Is this another run cheat?

I tried to send a fb message and it said message couldn't be sent. How do you even do that? How do you have a message button on your personal profile but stop all messages being delivered? He couldn't possibly have blocked me messaging personally as we have had no previous contact.

1st May

https://www.strava.com/activities/986954051/overview

Crazy Fast splits….pace crazy

3rd May

https://www.strava.com/activities/986954027/overview

- No Cadence, strange data

8th May

https://www.strava.com/activities/986954019/overview

- Completely fake?

17th May

https://www.strava.com/activities/992511347/overview

- Weird Cadence

18th May

https://www.strava.com/activities/994519748/overview

- Just Weird

12th May

https://www.strava.com/activities/986959840/overview

- Completely fake?

27th May

https://www.strava.com/activities/1008294196/overview

- Just Weird

4th June

https://www.strava.com/activities/1021115492/overview

- Dodgy Cadence @ same changeover spot

16th June

https://www.strava.com/activities/1039592305/overview

-Dodgy Cadence @ same changeover spot

21st June

https://www.strava.com/activities/1047424818/overview

- Dodgy Cadence @ same changeover spot

8th July

https://www.strava.com/activities/1073688639/overview

- Dodgy Cadence @ same changeover spot

27th July

https://www.strava.com/activities/1104254489/overview

- Dodgy Cadence @ same changeover spot

30th July

https://www.strava.com/activities/1109040423/overview

-Dodgy Cadence @ same changeover spot

23rd August

https://www.strava.com/activities/990098781/overview

-Legit Run 50km

I think by June he was getting the hang of faking it with a bike....

Don't contact the suspect, as he might get suspicious and destroy evidence.

I hope someone has copied all his Strava and Garmin data.

DataIsGood wrote:

4th June

https://www.strava.com/activities/1021115492/overview

- Dodgy Cadence @ same changeover spot

16th June

https://www.strava.com/activities/1039592305/overview

-Dodgy Cadence @ same changeover spot

21st June

https://www.strava.com/activities/1047424818/overview

- Dodgy Cadence @ same changeover spot

8th July

https://www.strava.com/activities/1073688639/overview

- Dodgy Cadence @ same changeover spot

27th July

https://www.strava.com/activities/1104254489/overview

- Dodgy Cadence @ same changeover spot

30th July

https://www.strava.com/activities/1109040423/overview

-Dodgy Cadence @ same changeover spot

23rd August

https://www.strava.com/activities/990098781/overview

-Legit Run 50km

I think by June he was getting the hang of faking it with a bike....

Now that's what I call an eye for the details. "Same changeover point". And let me add my bit here too, that point is ALSO the start and end point of the run!!! Means, "convenience". This is the so called base point for this runner. All logistics are probably taken care of here, and any changeovers happening as a convenience happen here!

Beware the mileage champions wrote:

Now that's what I call an eye for the details. "Same changeover point". And let me add my bit here too, that point is ALSO the start and end point of the run!!! Means, "convenience". This is the so called base point for this runner. All logistics are probably taken care of here, and any changeovers happening as a convenience happen here!

It can't be a coincidence that at the same park along the same route his cadence turns to rubbish on each day.

p.s. It also happens to be the base of operations for the Marathon Films run by Vikram Bhatti and Vandana Bhatti, the duo who run his facebook page and are supporting him.

Just another coincidence I'm sure.

Good work.

I don't think the May 8th and 12th runs are completely fake. Likely just biked the whole way. He does a lot of u-turns, running back and forth, and for each u-turn there is a large dip in pace when he slows to turn around. I doubt he would have gone into such detail on a fabricated file.

Rtfm wrote:

Don't contact the suspect, as he might get suspicious and destroy evidence.

I hope someone has copied all his Strava and Garmin data.

Agreed, lets make sure scam_watcheroo gets everything he needs first.

DataIsGood wrote:

It can't be a coincidence that at the same park along the same route his cadence turns to rubbish on each day.

p.s. It also happens to be the base of operations for the Marathon Films run by Vikram Bhatti and Vandana Bhatti, the duo who run his facebook page and are supporting him.

Just another coincidence I'm sure.

Ok, my next observation is even more interesting and damning.

I selected each of the bar graphs shown on Strava (one bar representing one week), and selected each one of them, and then selected the first run on top which shows up. So, around 11 runs. So, basically Mumbai is prime/south-mumbai, and downtown/north-mumbai. His bad section is always (except one run) in the downtown area where it is easy to give a miss, and south is where there is the highest probability for someone to catch him and recognize him, and also his opportunity for the daily photo-op.

The exception run I mentioned above is this:

where the bad pattern does not follow daily trend.

scam_watcheroo wrote:

I have found a duathlete Garmin user (named Erwin) on Strava that has an upload of his run, bike, run and the cadence data oscillates between 0 and 124 spm just like for our suspect, Samir Singh (guess everyone has guessed who it is by now), who's cadence oscillates between 0 and 120 spm.

I had wondered if the 60 rpm readings were a built in default that the device spits out when the cadence measurements from the accelerometer are highly erratic. In other words the software says this guy is obviously moving so I will assume normal walking cadence. I guess Garmin will never admit to that.

Did this duathlete also have significant spikes in his speed measurements? I still have a problem rationalizing the speed recordings during the low cadence phases. There are frequent wicked accelerations and decelerations happening every couple of minutes. I did some quick calculations and reckoned at times this guy must be putting out over 1600 watts. That puts him in the same league as Mark Cavendish, a sprinter who has won 30 Tour de France stages. That frequent accel/decel profile would be more typical of a bike courier than a guy rolling along the open road.

I had always assumed that the speed calculation was based purely on GPS position and time, and that cadence would have no effect. I am wondering now if the algorithm includes some form of adjustment based on perceived acceleration or deceleration. The intention being to smooth the speed profile and avoid sudden step changes in data output. Unfortunately, if you get a sudden step change in one of the inputs - cadence - the output gets distorted. Obviously we cannot see the raw data that goes into the processor, all we see is the processor's output.

If the duathlete's speed data also shows this problem then that would go along way to supporting the bike theory in that we have a totally independent measurement system producing effectively the same result. (It also would be bad news for Garmin as it would imply this device is useless on a bike unless you can turn off the cadence feature.)

Somebody earlier mentioned that the physical transition point between profiles was consistent. Have you confirmed this? This would be very important.

I am prepared to accept that the evidence is starting to come together to infer deliberate cheating but I do have a very high threshold before publicly accusing someone. If we can rationalize the erratic speed values I think we will have crossed that threshold.

Just based on the fact he is riding around the streets of Mumbai I would expect more accel/decel than constant cruising speed. Of course I don't know anything about the specific roads at the times of day he is riding. He also zig-zags around town, and does a lot of u-turns in short out and back section over and over. A lot of 20+mph sections occurred on a downhill. I would not put much stock in acceleration wattage based off even the cleanest gps data, then add in turns, trees, buildings, wind, and hills.

I have gone back to look at some of my friends' stroller runs. The same cadence glitch happens when his watch hand is on the handle, but his pace accuracy is unaffected (we ran together). So I don't think cadence influences pace with Garmin.

The transitions to and from the bike are almost always at a start or end point of his run. He does multiple loops and out and backs each day and will run-bike-run-bike in various sequences. I have not seen evidence of two bikes, where he stops biking at one location and starts again at a different location. If he stops and stashes the bike, he will run for a while and circle back to the same spot and start biking again.

Yes, the two duathlete Strava uploads I analyzed also show a big jump in speed when they went from running to biking, similar to Samir's changes in speeds. You will notice that the duathlete's speeds stay up high instead of dropping down every so often, this is expected because the duathletes are trainining/racing and don't need to stop or slow down for something whereas Samir likely needs to slow down to take city corners and stops.

I have not checked if all the physical transition points between the profiles are consistent but I think some posters above have checked some of them and say a lot of them occur at the same spot. It will take me some time to looking through all this data.

Look at below run data, after turning on cadence. There are 3 points in the run where the cadence goes from good to bad to good to bad. Each of these 3 transition point location is same. So, the question is, what's happening at this location? Is there any other possibility which you guys can think about without the bike angle (like moving the watch to something else for charging, maybe another runner biking with him and manage his charging while he runs, a possibility which cannot be ruled out, though then the data will not be acceptable to purists. Or moving the watch to a backpack which has some charging equipment).

https://www.strava.com/activities/1026920333

Just found the first instance of two bikes (for what it's worth). He bikes from miles 12-17 starting and ending at the same spot, then runs back to his original starting point (not where he left bike 1), and bikes from miles 45-62 (end of run).

https://www.strava.com/activities/992511347/overview

1 2 3 green wrote:

Just based on the fact he is riding around the streets of Mumbai I would expect more accel/decel than constant cruising speed. Of course I don't know anything about the specific roads at the times of day he is riding. He also zig-zags around town, and does a lot of u-turns in short out and back section over and over. A lot of 20+mph sections occurred on a downhill. I would not put much stock in acceleration wattage based off even the cleanest gps data, then add in turns, trees, buildings, wind, and hills.

I am stunned that somebody doing any kind of ultra activity would be doing it in city streets. The stop - go of crossing streets would drive me nuts. I agree that u turns and sharp turns will mess up the data, it does on my watch. It will have the effect of amplifying the speed variations already present from the stop start nature of city travel.

I have only a little data from running in city centers but my best example is the Toronto Waterfront Marathon where I've run the half several times. Most of the race is on open boulevards and along the Lake Ontario waterfront and the data is good.

But the last few kilometers, after you pass the CN tower, is in the oldest part of the city center with relatively narrow streets and high rises either side. At that point the data goes nuts showing a pace of 16 seconds a kilometer at one point.

Scam W:

For completeness you should update your report and include comments about the cadence, the run/bike transition locations and how the city route he is following will affect the quality of the speed data. The most damning evidence the the two distinctly different average speeds and the consistency of the transition location. The loss of cadence data in the bike phase indicates he is no longer moving his arms and the erratic speed changes are a function of his chosen location.

You almost have to admire this guy, going out each day for 15 or more hours and jogging and cycling around city streets must be unbearably stressful. If he had claimed this as some sort of duathlon record you would have to give it to him. Also, he has achieved the paradox of producing data that is both absolute garbage and yet logically explainable. Not even Dave Reading or Amy Hughes achieved that.

I assume you are Indian.

- How well known is this guy?

- Would people easily recognize him if he was on a bike in Mumbai? Or is it so crowded that nobody pays attention?

- Is it common for Indian cyclists to hide their face (e.g. with sunglasses, mask against pollution, helmet)?

I am asking these question to assess if he would have done the biking parts himself or if he used an accomplice.

The watch he is using (Garmin Forerunner 235) is a watch aimed at runners. The cadence feature isn't meant for cyclists. They use other tools to measure their "cadence", by attaching something to their bike, not by wearing a watch on their wrist.

His new stunt run will take him through almost all Indian states. Seems like that will be a lot harder to cheat using a bike than running loops in a city. If he will be biking part of that course, an accomplice with a vehicle would be necessary.

Beware the mileage champions wrote:

Look at below run data, after turning on cadence. There are 3 points in the run where the cadence goes from good to bad to good to bad. Each of these 3 transition point location is same. So, the question is, what's happening at this location? Is there any other possibility which you guys can think about without the bike angle (like moving the watch to something else for charging, maybe another runner biking with him and manage his charging while he runs, a possibility which cannot be ruled out, though then the data will not be acceptable to purists. Or moving the watch to a backpack which has some charging equipment).

https://www.strava.com/activities/1026920333

Some interesting information:

Oh – one final interesting tidbit, you can indeed wear the watch while charging it at the same time, such as this: It might be a bit finicky to do this while running since the clamp design isn’t super-secure, but it would probably work if you were careful about how you arranged it. The unit DOES record during GPS activities while plugged into a USB battery pack. It does NOT record if plugged into a computer. Additionally, it does NOT record continuous HR while plugged in (likely because that lets in too much light).

The Garmin Forerunner 235 has an optical heart rate monitor built-in. It seems like he turned this off, maybe to save battery.

This device has a battery life of "up to 16 hours", while some of his Strava activities have an elapsed time of 19 hours. So it seems likely that he has charged the device during the activity, or he used multiple watches. However, I doubt that can explain all the inconsistencies.

Rtfm wrote:

Don't contact the suspect, as he might get suspicious and destroy evidence.

I hope someone has copied all his Strava and Garmin data.

If he destroys all evidence of other runs that makes him look even MORE guilty so i'm sure he wouldn't be crazy enough to do that.

Is he on Twitter, Instagram, etc? He cannot be contacted on facebook anyway, as has somehow found a way to block all messages getting to him whilst displaying a message function. Odd.

Raddison wrote:

You almost have to admire this guy, going out each day for 15 or more hours and jogging and cycling around city streets must be unbearably stressful. If he had claimed this as some sort of duathlon record you would have to give it to him. Also, he has achieved the paradox of producing data that is both absolute garbage and yet logically explainable. Not even Dave Reading or Amy Hughes achieved that.

Admire? Have you seen how much money he's making in donations? Over 12 grand just for the latest faked stunt. I bet those dollars go a long way in India. You think people would give him money for cycling?

The big deal is it's supposed to be RUNNING.

But yes, he's gone to a bit more effort with data than others.

Battery life

The battery life on the watch should be no issue. According to interviews he runs from 5am-1pm and then resumes at 4pm to finish off remaining millage. Heaps of time to recharge. It's fairly consistent with the 16 hours durations with 12 hours moving.

Cadence.

Somebody mentioned it the watch could be charging in a backpack but I believe cadence still works on these watches in a backpack. Hopefully somebody could easy prove that.

Bike Theory.

Many people are assuming he rides the bike, I think it's just as likely he has an accomplice who would do it for him. Would mean no chance of recognition and more time to eat/sleep/recover.

Transition.

This run

https://www.strava.com/activities/1047424818/overview

shows after 8 hours the cadence changes. A 5am Start time lines up nicely with a 1pm Stop and him returning to the start. The second run being biked after a rest. Bike done by him or an accomplice. It does look like he does the last 6km on foot again on that particular run.