Power Output and Strava data

[oldgreyandslow]

Has anyone used Strava to "calculate" their power output?

Apparently if you input your weight and that of your bike it calculates the power you use during any given segment and the power for your oveall ride.

I'm sure its not particularly accurate but it does at least give an indication of some sorts and I'm pretty surprised to see how low my overall output is. Even "pressing on" on a particular hilly bit of my normal ride I'm struggling to push out more than a Strava calculated 270W.

So how do I get my power up?

 

[Rob3rt]

I have heard some people say the estimates are quite good, based on my experience (power meter user, so I know the true values vs what Strava guesses) they often nothing but junk. I wouldn't waste 5 minutes on it!

Example on a recent 10 mile TT, my return leg was 28.1 mph, averaged approx 350W (measured), another rider did a 28.6 mph return leg, 346W (measured), the variance is not unsusual, he may be more aero than me. The Stava estimates for people in and around both of us are around 450W (to contextualise this figure vs my power output, I would struggle to hold this value for a couple of minutes). 100W difference! Now that is massive!

It also estimates poorly for someone that I know (We have compared data previously and I race against him every weekend!) according to Strava he averaged 404W for a 27.7 mph return leg on the same day. Now here is where things get a little interesting, I know for a fact this person goes faster than me, for less power! So if his return leg was slower than me on this day, then his power must have been quite a bit below mine, but Strava estimate he put out about 53W more, yet went slower. For the record, he beat me over the full distance despite my faster return, he put quite a bit of time into me on the outward leg.

 

[Hip Priest]

I take everything on Strava with a pinch of salt.

 

[Rob3rt]

I take everything on Strava with a pinch of salt.

Probably best

 

[T.M.H.N.E.T]

A pinch isn't enough

 

[Rob3rt]

Has anyone used Strava to "calculate" their power output?

Apparently if you input your weight and that of your bike it calculates the power you use during any given segment and the power for your oveall ride.

I'm sure its not particularly accurate but it does at least give an indication of some sorts and I'm pretty surprised to see how low my overall output is. Even "pressing on" on a particular hilly bit of my normal ride I'm struggling to push out more than a Strava calculated 270W.

So how do I get my power up?

Go out and hit a few segments with a tail wind, you will gain about 100 virtual watts!

Seriously though, the estimates will vary wildly with segment type, wind conditions, rider CdA values etc etc, to the point that even if some were reasonable approximations, the uncertainty surrounding which were good and which weren't would be so large as to render the data useless. you couldn't take anything as stated with confidence.

BTW, you say you struggled to push out 270W (guestimate), this is an average, your instantaneous power could be much higher. That said, a lot of people would suffer trying to hold 270W for 20-30 minutes (don't forget to compare power from person to person it is good practice to normalise against mass or CdA).

People seem to assume that since the pro's are hitting 1600-1800W in a bunch sprint that they should be able to ride around at 3-400W easy, it is only 1/4 of a pro's power right? Not so!

 

[VamP]

Agree, Strava power is nothing but a random stab at guessing power. The algorithm is poor for road situations, and completely useless off-road.

 

[Sittingduck]

270 Watts over a sustained period sounds ok to me! Pity is, we have just established the estimates are codswallop

 

[DWiggy]

I know two guys that started a short segment together but the times were different when uploaded to Strava (Garmin 500) the guy that physically came first by about 6 seconds was down on Strava as losing the segment by 6 seconds!
I'm guessing that its down to the accuracy of publicly available GPS and would be interested to know how the garmin 510 & 810 compares as apparently they use the Russian GPS system?

 

[Rob3rt]

I know two guys that started a short segment together but the times were different when uploaded to Strava (Garmin 500) the guy that physically came first by about 6 seconds was down on Strava as losing the segment by 6 seconds!
I'm guessing that its down to the accuracy of publicly available GPS and would be interested to know how the garmin 510 & 810 compares as apparently they use the Russian GPS system?

Only the 510 uses GLONASS.

The error in this example could be down to Strava's algorithms (they are hardly the best) possibly.

I used a Garmin Edge 500 for about 3 years and the accuracy was never a concern. However I did also use a GSC10 sensor to get more accurate speed and distance data for when it matters, but Strava doesn't use this data, it only uses the GPS data.