cal on this morning's ride on my training bike. That's 500 Cal different, or most of a 100g block of chocolate!

So I'm not going to use the Garmin to plan my diet :-)

SteveR_100Milers

Well I experimented with my 205 this morning on my car commute from ystrad to ross on wye distane oof 42 miles in 54 mninutes, I managed a top speed of 147Km/h yet only burnt 1669 calories which is even more amusing. I also noticed that the you cannot own a bike less than 10Kg in weight according to garmin. #

I have always estimated calories on a rough estiamte of 500/hr at sun day club pace, 600/hr at tempo and 750/hr at TT pace. Bet thats closer than garmins attempt!

macka86

Hey all i am a traithlete and my Garmin Edge 305 today sad i burned nearly 5000 calories for my ride of 4 1/2 hrs (140km) I have had a vo2 test and my max watts was 450 and v02 was 62 so i think i would average about 300 watts on a good day. I think it would be as good as it can be for a computer that only has limited info so take it as a guide bu not an exact measurement