Claiming a PR from GPS instead of a race

GIS wrote:

They just keep coming. The idiocy of an app-for-everything world.

You CANNOT use commercial GPS to accurately measure a running distance. It does not work. It is in accurate by at least 15 m/s. Trees obsfucate GPS transmission. Commerical GPS can only get down to 15ft above the ground if nothing else comes first.

Jones counter. Steel tape. Chains. Even a Keson wheel. But not GPS.

GPS does not have a lot of precision.

You are saying a GPS is inaccurate by 'at least 15 meters'? Over the course of a 10K? That's plenty accurate to call a time/distance run legit. For a 35 minute 10k runner that's a matter of a few seconds.

GPS wrote:

You are saying a GPS is inaccurate by 'at least 15 meters'?

Per second.

The person you replied to is right. What everyone recognizes as GPS is extremely inacurate.

15 meters per second? Four minute mile pace is 6.67 meters per second. What kind of math is this?

Last two 5Ks I ran on certified courses, GPS showed 3.09 and 3.12.

Going from memory, the 15m accuracy is the theoretical accuracy of the pseudorange solutions used by cheap GPS receivers. In practice, many of the sources of error are correlated so something near 8m is expected with clear skies and good satellite coverage. A couple things about watches make things worse (crappy antennas, your body blocking half the sky --> fewer satellites and worse geometry...) , I imagine most watches/apps apply a smoothing algorithm, & possibly use accelerometers to minimize problems from dropped signals etc.I use an iPhone with strava. The distance measurements don't seem too bad, but the instantaneous pace is atrocious. I did a 5k, official time of ~18:50 but strava noted an 18:05 5k... Not sure if strava just thinks a 5k is three miles or if the gps was that wrong. A similar thing happens when I run on a track that has a strava segment on it... it extrapolates my lap times to ridiculous mile paces.

GPS wrote:

GIS wrote:

They just keep coming. The idiocy of an app-for-everything world.

You CANNOT use commercial GPS to accurately measure a running distance. It does not work. It is in accurate by at least 15 m/s. Trees obsfucate GPS transmission. Commerical GPS can only get down to 15ft above the ground if nothing else comes first.

Jones counter. Steel tape. Chains. Even a Keson wheel. But not GPS.

GPS does not have a lot of precision.

You are saying a GPS is inaccurate by 'at least 15 meters'? Over the course of a 10K? That's plenty accurate to call a time/distance run legit. For a 35 minute 10k runner that's a matter of a few seconds.

I'd say anything under 30:00 needs a race to be measured. Over 30:00 you shouldn't care.

Mind telling us why you are following this guy on instagram?

Bubbadeeboos wrote:

I did a 5k, official time of ~18:50 but strava noted an 18:05 5k... Not sure if strava just thinks a 5k is three miles or if the gps was that wrong. A similar thing happens when I run on a track that has a strava segment on it... it extrapolates my lap times to ridiculous mile paces.

Not following this. I can understand a GPS watch having inaccurate distance, but not time. Even a dollar store watch keeps accurate time. Were you truncating the distance to 5k?

wut wut wrote:

15 meters per second? Four minute mile pace is 6.67 meters per second. What kind of math is this?

Last two 5Ks I ran on certified courses, GPS showed 3.09 and 3.12.

On a very simple level the kind of math is a combination of algebra and geometry (above most of LRC's heads). The earth is not flat*. GPS estimates are the result of triangulation* so at any given moment if the transmission is off you might be allocated 15m or more from your actual position.

Where did I get the 15 m/s margin of error? Class. I went to a top 25 institution and in my GIS courses this figure was beat into our heads. It was repeatedly asked on homework, quizzes and exams to the point where it became a gimme. I think there was an overall point conveyed by its frequent appearances in graded material.

GIS wrote:

They just keep coming. The idiocy of an app-for-everything world.

You CANNOT use commercial GPS to accurately measure a running distance. It does not work. It is in accurate by at least 15 m/s. Trees obsfucate GPS transmission. Commerical GPS can only get down to 15ft above the ground if nothing else comes first.

Jones counter. Steel tape. Chains. Even a Keson wheel. But not GPS.

GPS does not have a lot of precision.

Agreed! I was just about to comment on this.

How many times have we covered this???GPS is not accurate - it's a guide at best. Maybe if you’re out on the farm running after cows to milk, but not for city folk. Yes it's pretty damn close, but unless a certified course is bragging about measuring their course on device X3948BCKY80009Z and it's the same as the wheel, then it's not accurate tool.

Onto the PR saga. Keep it simple, no gray areas with PR's, wins, or whatever else. Doesn't matter if it's instagram, hipstagram, stravagram, garmingram, or whatever the latest flavor is of the day to engrave your trophies with. No results count unless it's a certified race. Period.

pop_pop!_v2.2.1 wrote:

GPS wrote:

You are saying a GPS is inaccurate by 'at least 15 meters'?

Per second.

The person you replied to is right. What everyone recognizes as GPS is extremely inacurate.

Are you telling me you run 15 meters a second?

For one track that I run on, GPS estimates my speed as being ~20 secs/mile too fast. On another track it's ~20 secs/mile too slow. I'd trust GPS if the course follows a more or less straight line if there aren't any obvious problems in the data (if the course jumps around a lot), but not if it were from a track.

Why the fukk would you need a gps on a track? Do you need one on a treadmill?

I use a much more reliable way to measure my runs. I go by the mile markers on the highway.

morans... wrote:

pop_pop!_v2.2.1 wrote:

Per second.

The person you replied to is right. What everyone recognizes as GPS is extremely inacurate.

Are you telling me you run 15 meters a second?

No, idiot!

GPS is the result of continuous triangulation. For every second that your gizmo transmits data (which is every second) the triangulation from commercial equipment (device and satellites) can be 15m off. That's where you get the 15 m/s from.

Would it count if it were on a track but not a race?

morans... wrote:

pop_pop!_v2.2.1 wrote:

Per second.

The person you replied to is right. What everyone recognizes as GPS is extremely inacurate.

Are you telling me you run 15 meters a second?

No. Imagine it another way. Each second the little computer tries to find some satellites. Maybe it does.. Maybe it doesn't.. each sample is off by meters. Roll all those bad readings and estimates up into a GPS file.

Strava accepts this garbage and acts like it is accurate. It's a bunch of wildly varying points. Nice pictures and graphs of garbage though.

im enjoying people trying to out this guy as a hobbyjogger after he said he followed this guy as someone to look up to, despite only having a 35 min 10k.

No, you can't claim it as a PR. "Records" and "PRs" must happen in a timed race with competition. If Ritz went out tomorrow, ran a long run, then uploaded a photo of his GPS watch that said "26.2 miles, 2:01:03" would you award him the new WR for the marathon? No.

Also, I get it. The dude is a 35 min 10k guy, so why should anyone care? I agree, there's no point in getting worked up about it, but he's only screwing himself over if he is claiming it as his own PR.

GIS wrote:

On a very simple level the kind of math is a combination of algebra and geometry (above most of LRC's heads). The earth is not flat*. GPS estimates are the result of triangulation* so at any given moment if the transmission is off you might be allocated 15m or more from your actual position.

Where did I get the 15 m/s margin of error? Class. I went to a top 25 institution and in my GIS courses this figure was beat into our heads. It was repeatedly asked on homework, quizzes and exams to the point where it became a gimme. I think there was an overall point conveyed by its frequent appearances in graded material.

Either you're confused by what you heard or misinformed. Yes, you can get errors of 15m from GPS. No, it's not 'per second', it's strictly for the individual readings.

I don't think pop_pop or GIS really understand this GPS thing after all.... You guys are aware that the watches don't ONLY rely on GPS data, right? You also understand that they don't use RAW GPS data, right? You also understand that between Strava and Garmin you'll see differences in track data/overall distances, right? There's this amazing thing called smoothing, which helps to eliminate garbage data. Strava also has this other amazing thing where they take the "popular" track data and use that to fine-tune yours to... get this.... eliminate bad data. I know, who would've thought that in 2016 we have the ability to smooth out data points? Crazy.