Whats the deal with Power in Strava

NITR8s

Can some please explain to me what it means, as I am lost on the concept.

For example I have come 5th in a local segment yesterday and it has recorded my power as 256w, whereas people who are slower on the segment have more power such as 7th place who is 3 secs slower than me has a power of 318w.

Cans someone please explain.

Mike39496

The power is estimated on the weight of your bike and yourself. If someone weighs more they need more power to get them up the hill hence why slower people can have more power if they weigh more.

I wouldn't worry too much about the power estimates anyway, if it's really troubling you you'd be better getting a power meter (but they are in excess of £1000).

NITR8s

Its not really troubling me, just wanted to understand it a bit better. So basically the heaver person put in more effort than me, but was still slower.

This means I have room to improve, first place here I come, KOM baby. Althogh I averaged 14.5 on the segment and first place averaged 18.6(wow), better get then

slowondefy2

The power estimates obviously don't take any wind advantage/disadvantage into account, which can massively skew the power estimates. Large pinch of salt required!

springtide9

As above really.

You'll notice that professional (mainly 'Tour riders') are often rated by their watt/kg (wattage per kg). This is their power output divided by their weight, which is a useful metric when stating a riders performance.

On hilly routes, it's gravity that is the main limiting factor to speed (so weight plays a bit part). But on flat routes wind resistance main limiting factor.

This is obviously simplified as there are many other factors involved, but this is why races like the Tour De France are do interesting (as the 'tour days' are a complete mix of events)

Strava estimates the power based on the incline/decline of the route and the riders weight. And powermeters start at just less than £600 new, as I bought one yesterday :-) [although this was with an extra 10% off voucher with Wiggle which has now expired. So actually £675]

okgo

NITR8s wrote:

Its not really troubling me, just wanted to understand it a bit better. So basically the heaver person put in more effort than me, but was still slower.

This means I have room to improve, first place here I come, KOM baby. Althogh I averaged 14.5 on the segment and first place averaged 18.6(wow), better get then

They mean nothing from what I can see. I really wouldn't pay ANY attention to them, none at all.

They're literally miles away from the correct figures, so not even worth looking at.

Lightning

Don't bother with it at all. At first I thought it was because of wind and all of that, but now I don't even know how it's supposed to work since it's incredibly wrong.

Just for you to have an idea of how wrong it can be, I have a few KOMs where I went all out and my average power is 50 watts tops. On the same segments, on different days, when I didn't go all out at all, my power is way over that. It's really random.

Oh, and then there's also the descents where I'm barely pedaling and still score amazing power.

springtide9

Lightning wrote:

Don't bother with it at all. At first I thought it was because of wind and all of that, but now I don't even know how it's supposed to work since it's incredibly wrong.

Just for you to have an idea of how wrong it can be, I have a few KOMs where I went all out and my average power is 50 watts tops. On the same segments, on different days, when I didn't go all out at all, my power is way over that. It's really random.

Oh, and then there's also the descents where I'm barely pedaling and still score amazing power.

I must say that the power displayed in Strava is generally much more consistent than that for me. i.e. if I look at my results over time from riding segments, I can see that it's related to the segment itself (I guess the "profile" of the route) and my weight. I'm not saying it's right at all, but that it is generally more consistent, and some segments that have been ridden by people with powermeters are a lot closer than I have expected.

I was talking to a friend about GPS accuracy (who happens to be very knowledgeable on the subject), particularly about height accuracy (as well as barometer) that are fitted to GPS units.
He was saying that it's not uncommon to see 50m height swings with the raw data from a GPS, and that the GPS units will try and smooth out the data as best as they can. Also any 'height' data from satellites is also very low resolution.

I guess what I'm saying is that as Strava is using height data to calculate power (and is not aware of other factors such as road surface, the weather etc), it's not really surprising it is very much of a 'finger in the air' value.

I would assume it's probably more accurate on flat sections and/or areas where there is greater height accuracy for the segment or route.

My Powertap arrived today; it will be interesting to see segments where Strava has estimated my power compared to actual power.

nickel

springtide9; could you post in this thread how your powertap compares to strava when you've had a go with it? Would be interesting to see how they compare. For what its worth the strava power thing is fairly consistent for me.

napoleond

I use Strava and also a Powertap. When I compared the figures before the Powertap and after, the power figure estimates on Strava were wildly inconsistently over or under estimating. Miles out. If any were close (within 20-30w then it must have been luck...

springtide9

Finally got out on the bike today with the Powertap, and did my 'restricted time' loop :-)

Here is how the estimated data on Strava compared to my Powertap Pro (the bold one is the Powertap)

Date Speed HR Power VAM Time
Apr 08, 2012 19.0mi/h 145bpm 221W - 1:51:00
Jan 29, 2012 17.8mi/h 146bpm 191W - 1:58:40
May 20, 2012 17.2mi/h 136bpm 231W - 2:02:11
Jan 08, 2012 16.3mi/h 145bpm 179W - 2:08:57

Date Speed HR Power VAM Time
Jan 29, 2012 17.7mi/h 161bpm 294W - 3:28
May 20, 2012 17.0mi/h 147bpm 319W - 3:36
Mar 05, 2012 17.0mi/h 152bpm 271W - 3:37

Date Speed HR Power VAM Time
Apr 08, 2012 11.7mi/h 152bpm 283W - 6:31
May 20, 2012 11.1mi/h 150bpm 301W - 6:53
Jan 29, 2012 10.7mi/h 153bpm 256W - 7:07

Date Speed HR Power VAM Time
Apr 15, 2012 14.9mi/h 152bpm 259W - 3:01
May 20, 2012 14.7mi/h 143bpm 294W - 3:04
Jan 29, 2012 14.5mi/h 147bpm 250W - 3:06
Mar 05, 2012 14.3mi/h 139bpm 238W - 3:08

Date Speed HR Power VAM Time
Mar 05, 2012 21.1mi/h 158bpm 650W - 0:43
Mar 02, 2012 18.5mi/h - 519W - 0:49
Jun 16, 2011 16.2mi/h 155bpm 392W - 0:56
Apr 08, 2012 15.9mi/h 153bpm 452W - 0:57
May 20, 2012 15.9mi/h 153bpm 442W - 0:57
Feb 13, 2012 15.1mi/h 148bpm 418W - 1:00
Apr 20, 2011 13.7mi/h 152bpm 329W - 1:06

Hope this is useful.

peejay78

strava has some savage inconsistencies in a number of areas. most notably the use of iphone.

springtide9

peejay78 wrote:

strava has some savage inconsistencies in a number of areas. most notably the use of iphone.

I think the main issue with the iPhone (and any most phone GPS's) - is that the GPS data is poor compared to a dedicated GPS. Not sure you really can't blame Strava because your GPS data is inaccurate.
(As the saying does, "Crap in, crap out")

There are other inconsistencies, but I'm convinced that it's related to the data smoothing (as in there is less in Strava than other apps - at a guess).

I've uploaded some old data from a FR 301 into Strava, which was very inaccurate compared to my 310XT, and the data was all over shop.

peejay78

i don't use an iphone. but some people do. some of the times are strange.

springtide9

peejay78 wrote:

i don't use an iphone. but some people do. some of the times are strange.

Today on more than one occasion my 800 started beeping at me stating that I was off-coarse (following a course put together with RWGPS), only to get 'Course Found' a few seconds later. I also have seen this on Sat Navs on my previous and current car, where I get the 'Off Road' message (when I an obvious on the road!)

Unfortunately I think that is the 'nature of the beast' with using a GPS for applications like Strava. GPS data is not always accurate, which makes trying to process this into something useful pretty tricky.
At best Garmin's only record position once per second, so the granularity of the data also isn't that great either.

Here you go... loops around a singletrack for 8 hrs. The singletrack was never more than about 1 metres wide:


The data was uploaded into Garmin Connect (so nothing to do with Strava) and was recorded with a Garmin 310XT (which is pretty accurate compared to other phone GPS's).

And now take a look at this one... same area - although I've had to zoom out to 2x scale... but this time recorded with a Garmin 301



And to compare, my FR301 used to give more accurate GPS data than my mobile phone.