This time I look at the effects of WEIGHT and HEIGHT on the Garmin Running Power numbers. Does it take weight into account as it claims?
In an attempt to partly fill my own curiosity I have looked at some of the inputs to Garmin’s Running power algorithm over the last week or so. These links show what I anecdotally found.
So what else can I look at? Well, the table below is directly from Garmin.com and nicely summarises the inputs to their running power model.
| Component of Running Power | What is it? | Source of Data Used to Compute |
|---|---|---|
| Kinetic Power | Power required to change your pace | Speed from the watch* |
| Potential Power | Power required to run up or down a hill | Elevation data from the barometer on the watch |
| Vertical Oscillation Power | Power required for vertical oscillation on each step | Running dynamics from an HRM-Run, HRM-Tri, or Running Dynamics Pod |
| Horizontal Oscillation Power | Power required for horizontal oscillation on each step
(you brake a bit when you hit the ground, then accelerate again as you push off) |
– Speed from the watch*
– Running dynamics from an HRM-Run, HRM-Tri, or Running Dynamics Pod |
| Wind/Air Power | Power to overcome air resistance, greater if running into a
headwind and less if you are running with the wind at your back |
– Speed from the watch*
– Heading from the watch – Reported wind conditions from weather services – Barometric data to detect local conditions |
As a note to that table, Garmin point out the following, (complete with typos):
*The running power app uses the same speed that is displayed on the watch and logged in the activity file. If you have the footpod selected as your speed source, then that is what is used by the running power app.
Note: The equations to compute these components of running power also require some constants (sic) values such as your weight, acceleration due to gravity, and the density of air.
When you run, you will notice that running power responds quickly when you speed up or slow down. You’ll also find that running power is higher when you are running up hills than when you are running at the same pace on flat ground. Wind also effects (sic) running power.
So that leaves me two things to have a play with:
- Weight – let’s adjust the weight of ‘me’ on the watch and see how the Garmin Power numbers change. Christmas will come early on Today’s run as I miraculously consume several pre-run turkeys all by myself ie I will double my weight,
- Height – I’m going to go from mini-me to mighty-me with a doubling in my height. Garmin do NOT say above that height is a factor although I would have thought in reality it might have some CdA effect. Even if it did have an effect at that speed, I’m guessing it would be 0.1% or something of that kind of tiny magnitude.
- I’ll look at all 4 combinations of those 2 factors over about a mile; running at 5:00/km in a tiny bit of wind over a slightly down and up route around my block.
- I have STRYD and RunScribe as control devices (although I won’t show their data below)
Here is what I found:
In time order, from top to bottom they are: blue (2x height and 2x weight); red (normal weight and 2x height); green (2x weight and normal height); yellow (normal).
Observations
- Clearly weight is factored into the algorithm. I can’t say if that is done correctly.
- Doubling the weight seems to just about double the power. Someone can say below if that makes sense in the overall scheme of running with power.
- There is no indication that height is included as a factor. As Gamin didn’t mention it (earlier) I would assume they have not included it (as they say)
- The only thing that looks strange to me is that there appears to be a disproportionate increase in the variability of power with the higher weight setting. The yellow and red lines are much flatter – doubling the yellow/red lines would not produce variation of the same magnitude as the other two lines.
Disclaimer: I’m clearly not a scientist. I just found this interesting to do.
If you wnat to play with Garmin’s ‘freebie’ Running Power then you will need one of these…. 😉
From Cycling, Row- & SkiErging and Running (Stryd) I guess have a good feeling for watts and calorie expenditure. Can’t help myself, but Garmin’s power numbers don’t look right at all for me.
I still am very skeptic about power in running. Since it cant take wind into consideration, it is nothing more than grade adapted pace.
Since both stryd and garmin claim that it’s not that simple and they have developed ‘advanced’ algorithms, maybe You could do another fun test: try to run normally vs ‘ineffectively’ (low cadence, long stride, ‘heavy’ steps) in same pace? I wonder if then You will have different power levels? (You should if garmin’s and stryd’s claims are true.
hmm. i might try that. nice idea. might be hard to execute but i might give it a go.
there is one test on a treadmill at a constant speed where you vary your cadence and watch the power. the lowest power repsents the most efficient cadence/form (for that speed)
garmin DO claim to take into account wind BTW.
BUT: what is wrong with LIVE grade adapted pace (GAP)? i agree it’s (mostly) basically the same thing as power but live GAP/power IS useful. having a integer WATTS number is also a bit easier for subsequent post-workout maths for all the other clever stuff (as opposed to xx min/km)
“BUT: what is wrong with LIVE grade adapted pace (GAP)?” Nothing. It’s currently the
bestonly thing we have.how do you get LIVE grade adapted pace/normalised graded pace? ie on your watch. I can get it after the fact but that is les actionable
You can find some formulas in Suunto Apps, both for GAP and power (they actually use same formula).
Yep.
ah. my lack of an ambit for a while would explain that one 😉
ty
0,5ρcdAv3+20,5ρcdAvwv2+(0,5ρcdAvw2+crmg+i/100mg)v-Pη=0 is the formula for running power, which is explained in this book:
https://www.bol.com/nl/p/the-secret-of-running/9200000075008518
M is mass in this formula. To explain the other metrics goes a bit far for this post. 🙂
It introduces a model for running with power and contains a lot of references to research in physiological aspects of running. I quite liked the book!
I got lost at 0,5ρcdAv3 😉
seriously. that formula doesn’t say that 2xMass=2xPower tho does it? which was what my ONE EXAMPLE (not science!) came up with.
@Joop, can you comment on roughly what an increase in power should come from an increase in mass. (from a given speed/CdA etc. ceteris paribus; if my latin servies me correctly ??)
i/100mgv represents the power needed for climbing. i is the steepness gradient and has to be devided by 100 therefore. m is mass, g acceleration because of the gravity of the earth and v is the speed. If someone (weight 70kg) wants to go up Alpe d’Huez in France with 3,35 m/s speed (famous mountain from the Tour de France, 7.4% incline), the power needed for ascending is: (7,4/100)709,813,35 = 170 Watt. However, running itself on a level road also requires power (= cmv): 0,9870*3,35 = 230 Watt. In total required is 400 Watt. In the book they use a “marathon man” of 70kg who can produce 235 Watt during the period of time needed for the ascend. As 400 Watt is too much for him, he has to slow down. They calculate then that he can do 7,1 km/hour.
It’s really a nice book. It shows correlations for power versus heat, gradient, weight, all sorts of factors. You can mail them for the spreadsheet with their formulas as well (I received it as well). They are Stryd fans also, by the way.
The Running with power book of John Vance is nicely complementary to this book. This book shows the physics behind running with power
So, as far as I see it, mass and power correlate indeed
0.5ρcdAv3 + 20.5ρcdAvwv2 + (0.5ρcdAvw2 + crmg + i/100mg)v – Pη = 0
I can’t see how that formula can be arranged to give: mass multiplied by (something complicated) = power
therefore i cant see how doubling mass doubles the power.
Sorry for that 🙂
See https://stryd.blog/2017/05/26/beat-the-pace-game-run-with-power-for-a-breakthrough/ for their basic approach or just read the book….
Dear 5krunner,
Joop asked me to reply to your question. The power required for running is directly proportional with body mass, both on a flat course as well as uphill. However, the power required to surmount the air-resistance is independent of power. In general the air-resistance is in the orde of 5-10% of the total power, see http://www.thesecretofrunning.com
Best regards,
Hans van Dijk
so if the weight doubled you’d expect something like 190% increase in watts. which would not be out of line with what i found.