Easy that when to send power data in from outdoor racing
Why did you leave out the last part of Paulās statementā¦ āand absurd power numbers beyond what pro cyclists can produceā?
Is it too much to ask of Zwift to make new users go through a tutorial before entering their first race in which you could educate your users about reasonable power figures for untrained, trained, elite, pro, etc cyclists?
1 Like
By way of example; Would it be easier to just bulk email anyone who is over 60 with an FTP of 6.0 and then verify the outliers that are real?
Cones used to fix this but seem to be less common now. The current approach is that the community randomly accuses someone of cheating mid-race and asks for others to flag them. This may or may not work but you end up in this weird toxic holding pattern.
A good example yesterday is where someone who had previously done hilarious numbers a few races back was off the front at fairly realistic numbersā¦ but people wanted to flag them because theyād cheated before. This is their Zpower profile which is āfantastic if trueā but they were doing 4.2 off the front at the time they were being accused of being a ārepeat offenderā.
Respectfully āenjoy a workout that feels a bit like a raceā might be a more accurate framing here. Itās not racing if the numbers are made up. Itās not really racing if genuine riders are all in, while someone on a spin bike is doing the same power at a Z2 heart rate.
at 6.02 wkg he need to be with the PRO and we need to see him irl on TV
If thatās what heās capable of sure. The problem with user-centric reporting is that he could have also had a weird calibration or borrowed a spin bike (hotel trip?) on one specific race and now be legit. Zwift is in the best position to know that. In game controls could solve thisā¦ but they donāt exist yet.
then all people need to do is email zwift please can you remove my race data from the last race as I think it wrong
1 Like
I would start by looking at the power output of pros. If you can beat Gannaās hour record numbers, you need to prove it or your equipment is faulty. If you can beat Jakobsen sprint numbers, you need to prove it or your equipment is faulty. Maybe there are random amateurs on Zwift who can do that. Bully for them, and prove it. I have no idea what the current limits are for coning riders, but those limits seem obviously too high if a random guy on a spin bike can do 600W+ for an entire race. Itās OK to start by turning the screws just a little and see what comes out. Then do it some more. Incorporate reported age into the calculation. Seek out the science on performance to understand what is reasonable and what is not, set the limits a little bit above reasonable. Itās OK to not detect everyone with bad equipment, just catch more of them, and then some more, and then some more. It may not be something that you can reasonably eradicate instantly.
3 Likes
āI would start by looking at the power output of prosā
There are plenty of cyclist who can produce more power that Jakobsen (not much over 1400 Watts), they just canāt do it for 18 seconds after 3 hours over a 160km course, somehow staying near the front and sprinting at exactly the right moment when heading for the line. Thereās a lot more to being a pro sprinter than just raw power.
A Gannaesque 440Watts - 5.4W/kg. There are riders who can beat those figures for 5 minutes, 10 minutes, 20 minutesā¦ So how long do you wait before you say DQād. Waiting the full hour isnāt going to solve the original posters desire to have quicker identification and elimination of cheaters.
So back to my question āand your definition of absurd power is what? ā¦for how long?ā Come up with some form of average power/duration curve outside of which you think automatic DQ should apply. Weāll let you have 400 Watts over 1 hour as an end point and maybe an Andre Grippel 1800 Watts for 15 seconds as the start point, but thatās a pretty poor graph so far.
Iād probably start with power (W or W/kg) that significantly exceed the P90 numbers (10% to be really generous):
So anyone of Zwift doing 5min power of 600 watts, or 8W/kg should appreciate the email from Zwift effectively offering a new career as a Grand Tour rider
2 Likes
Thank you Paul, thatās a more meaningful and useful answer that any so far for what absurd levels might actually be. Maybe add a āYouāre a Proā badge to let those that get DQād down gentle.
1 Like
And maybe itās a table like this that allows some results in, and DQs others, as itās looking at just wkg. A good example is under ZP profile 592010
However, if this is the case, then Zwift neglects to layer on other data, such as HR, to lend feasibility to the results.
Oh yeah - just the W or W/kg over a time is a relatively poor way of doing it - advantage is it is simple and pretty categorical, but on the other hand lots of false negatives. Zwift have much more data and can get much more precise with both just āin raceā data and also historic data (cranking out 600W at 140bpm and also 200W at 190bpm on different days would invite questions).
ETA: Oh boy, I looked at the profile. Apart from the ridiculous numbers the power curve is practically flat - 1min power ~= 20 min power, which is very unnatural especially in a crit.
ETA2: if you follow the bread crumbs you can find he was doing these mega-efforts (>300W) at <100bpm.
Yeah. But if you look to compare results that got through vs DQād, you might perhaps notice a trend that the allowed results seem to be less your chartās P75 20m and 5m numbers, and results that exceed the P75 (or maybe the P90 actually?) numbers are DQād. Though could easily just be a coincidence.