strava - getting lame?
#76
Senior Member
Join Date: Nov 2010
Posts: 195
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
#77
Senior Member
Join Date: Feb 2012
Location: Nova Scotia, CANADA
Posts: 237
Bikes: CCM 21 speed big box special.....
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
I have a tablet, how do I start a segment? Go to the start and then start my ride! total noob here
#78
Beer >> Sanity
Join Date: Oct 2010
Location: Colorado
Posts: 3,449
Bikes: 2014 Evo DA2, 2010 Caad9-4, 2011 Synapse-4, 2013 CaadX-disc
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
#79
Token Canadian
Join Date: Nov 2008
Location: Gagetown, New Brunswick
Posts: 1,555
Bikes: Cervelo S1, Norco Faze 1 SL, Surly Big Dummy, Moose Fatbike
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 200 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
and there's no easy way to look at historical data comparatively.
Strava rocks.
DG
#80
Junior Member
Join Date: Oct 2012
Location: Sheffield, UK
Posts: 6
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
English have never been my strong point! Please email me (link on the website) with any corrections that I can make to improve things. I rely pretty heavily on spell checkers but unfortunately they don't exist in development tools.
@bikerjp I've been a bit confused about the "percentile" thing since I was asked in another forum to add that column as the only time I've come across it has been in terms of my children's head size and weight but I take it those are based on normal distributions where as your placings isn't. Is that value actually just a placing percentage? If so then theoretically a percentile could be calculated from everyone's times but that would be hideously time consuming and API killing to find out.
Even if percentile was the correct term then I'm not quite sure why 13/499 would be 96th percentile rather than 3rd. My new boy's head is huge and so he's in the 98th percentile which would be the equivalent of being in say 480/499? I avoided stats at all costs during my maths degree so I'm more than happy to be educated to how this should be used :-)
In terms of your incorrect placings then see the https://veloviewer.com/blog/known-issues/ page and read "Optimistic placings" (linked from the top of your segment list also). Due to a Strava API bug unfortunately but hopefully they will fix in the new year.
Ben
@bikerjp I've been a bit confused about the "percentile" thing since I was asked in another forum to add that column as the only time I've come across it has been in terms of my children's head size and weight but I take it those are based on normal distributions where as your placings isn't. Is that value actually just a placing percentage? If so then theoretically a percentile could be calculated from everyone's times but that would be hideously time consuming and API killing to find out.
Even if percentile was the correct term then I'm not quite sure why 13/499 would be 96th percentile rather than 3rd. My new boy's head is huge and so he's in the 98th percentile which would be the equivalent of being in say 480/499? I avoided stats at all costs during my maths degree so I'm more than happy to be educated to how this should be used :-)
In terms of your incorrect placings then see the https://veloviewer.com/blog/known-issues/ page and read "Optimistic placings" (linked from the top of your segment list also). Due to a Strava API bug unfortunately but hopefully they will fix in the new year.
Ben
#81
Beer >> Sanity
Join Date: Oct 2010
Location: Colorado
Posts: 3,449
Bikes: 2014 Evo DA2, 2010 Caad9-4, 2011 Synapse-4, 2013 CaadX-disc
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
If I'm 13/499 then I scored better than 486 others. So 486/499 = .974 or the 97th percentile (okay, I was off by one). The idea behind percentile is the higher the better (though not necessarily with baby head sizes because you wouldn't want your kid to look like a lollypop).
More info https://www.regentsprep.org/Regents/m.../quartiles.htm
Last edited by bikerjp; 10-31-12 at 08:49 AM.
#82
Junior Member
Join Date: Oct 2012
Location: Sheffield, UK
Posts: 6
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
I've tweaked the sums and now the percentile is showing as you say. I'm relying on my users to test it all for me :-)
Let me know if you see anything else that needs changing or if you have any other ideas I could implement.
Interesting thread this. One of the reasons I started putting VeloViewer together was to give a better way for comparing how well I had performed across all my own segments. The Strava placing doesn't really give you anything useful as it is not in the slightest bit comparable between segments due to all the reasons stated above. I also wanted to try and give something useful to those people who don't have heaps of KOM's as this is something Strava doesn't seem to do at all. Of course most people do like seeing their placings but hopefully they take them in context.
I completely agree with short segments (be they up, down or flat) being of little value, check out https://veloviewer.com/NewLeaderboard.php where I attempt to explain why.
#83
Junior Member
Join Date: Sep 2012
Posts: 21
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
I don't care if I am #2 or #3 on a segment. If I am 57 seconds behind the first place, I know I have a bit more work to do. If I'm one second behind, I know I'm reasonably well matched with that rider on that particular ride. Wind conditions being a real variable in South Florida.
The Strava API is quite powerful - and once V3 is released, I think we'll see some much better third party tools for analysis. I wrote two different tools to analyze my riding - one to find riders that have ridden local segments within the last 60 days, and the other to track progress on segments to see if I am improving. I have an idea for an app I've been toying with to run while I'm riding, but, alas, time hasn't been on my side.
But yes, the percentile on veloviewer is backwards, but, it is easy enough to figure out what was intended. It is what inspired me to write my segment performance tracker, though, minor hitches with the API prevented me from doing it precisely how I wanted.
#84
Senior Member
Percentile is usually used to show what percentage of observations fall below a certain level. If your performance is better than 96% of the other observations you are in the 96th percentile. There is some discrepancy over how to calculate percentile but it's not that important for this application. The basic math would look like this:
If I'm 13/499 then I scored better than 486 others. So 486/499 = .974 or the 97th percentile (okay, I was off by one). The idea behind percentile is the higher the better (though not necessarily with baby head sizes because you wouldn't want your kid to look like a lollypop).
More info https://www.regentsprep.org/Regents/m.../quartiles.htm
If I'm 13/499 then I scored better than 486 others. So 486/499 = .974 or the 97th percentile (okay, I was off by one). The idea behind percentile is the higher the better (though not necessarily with baby head sizes because you wouldn't want your kid to look like a lollypop).
More info https://www.regentsprep.org/Regents/m.../quartiles.htm
#85
Senior Member
Join Date: May 2012
Posts: 66
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
The most significant comment I have is that your app displays position without taking into account ties. For instance, if I am #14 on the Strava leaderboard, with the KOM at 2:01 and 12 others all tied for #2 at 2:02, with me coming in at 2:03, your app shows me at #3 overall (presumably, on the basis that I am just 2 spots back on an "absolute" basis), which from a view of the Segments tab, can lead me to believe that I am a lot closer to the KOM than I really am. This also makes the "Position %" inaccurate.
#86
Bourbon junkie
Join Date: Aug 2007
Location: NorCal
Posts: 722
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 5 Post(s)
Liked 0 Times
in
0 Posts
gaming is not cheating. If they're on their bike and not getting towed. Legit as far as I'm concerned.
Last edited by ricebowl; 10-31-12 at 12:55 PM.
#88
Beer >> Sanity
Join Date: Oct 2010
Location: Colorado
Posts: 3,449
Bikes: 2014 Evo DA2, 2010 Caad9-4, 2011 Synapse-4, 2013 CaadX-disc
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
I think you'd want to do this based on time rather than placement. 13th out of 499 could be 30 seconds behind 12th, and 1st through 12th might be separated by 10 seconds. Based on the speed of the slowest riders, you might be in the 98th percentile, but, if everyone is grouped after you, you might be in the 90th percentile.
#89
internettubes engineer
Join Date: Feb 2012
Location: Mesa, Az.
Posts: 305
Bikes: 2012 Felt Z85
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
There is seriously a lot of whining going on in this thread... For segments where they run a light, etc... Lame.
For segments where they KOM using groups / wind / whatever else... who cares? Life isn't fair, and if you thought that a sport that attracts competitive people would be played on fair ground... you're an idiot.
It just makes you sound like a ***** to come onto a bike forum and whine about that instead of HTFU and doing better, or accepting the fact that you'll never KOM on that particular segment and just to ****ing accept it.
For segments where they KOM using groups / wind / whatever else... who cares? Life isn't fair, and if you thought that a sport that attracts competitive people would be played on fair ground... you're an idiot.
It just makes you sound like a ***** to come onto a bike forum and whine about that instead of HTFU and doing better, or accepting the fact that you'll never KOM on that particular segment and just to ****ing accept it.
#90
abandoning
Thread Starter
Join Date: Mar 2006
Posts: 2,068
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 1 Time
in
1 Post
There is seriously a lot of whining going on in this thread... For segments where they run a light, etc... Lame.
For segments where they KOM using groups / wind / whatever else... who cares? Life isn't fair, and if you thought that a sport that attracts competitive people would be played on fair ground... you're an idiot.
It just makes you sound like a ***** to come onto a bike forum and whine about that instead of HTFU and doing better, or accepting the fact that you'll never KOM on that particular segment and just to ****ing accept it.
For segments where they KOM using groups / wind / whatever else... who cares? Life isn't fair, and if you thought that a sport that attracts competitive people would be played on fair ground... you're an idiot.
It just makes you sound like a ***** to come onto a bike forum and whine about that instead of HTFU and doing better, or accepting the fact that you'll never KOM on that particular segment and just to ****ing accept it.
#92
Junior Member
Join Date: Oct 2012
Location: Sheffield, UK
Posts: 6
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
#93
Administrator
Join Date: Apr 2005
Location: Hudson Valley, NY
Posts: 32,978
Bikes: Merlin Cyrene '04; Bridgestone RB-1 '92
Mentioned: 325 Post(s)
Tagged: 2 Thread(s)
Quoted: 11952 Post(s)
Liked 6,601 Times
in
3,466 Posts
Let's tone it down in here, gents.
__________________
See, this is why we can't have nice things. - - smarkinson
Where else but the internet can a bunch of cyclists go and be the tough guy? - - jdon
#94
abandoning
Thread Starter
Join Date: Mar 2006
Posts: 2,068
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 1 Time
in
1 Post
feel free to delete mine if you want, billyd. it was kind of tongue in cheek meme, but seems a bit more intense when followed by a post calling me a *****.
#95
Junior Member
Join Date: Sep 2012
Posts: 21
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
V3 supposedly fixes a lot of issues, but, makes it much harder to write single-page apps. Even my tweetbot got throttled by their API and it was only checking to see if I had ridden, though, perhaps a bit aggressively.
Even though they send through the position number incorrectly, the time that they are sending in the JSON response is their 'corrected' time. So, with a condition check to see if the current time matches the prior segment time and a tiny bit of ugly code, you could replicate their leaderboard.
#96
Senior Member
Join Date: Mar 2010
Location: Seattle, WA
Posts: 23,208
Mentioned: 89 Post(s)
Tagged: 0 Thread(s)
Quoted: 18883 Post(s)
Liked 10,646 Times
in
6,054 Posts
#97
Senior Member
Join Date: Oct 2008
Location: Chicago, IL
Posts: 2,745
Bikes: S-Works Roubaix SL2^H4, Secteur Sport, TriCross, Kaffenback, Lurcher 29er
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
I love it.
4 pages of hand-wringing, postulating, reverse engineering, and general whatnot about .... average speed.
4 pages of hand-wringing, postulating, reverse engineering, and general whatnot about .... average speed.
#98
Descends like a rock
is this supposed to be a gotcha? of course i care about the numbers if they are being manipulated. i care that whenever it is really windy, my phone blows up with notifications saying that a kom on a flat segment that i didn't know existed was beat by 3s by one of the gamers. i care that when i look at the kom climb that starts with a stop sign, i have to parse through the performance graphs to see what is real and what isn't. that is lame.
so you wouldn't mind if all of those veloviewer segments in your signature got bombed by dudes on mopeds? how about if we just stocked all of the segments with random data - you wouldn't care? look, i don't care what place i am on a segment. what i DO care about is the fairness of the segment times and the ease of parsing through the crap for realistic results.
so you wouldn't mind if all of those veloviewer segments in your signature got bombed by dudes on mopeds? how about if we just stocked all of the segments with random data - you wouldn't care? look, i don't care what place i am on a segment. what i DO care about is the fairness of the segment times and the ease of parsing through the crap for realistic results.
Yes, it would be lame for someone to go through on a moped and screw up the data, but that's part of having an open public data service like this. But anything anyone does on a bike is fair game. If it happens, I won't really be upset.
I have some KOMs I got on windy days because many of the segments I ride are on my commute. I ride them every day. My odds of getting the KOM are decent on my route to and from work because if there's a day with perfect wind conditions, I'm probably going to be riding. There is one day I had to drive and it was really windy and someone stole my KOM. That sucked, but I laughed because its funny - not something to get all worked up over and declare that Strava is lame.
#100
Descends like a rock
To make a new segment, go ride it with a supported GPS device - Garmin or cell phone app. Then log into Strava and upload your ride. (or it will already be there if yo used a cell phone app). From the ride, at the bottom where all the segments are, there's an option to create a new segment. From there, you set the beginning and end point of the segment.