This thread was originally titled, "Incredible development in the $612,000 Transcon Goodge run, currently ongoing" but the new title is more descriptive. The description of the run is here.
I trust your credentials but I just want you to do the dang analysis. Can you use the Strava data for now and point out areas where cheating may have occurred? Then whenever (if ever) Paul releases his garmin files we can use those to dig deeper and verify your findings?
You know you can download the Strava data behind those charts? Its all sampled at once per second. Is that not precise enough for your analysis?
Not a .fit format expert. But these fancier watches record more than Strava displays. It's not the resolution of 1 sample/second. It's that the original .fit files have stuff like ground contact time and other little metrics that could very much help confirm or deny his cheating.
I'm sure Runroblarun knows how to get the data from Strava. It's not even CS101. It's "open up dev tools 101."
I like how Sneakers sounds like the econ kids that quit math and compsci after one class, took their remedial "programming" courses for econ students, and now think they're gods gift to data science and software. Sweety, while you studied Excel, I mastered x86.
Well, I am a veterinary surgeon, with a PhD, the study of both involved stats, so maybe I don't claim to be the International King of Stats, but I am pretty handy when it comes to looking at data and evidence. So...probably have sufficient skills and maybe if I'm ahead of the curve to some, I still think a decent majority could see the Garmin "light", especially if I showed them, which I have, of course, offered to do.
Here's one of my published papers, by means of proof of the PhD, as it seems you're sceptical of everyone else's scientific ability apart from your own.
Royal Veterinary College (where I also gained a third higher degree), University of London, graduating 2001, 2002 and 2010, if you want to dig further.
If you think you are "pretty handy when it comes to looking at data and evidence" then how about demonstrating it to me and show me a GPX file you download from PJ's Strava pages that contain the 1 second interval data along with all the fields, including the cadence. Then, plot the raw cadence data and tell me what you see compared to the Strava graph of cadence. I think you will find that very enlightening and then you will feel very embarrassed if what you wrote about having skills is true.
If you think you are "pretty handy when it comes to looking at data and evidence" then how about demonstrating it to me and show me a GPX file you download from PJ's Strava pages that contain the 1 second interval data along with all the fields, including the cadence. Then, plot the raw cadence data and tell me what you see compared to the Strava graph of cadence. I think you will find that very enlightening and then you will feel very embarrassed if what you wrote about having skills is true.
I like how Sneakers sounds like the econ kids that quit math and compsci after one class, took their remedial "programming" courses for econ students, and now think they're gods gift to data science and software. Sweety, while you studied Excel, I mastered x86.
Not a .fit format expert. But these fancier watches record more than Strava displays. It's not the resolution of 1 sample/second. It's that the original .fit files have stuff like ground contact time and other little metrics that could very much help confirm or deny his cheating.
I'm sure Runroblarun knows how to get the data from Strava. It's not even CS101. It's "open up dev tools 101."
I highly doubt Runroblarun even knows how to open up dev tools.
As for ground contact time, you do know that is just derived from the accelerometer, right? The exact same accelerometer used for cadence? If you are using trekking poles like PJ or a push stroller like Rob, the ground contact time will indicate both PJ and Rob cheated in their runs across America. I'm sure Rob will appreciate that accusation.
Not a .fit format expert. But these fancier watches record more than Strava displays. It's not the resolution of 1 sample/second. It's that the original .fit files have stuff like ground contact time and other little metrics that could very much help confirm or deny his cheating.
I'm sure Runroblarun knows how to get the data from Strava. It's not even CS101. It's "open up dev tools 101."
I highly doubt Runroblarun even knows how to open up dev tools.
As for ground contact time, you do know that is just derived from the accelerometer, right? The exact same accelerometer used for cadence? If you are using trekking poles like PJ or a push stroller like Rob, the ground contact time will indicate both PJ and Rob cheated in their runs across America. I'm sure Rob will appreciate that accusation.
1) Strava doesn't return a gpx, it returns a JSON that approximates a GPX. So you're already wrong there.
2) Yes, I know ground contact time is derived from accelerometer. That does not mean it uses the cadence data reported and it's purely derivative of the data saved in the "GPX" that you think you're getting. There could be a black box function between the accelerometer and the recorded cadence values that calculated ground contact time.
Nope. Don't want to, as that will involve more work than I can be arsed with and seeing as I haven't looked at the files, I don't know if they are as accessible to the general public, which was my goal. Seeing as you two are so good, why don't you do everyone a favour and do it yourself. Produce a graph of the dodgy days. Take a photo, load on imgur. Post a link. Anyway, that's the last comms I have with you two on this subject. If you ACTUALLY WANTED to prove Paul right, or defend him, you'd have produced an overlay graph already.
But yeah, I'm done with trolling and can't wait for that new transcon attempt, where I'll be supportive once again and revel in their accessible Garmin plots.
1) Strava doesn't return a gpx, it returns a JSON that approximates a GPX. So you're already wrong there.
2) Yes, I know ground contact time is derived from accelerometer. That does not mean it uses the cadence data reported and it's purely derivative of the data saved in the "GPX" that you think you're getting. There could be a black box function between the accelerometer and the recorded cadence values that calculated ground contact time.
1) Do you need me to teach you how to download from Strava? It's very, very simple....
2) So you just want garbage data from a black box function? How is that supposed to prove anything?
We generate completeness maps and leaderboards for people that want to run every street in a city. It shows you which streets you've run and which ones you haven't based on your Strava data. So you "ferret out" all the streets. Same for cycling.
The live tracker was a side project we did for Paul because he asked us to one night during our run club and we thought it would be a fun thing to build since we were already integrated with Strava.
Is this different than CityStrides?
We're low key but we've been around for five years. Unlike our competition, we actually update our underlying map data on a reasonable timeframe (normally within a week). I always had streets that didn't exist in CS stick around for years because they would do map updates so infrequently.
We also do cycling, so you can see streets cycled and streets run in separate maps.
We don't do silly things like call a street complete after 90% completion or because you ran both ends and not the middle.
Basically we built our own site to do street hunting in order to fix all the things we hated about the alternatives.
Nope. Don't want to, as that will involve more work than I can be arsed with and seeing as I haven't looked at the files, I don't know if they are as accessible to the general public, which was my goal. Seeing as you two are so good, why don't you do everyone a favour and do it yourself. Produce a graph of the dodgy days. Take a photo, load on imgur. Post a link. Anyway, that's the last comms I have with you two on this subject. If you ACTUALLY WANTED to prove Paul right, or defend him, you'd have produced an overlay graph already.
Rob, I already created some of those graphs using PJs detailed data and posted them multiple times in this very thread. Looks like you just ignore the detailed graphs I made because they don't support your prejudices.
I knew you would chicken out when I called your bluff about having any technical skills in analyzing data. I didn't misread your character. I will let PJ know to continue ignoring you.
2) So you just want garbage data from a black box function? How is that supposed to prove anything?
You realize the cadence data is already "garbage data from a black box function," right? If you got raw accelerometer data, it'd be nonsense. It's been run through a black box to give you the numbers.
Nope. Don't want to, as that will involve more work than I can be arsed with and seeing as I haven't looked at the files, I don't know if they are as accessible to the general public, which was my goal. Seeing as you two are so good, why don't you do everyone a favour and do it yourself. Produce a graph of the dodgy days. Take a photo, load on imgur. Post a link. Anyway, that's the last comms I have with you two on this subject. If you ACTUALLY WANTED to prove Paul right, or defend him, you'd have produced an overlay graph already.
But yeah, I'm done with trolling and can't wait for that new transcon attempt, where I'll be supportive once again and revel in their accessible Garmin plots.
Goodgebye all.
Rob, I'm on your side. I'm literally at work now and quite busy but keep being drawn back here.
You said in your previous post that Rob already knows how to get the Strava data because he is so skilled, so why are you showing him that link again?
I'm willing to admit when I'm wrong :)
I can respect that.
When you get the graphs derived from the 1 second interval data, post them up and we can drag Rob back into the thread. I would upload all my graphs and GPX files but I'm the same as Rob, can't be arsed to do something very simple and quick.
Nope. Don't want to, as that will involve more work than I can be arsed with and seeing as I haven't looked at the files, I don't know if they are as accessible to the general public, which was my goal. Seeing as you two are so good, why don't you do everyone a favour and do it yourself. Produce a graph of the dodgy days. Take a photo, load on imgur. Post a link. Anyway, that's the last comms I have with you two on this subject. If you ACTUALLY WANTED to prove Paul right, or defend him, you'd have produced an overlay graph already.
But yeah, I'm done with trolling and can't wait for that new transcon attempt, where I'll be supportive once again and revel in their accessible Garmin plots.
Goodgebye all.
So you're going to accuse him of cheating but not put in the effort to prove it? I no longer believe that you're doing this for "transparency". You must be trolling lol
I'm just saying this in advance: I don't hate PJ. He's a qualified SWO, and I respect him. He's clearly on the spectrum and somewhat mentally ill, but he's not an evil dude. He has an unhealthy outlet (running and social media) and should get professional help.
I'll ask around for more stories, but here's one y'all should enjoy.
It's pretty normal for Navy commands to have meme pages, and apparently, when he was called out in one meme (on the Cowpens), he pitched a hissy fit and tried to have it shut down.
Nope. Don't want to, as that will involve more work than I can be arsed with and seeing as I haven't looked at the files, I don't know if they are as accessible to the general public, which was my goal. Seeing as you two are so good, why don't you do everyone a favour and do it yourself. Produce a graph of the dodgy days. Take a photo, load on imgur. Post a link. Anyway, that's the last comms I have with you two on this subject. If you ACTUALLY WANTED to prove Paul right, or defend him, you'd have produced an overlay graph already.
But yeah, I'm done with trolling and can't wait for that new transcon attempt, where I'll be supportive once again and revel in their accessible Garmin plots.
Goodgebye all.
So you're going to accuse him of cheating but not put in the effort to prove it? I no longer believe that you're doing this for "transparency". You must be trolling lol
I find it hilarious that Rob Pope is putting tons of hours into accusing PJ of cheating with comments everywhere and demanding the Garmin Connect links, yet won't put a single minute into plotting a simple cadence graph that would show him exactly what the raw data shows and how wrong his accusations are. In the words of UltrarunningKid, "This is what bad faith looks like."