Data and techie stuff thread for BAFS 2016

[DismalScientist]

Ummm… SteveO, check the distances on those rides you posted above.:rolleyes:

[Vicegrip]

Dave and Amalitza might not have ridden the same 1/2 route

Strava knows where the bike trails are. If you have not gone to Strava via a laptop and sorted your profile and desired settings you might want to. there might be a setting that lines toy up with trails over roads when close.

I have compared the eye-phone to a couple of Garmin Edges and found the phone to be slightly less accurate in plotting and a fair bit worse on elevation in steady weather. During a weather change the phone is better as my Edge 1000 has barometric compensation and it gets faked out by pressure changes. Both were WAY better than my old Crackberry.

Took a look at some of the monster cross rides. There seems to be some agreement when close to the same device is used over the same course. Some people skipped a loop here or there and the miles show this. Some rides that were recorded using a Garmin Edge 800 or 810 are about the same. Joe B used a Garmin forerunner which might not have barometric compensation.

[Anonymous]

We did ride the same route. The 25 mile ride was one lap. The 50 mile ride was two laps of the same route. SteveO is correct that doubling my data *should* result in the same amount of climbing as everyone else.

As it happens, I am not unaware of the tendency of my rides to somehow involve less climbing than the other people I’m riding with… I’ve even had a theory about that … https://www.strava.com/activities/252314836

However, it does seem to fairly consistent with itself– the routes that I ride regularly do not have wildly varying reported elevation changes, so it’s useful enough in comparing this ride vs. that ride in terms of difficulty, or this year vs. last year have I done more climbing.

[Vicegrip]

as others have pointed out I think it might be a phone rather than Garmin device thing. Elevation is harder to ascertain via GPS than lat/lon. When you stop and think about how a GPS system works and how much goes into it the mind boggles how accurate it is.

[LeprosyStudyGroup]

Strava FAQ about elevation is interesting

https://support.strava.com/hc/en-us/articles/216919447-Elevation-for-Your-Activity

If you track gps from a phone, any “climb” under 10m is ignored and the ending data is “smoothed” to reduce noise on top of that – which I already knew thanks to a few years of mildly hilly commute trips that always showed up as 0 elevation gain.

[bentbike33]

Thanks for all the input on my Strava/iPhone issue. I guess I’ll just live with what I get including the odd 40-mph uphill sprints through solid objects and other weirdness (maybe make them segments and finally get a KOM).

[Anonymous]

nah, I still think strava deducts feet if you don’t get to the top fast enough.

😎

[vvill]

GPS units with barometric altimeters are generally the most accurate for elevation but they are skewed by weather changes and can be completely useless if it’s raining (especially heavily) as the sensor can’t seem to detect air pressure properly. GPS units like the Garmin aren’t terribly consistent across units however. My Garmins all give quite different temperatures, elevation, etc. on the same ride, even with the same model (Edge 500).

Rides tracked with phones (or GPS units like the Garmin Edge 200 which don’t have an altimeter) will usually have lower elevation because they’re based on long-lat lookups to known terrain profiles. Fine for a long steady climb, but you can miss a lot of the elevation on rollers, trails, etc. since the terrain data is not that detailed. The main exception is if you cross a river on a bridge you basically get credit for going into the water and climbing back up the bank – you can see this on the Key Bridge over the Potomac, for example. If you wanted some easy climbing you could probably do bridge repeats on a device without a barometric sensor.

And as mentioned, Strava (and actually pretty much anything that derives elevation gain from ride data) will also smooth the data to some extent, and usually if there’s more smoothing you’ll end up with less elevation gain.

Most Garmin units have an optional “accuracy” field you can add to see how accurate the GPS reading is at the time – you might be surprised!

[Vicegrip]

Sample rate can skew things too. Phones tend to make less plot points couple this with Strava rounding and you can get some odd errors.

Spot on about baro and storms. I did a nice long bills repeat ride and the garmin recorded it as a single arched curve plotting the pressure drop of the storm. I corrected it in Strava and the climbs reappeared

[vvill]

@Vicegrip 135452 wrote:

Sample rate can skew things too. Phones tend to make less plot points couple this with Strava rounding and you can get some odd errors.

Spot on about baro and storms. I did a nice long bills repeat ride and the garmin recorded it as a single arched curve plotting the pressure drop of the storm. I corrected it in Strava and the climbs reappeared

Good point. A Garmin with 1 sec sampling vs a phone (or a Garmin with “Smart” recording on) will interpolate data more. I suspect it’s actually easier to get faster segment times on a phone because of this, but don’t quote me on that.

The correction on Strava will just go to lat/long lookup, I think.

[ChuckS]

Hello –

This ride doesn’t appear in my Team’s Activity Feed (Team 10 – Team X) ? :confused:

https://www.strava.com/activities/498236866
Chuck Seiders – 2:16 PM on Sunday, January 17, 2016 [actually ridden Saturday, February 20, 2016]
#BAFS2016 Afternoon Ride – 43 of 51 days
Bike Manassas group of 10 riders today temps almost spring like, and a mild head wind which ever way we turned(or so it seemed). “Elapsed Time 819:09:22” …wth is up with that??

Perhaps a device mishap??
How /where can one verify it has been included or excluded from the database that feeds the team data report http://freezingsaddles.com/leaderboard/team_text?

[rcannon100]

aaaaaaaaaaaahhhhhhhhhhhhhhhhrrrrrrrrrrrrrggggggggggggggg!!!!!!!!!!!!!!!!

[attach=config]11055[/attach]

[hozn]

@rcannon100 135543 wrote:

[ATTACH=CONFIG]11055[/ATTACH]

Yeah, no idea … Nothing changed, cron (scheduled) scripts running fine. The shared host seems to have broken. If they don’t get it up in short order, I’ll setup a new server. Don’t worry no data will be lost, but you might be riding blind for a bit (hopefully they can restart the service and it’ll be back up shortly; worst case I should still be able to build a new system within a day or so) !!!

[cvcalhoun]

@hozn 135544 wrote:

Yeah, no idea … Nothing changed, cron (scheduled) scripts running fine. The shared host seems to have broken. If they don’t get it up in short order, I’ll setup a new server. Don’t worry no data will be lost, but you might be riding blind for a bit (hopefully they can restart the service and it’ll be back up shortly; worst case I should still be able to build a new system within a day or so) !!!

Well, dang, my photo of this evening’s bicycling conditions will probably never make it to the front page!

[ATTACH=CONFIG]11056[/ATTACH]

[dbb]

@rcannon100 135543 wrote:

aaaaaaaaaaaahhhhhhhhhhhhhhhhrrrrrrrrrrrrrggggggggggggggg!!!!!!!!!!!!!!!!

[attach=config]11055[/attach]

It looks like the Cannon filter is working fine!

[Author]

Posts