Dusty Bin
Guest
Why would Strava want to routinely and pre-emptively alter the raw data given to it by a third party's input device?
Top speed. Chain rings ?????
- Thread starter Doyleyburger
- Start date
Page may contain affiliate links. Please see terms for details.
Why would Strava want to routinely and pre-emptively alter the raw data given to it by a third party's input device?
Rob3rt
Man or Moose!
- Location
- Manchester
In order to extract the most accurate metrics they could from a "noisy" data set.
HLaB
Marie Attoinette Fan
Not too much, some of the best descenders I know are compact units it's more to do with aero, skills and courage. I also know that I lack the latter two but when I adopt an aero position on a non technical descent, I shoot by folk heavier than me.
Dusty Bin
Guest
Who says it is 'noisy' ? And who says this 'noise' is materially affecting the integrity of the raw data?
Rob3rt
Man or Moose!
- Location
- Manchester
The scientific community, that is mathematicians, engineers etc.
Dusty Bin
Guest
Noise implies a random distortion, which is not something you can economically address on what is effectively a free web app. If people are seriously worrying about losing a few nanoseconds off their top speed, then they obviously don't have enough to worry about...
Rob3rt
Man or Moose!
- Location
- Manchester
I can't help but feel you don't really know what you are talking about.
lukesdad
Guest
..and probably some of the best descenders you know are heavyweights too ?
Rob3rt
Man or Moose!
- Location
- Manchester
With regards weight, it does have an impact and assuming all else equal a heavier rider will have a higher terminal velocity. However a heavier rider will also generally be larger and thus have a higher (generally speaking, not always true as drag is not solely dependant on presented area) CdA, which will counteract the increase in mass.
It would be interesting to see how much effect it has at various levels of rider though because for competitive riders, heavier and lighter riders won't necessarily have that much difference in their presented frontal area as positions are dialled and physiques converge, whereas at lower levels, where you get all shapes and sizes and loads of weird and wonderful bike fits, the chances are a heavier rider is bigger, which is reflected in their ride position and a generally larger frontal area such that the CdA vs weight is not so clear cut.
Having said all of this, terminal velocity is not the end of the story either.
It would be interesting to see how much effect it has at various levels of rider though because for competitive riders, heavier and lighter riders won't necessarily have that much difference in their presented frontal area as positions are dialled and physiques converge, whereas at lower levels, where you get all shapes and sizes and loads of weird and wonderful bike fits, the chances are a heavier rider is bigger, which is reflected in their ride position and a generally larger frontal area such that the CdA vs weight is not so clear cut.
Having said all of this, terminal velocity is not the end of the story either.
Last edited:
fossyant
Ride It Like You Stole It!
- Location
- South Manchester
This.
You dont need a bigger gear. I have 53x13 which is like 50x12 and don't need any more. I can pedal till about 40 to 45 but its not as efficient as an aero tuck which you will go faster.
Being able to ride up hills faster is key. Down fast is technique, skill, positioning and being able to spin.
I also think you are probably grinding the gears too much (low cadence)
HLaB
Marie Attoinette Fan
Nope as said they are compact units, lighter than me.
Dusty Bin
Guest
Why - because you have no credible response?
I'm saying it doesn't matter. Your turn now - what are you talking about?
Rob3rt
Man or Moose!
- Location
- Manchester
Why - because you have no credible response?
I'm saying it doesn't matter. Your turn now - what are you talking about?
What do you want by way of credible response?
You have questioned the fact that GPS data is noisy, something that is accepted by the scientific community, as I have informed you. A quick Google search on your part should confirm that this is the case.
You have then assumed that the effect of this noise is minimal, which is not strictly true. The error introduced can be substantial!
Further you have made the statement that it is not economic to reduce the effects of such issues, which again is not true and some basic statistical methods could be implemented reduce the impact of such noise. I suggest you look up a few methods employed in GPS data interpretation. The Extended Kalman Filter might be a good place to start. To help you understand this, start with some basic Bayesian Estimation!
Last edited:
Dusty Bin
Guest
Ideally, something which doesn't involve you running off to google every time you need to answer a question. I work in the microwave industry (no, not the ovens, ffs!
), so if you want to get technical, you will need to step away from wikipedia.Multipath noise and other interference is inevitable in radio signals - no argument there.
Neither of us knows the precise noise levels involved here, so your comment is irrelevant.
You need to go back and read what I said. I said it was probably not economical for Strava (a free web app) to do this. Why should they - and how much more accuracy could they provide? For that matter, how much more accuracy is needed for people who want to casually race each other on the internet? If you need high precision GPS data, then you won't using Strava anyway.
ho hum...
Last edited:
Rob3rt
Man or Moose!
- Location
- Manchester
I did not run away to Wikipedia, this subject being discussed employs many of the techniques used in my own area of research.
Further, I didn't initially comment on the extent of the noise, I simply said the data is noisy, you questioned this fact, then you further stated, in a roundabout way, that the effect was minimal. I simply said this is not strictly the case and that the effect of the noise CAN be substantial. So my comment is no more irrelevant than your own.
Not economic for a free web app? 1st, Strava is not wholly a free app! As for why should they do it? Well it would help with segment matching for starters which would save them time manually matching segments where the GPS data has not registered a segment and the user emails in asking for it to be matched (they tell you to do this on their website). How much more accuracy would it lend? Well, that depends on many factors, not least the desired metric you are extracting from the data.
