Digital EPO - how to tell when someone has used it

deadhead1971

Thanks to one of my blog readers for noticing this, and taking the time to write a comment.
I've written up a post with a few screenshots showing you how to tell if someone has used digital epo to increase their speeds in order to cheat their way up the leaderboards on Strava.

http://www.scarletfire.co.uk/2013/06/how-to-tell-if-someone-used-digital-epo-to-cheat-on-strava/

supermurph09

Interesting to know, thanks.

ddraver

S'cuse me one second - OH MY GOD WHO CARES!!!!!!!!

Ok back now - that's a very useful post...

diy

I think a bit of fake data on strava is probably a good thing. If strava gets too reliable we'll all face fines for racing on the highway.

Steve236

Strava users should probably be required to subscribe to a "digital passport" which will demonstrate the consistency of their speed traces of a period of time.

stueys

ddraver wrote:

S'cuse me one second - OH MY GOD WHO CARES!!!!!!!!

Ok back now - that's a very useful post...

Amen, until the UCI adopt Stava to determine races it's all a bit of fun isn't it?

lawrences

I'm going to switch back to regular EPO. Much, much harder to detect.

buckmulligan

I find the intricacies of data manipulation quite interesting really.

All this post really shows is how crude Digital EPO's 'data juicing' methods are. To me, it looks like it just removes GPS location data points at a set interval and shuffles the timing data to match; so for example for a 5% boost would remove 1 out of every 20 data points, producing a spike in speed every so often as you travel the distance for two location data points in the time for just one.

The raw GPS and timing data recorded by these devices is very simple and there's definitely more sophisticated ways to do this without a boatload of processing power. But I suppose as always, making it work (and look vaguely convincing) for the vast variety of devices and data formats out there is the real challenge!

Anyone interested in this kind of stuff might enjoy this (slightly more comprehensive!) article:

http://blog.veloviewer.com/41mph-the-ev ... s-article/

buckmulligan

Just thinking, that if my theory about their data processing is correct, using a 100% boost (i.e. removing 1 in every 2 GPS location points) wouldn't produce those little speed spikes they show, it'd be a nice smooth trace again. If you want to cheat undetected you could go for a nice gentle country cruise and then boost it by 100% when you get home and still sweep up all the KOMs!

I guess you'd have to make sure you don't accelerate or brake too hard because going from 0-50 kmph in 5 secs might be a bit of a red flag!

sloppyschleckonds

Proof if it was needed, that all discussions about Strava aren't dull.

thefd

Interesting to see...

How do you zoom in on the performance? I cannot seem to see how to do it!

adr82

TheFD wrote:

Interesting to see...

How do you zoom in on the performance? I cannot seem to see how to do it!

Just click and drag with the mouse over a section of the graph and it'll expand that section to fill the whole width.

Initialised

I lost a KOM this morning, flagged it up and I back to joint 1st now.

82.3MPH on a bike round these parts, as if. I recon he left the app running after getting in the car. I wonder if I should contact the police since strava will have the GPS logs.

sigorman85

Had the same on beat everyone's times by 4 odd mins .... What a joker lol

diy

Initialised wrote:

I wonder if I should contact the police since strava will have the GPS logs.

Given that you have both committed road traffic offences, I would think that unwise.

pkripper

diy wrote:

Initialised wrote:

I wonder if I should contact the police since strava will have the GPS logs.

Given that you have both committed road traffic offences, I would think that unwise.

And what would they be then? I can understand the 82mph one..