"Weight" doping


(James Newman) #1

I was riding this morning between 10 am and 11 am EST. During this time both the KOM and Sprint jersey wearers never broke 5 w/kg but had sub six second sprints and sub 30 KOM times. How is this even possible? They must have put in extremely light body weights in their profile is the only way that I can figure it out. Cheats like this ruin the overall gameplay aspect. Without structured workouts and allowing people to “game” the game I can see stick around much less paying for this service.


(Christian Wiedmann [X] 50) #2

The kg in w/kg is weight, so that would not be possible. Most likely the server isn’t sending updates to the w/kg number reliably.


(Eric Oshlo) #3

I’m not sure that’s correct Christian. It depends upon how Zwift is calculating riders’ speeds; particularly, uphill. IF the speed is drawn from w/kg (i.e. everyone generating the same w/kg would be going uphill at the same speed regardless of actual weight), then putting in a lower weight would generate a higher w/kg number and thus a faster uphill speed at a given power meter output.


(David Hewes) #4

Eh, give it time. The Zwift team will weed this stuff out. You are always going to have people that feel the need to one up everyone else, even if it means cheating to do it. Just like those kids you group up with playing any game and they would quit in the middle because they were losing. I have people that come into the shop and ride once a week and have got on Zwift and average 20+mph and 350+w every ride. I think there are people that just don’t enter the correct info. It is the best thing since Strava and they will clean up to get it right.


(Christian Wiedmann [X] 50) #5

@Eric - yes exactly. The key is that they would have higher w/kg. James was wondering if weight doping could explain a higher speed with the same w/kg number.
I suspect what happened is that the riders’ w/kg numbers weren’t updating for some reason so they weren’t actually maintaining 5w/kg - that’s just what James saw. I haven’t seen any evidence so far that Zwift’s model gets things so wrong that some riders are faster uphill with the same w/kg.
Both sub-six-second sprints and sub-30-second KOMs are within the realm of possibility, so I think in this case there may have been no cheating/bad setup.


(Jon Mayfield) #6

Theoretically a heavy rider could take both a KOM and a Sprint jersey in the current map with low’ish watts per kilogram. CdA is still the biggest factor, and for something like a sprint watts/CdA is really what matters. Even the KOM is not much of a climb in the current direction the course runs. Right now everybody has the same CdA regardless of weight.

I can’t say for sure people are or are not cheating without names and digging through the logs, but aside from some obviously bogus times (like 5 second KOMs due to bugs in the code) the riding I see generally looks legit.

Like strava we’ll eventually have to add self policing mechanisms so that performances can be flagged for analysis. We will never totally beat cheating, much like Strava will never know the KOM wasn’t on an E-Bike, motorcycle, or car.


(Christopher Pallotta) #7

Jon@Zwift: My understanding is that equal w/kg output would give similar performance on Zwift, despite whatever weight was input (and with no draft effect). Is that correct?


(D W - ltb LMDGD) #8

I’m curious if it’s actually a net advantage to ‘weight dope’ with a higher weight under certain combinations of rider ability and weight. Basically, what I’m wondering is if some people would gain more on the downhill than they lose on the uphill by inputting a higher weight. I do know that it can be a huge advantage for the current green jersey as I watched someone who apparently weighed 350+ lbs achieve an unreal time with believable wattage.


(Jon Mayfield) #9

Christopher, I mentioned in my previous post that watts/CdA is probably what determines who “wins” the current jerseys with this course profile in the the current direction of travel. Since everybody has the same CdA, it’s all about watts. The sprint has the added bonus of being downhills, where a heavy weight will be an advantage.

In short, on flat and downhill segments watts/CdA are king. On slightly upsloping courses its a mix of w/CdA and w/kg, and on 20% grades it’s entirely w/kg.

A person with a high weight therefore could get a jersey even though their w/kg were nothing stellar. w/kg is the perfect metric to determine if a user is cheating, which is why we show it. If their w/kg looks legit, it probably is.

WalterWong, totally. You’ll probably be slower on the orange and KOM, but the sprint might be that persons to own.


(Christopher Pallotta) #10

Jon@Zwift: Thank you for the detailed explanation. It seems, then, that at this time, heavier riders have an advantage in Zwift that they don’t have in real life. In the real world, a heavier rider would generally have a greater CdA which isn’t currently being accounted for. They’re bigger, so have more they have more air-blocking mass.

Heavier and taller riders can often produce greater raw wattage also, solely because of greater muscle mass. The benefits are negated to a certain degree by the added weight and CdA.

I would think CdA could be calculated by the height and weight to include this in the future. Keeping everyone’s CdA the same is skewing the results.


(Andrew Williams) #11

Provided that they have the power to match, heavy riders are definitely getting a great deal on Zwift nowadays.

My teammate who is over 200lb is easily doing 7:30 lap times at around 3.5 w/kg, which is threshold area for him. For me to match that, it takes a VO2max effort of darn close to 5 w/kg.

In real life, big strong riders might have a slight advantage on a flat course, but it’s not so pronounced. A course like this one with 50’/mi would not favor them (it wouldn’t really favor climbers either to be fair).


(Josh Currier) #12

How does ‘believable wattage’ even become a tangible thing on the sprint when you can only see the w/kg of the other rider? Isn’t that doing math with half of an equation? If you see me take the green jersey with 7.5 w/kg, how much wattage was I actually putting down? It’s a downhill, so w/kg is basically irrelevant.


(Jon Mayfield) #13

w/kg is how you would determine realistic human performance. The absolute value of the wattage doesn’t really matter.


(Josh Currier) #14

@Jon - In the context of the sprint, I think it definitely matters. A rider half my weight showing higher w/kg may not beat my time for that segment if I’m putting out 400w more than the other rider.

me @ 120kg, 900w = 7.5 w/kg
other rider @ 60kg, 500w = 8.3 w/kg

I’m fairly certain I’d win that sprint, so how is w/kg as a singular bit of information valuable for that scenario?

Again I’m talking about the sprint here. If we’re talking only about the climbs I’d obviously agree 100%.


(Jon Mayfield) #15

Josh, the point of showing w/kg is not to determine who you think might win a sprint, the point is to determine if its a possible human performance. In your example, both 7.5 and 8.3 seem reasonable, and thus we could assume some legitimacy.


(David Hewes) #16

It also comes down to what you can maintain for a given distance. Sure, most good riders can hit 900w at some point, but how long can they sustain that is what it comes down to. I am 186lbs, but I can maintain 1200+w for 8-10 seconds. Whereas someone else lighter may hit 1200w, but not maintain the effort for as long. I agree, I don’t think its possible to hit 58.1mph on that particular sprint with those “low” wattage numbers. At 1200w, my wpkg is 14.4, but only for that brief time. I have seen people fly by me and maintain 12wkg for minutes, not possible at all, but it is still young and things are still being hashed out. Overall, it is awesome!


(Jeremy Brazeal AETNA R/T (A)) #17

its possible they were drafting of one or more of the AI’s. Which would not explain the lack of a sprint though. I have seen riders do that quite a bit and it is completely legit and a good use of strategy. Keep in mind also that having a 5w/kg LT is elite level. Which would mean a 140-145 LB rider can average 320+ watts for 20 minutes or more. Its the riders who are averaging 7w/kg for long periods are the riders who are truly suspect.


(Josh Currier) #18

@Jon - thanks for the clarification, I see where you’re coming from now.


(mike mbarton (A)) #19

@Jon:

re: “downhill W/kg and Cda are king”  This isn’t entirely true, heavy riders should have an advantage on descents…should be: pure W, weight (numerator), grade, and Cda to calculate speed going downhill, right? 

For example, all things being equal on a 10% downhill at 250W, a 180 lb rider will be about 1 mph faster than a 160 lb rider.  This also jives with IRL where the big boys lead the skinny guys down the hill (unless the little guy is drafting in the aero-tuck).  But, in the aero-tuck, the little guy’s power is probably not 250 W…for example, at 125 W for the aero-tuck little guy, the speed difference jumps to almost 3mph. 

Here’s a simple online calculator that you can play with:  http://www.kreuzotter.de/english/espeed.htm