I ride with multiple friends in the ‘real world’ and we are all of similar fitness and power. We take turns at the front of the pack and have similar endurance.
But with Zwift I just cannot keep up with both of my RL friends and get completely smoked.
I think the issue is the weight mechanics.
I am a bit taller and do more gym work so I weigh in at 90kgs.
One of my friends is 78 and another 75.
That 12kg to 15kg weight difference seems to be completely punishing me in the app.
Yet in the real world it doesn’t seem to make much difference.
I am actually much stronger on hills in real life compared to both of my friends.
Just wondering if it is a bug or if I should just change my weight to the same as my friends?
I don’t really want to change my weight as I am sure it will unfairly impact leaderboards.
As such while it is really enjoyable riding with friends on the real world it really ‘sucks’ in Zwift. They need to keep stopping and waiting for me and it is getting very frustrating
There a few factors at play, including weight, height and power. If your weight and height on Zwift is the same as in real life, and your friends’ weight and height are all correct, is your power correct?
Are all your trainers different types and are they calibrated correctly? Is yours giving you the equivalent of having a poorly maintained bike in real life, or are theirs calibrated wrong? Could it be that they push harder on a trainer than in real life?
Then there’s the bike you choose in Zwift. This should only make a small difference - unless you’re on a TT bike (which doesn’t draft) or a mountain bike.
If you want things to be as realistic as possible, check each of these factors.
Once you’ve checked all of thise, you could try playing around with your weight on Zwift. Some people will see it as gaining an “unfair advantage” but others will see you wanting to level out an “unfair disadvantage.” Your friends could be the best judges of this.
The key thing is to make sure Zwift feels fun and fair - for you and the people you ride with.
We all use Wahoo KickR V5s for the smart trainers. I haven’t done any calibration on mine but ill check to see if they have made some changes.
The bikes on the machines are different. I have my old Giant OCR One and they have other bikes maybe a Trek and something else.
Unfortunately I don’t have a power meter on my outdoor bike so I am not sure if it is correct but I dont think it would be materially different.
In terms of wattage for example we might be going up a hill and on a skype call ill be doing 300 watts and they will be doing 250-270 but they will be increasing the gap by 1m every 8-10 seconds.
If we all do 250 watts then they will be gaining a meter every few seconds.
I guess I am just not convinced that the power to weight ratio that zwift uses is a fair mechanic. It probably sounds good but its nothing like the real world.
I ride with a couple of blokes that are 6"6 and 110kgs and they are beasts on the road super strong.
I bet they would really struggle on Zwift with those dimensions.
Interested to hear if anyone else has different weight from their friends in the real world but struggles on Zwift to keep up or if they get too far in front.
Using “keep together mode” in meetups or group workouts isn’t cheating - it’s just a feature of the game to make it more sociable for people of different abilities. Some days you’ll be the slower person getting pulled along quicker than normal, some days you’ll be the faster person getting slowed down.
It’s a feature I’m not personally a fan of - because I like my riding to feel more realistic - but you should try everything in Zwift at least once.
Try the details above first though - and if you have a Kickr v5 you should also make sure you - and all your friends - have updated the firmware on it.
Gary, you’re not going mad. W/kg and height play a massive difference on Zwift, much more than in real life. If you’re taller and/or heavier than your friends, you will be penalised even on flat roads.
Similar to your experience, I can (just about) keep up with my wife on real flat roads; on Zwift, I’d be in z6 if I tried. And it’s absolutely and categorically not a miscalibration or erroneous power readings. Zwift have to simplify the calculations or it would take a dozen Crays to account for all the variables.
Unfortunately, however, there are still those deluding themselves that flat roads on Zwift are about absolute power and w/kg counts only on climbs.
The misconception here is that while absolute power is the main driver of the motion equation on the flat all other things being equal and power over weight is the main driver of the motion equation in climbs all other things being equal, that somehow gets translated into “weight has no impact on the flat”. Which of course is false.
No Cray (and even less a dozen thereof) is needed to calculate linear translation motion equations; only accurate data is needed. But since it’s unreasonable to ask every Zwift rider to take a day in a wind tunnel to measure CdA under various positions and enter those into the app, we’re left with a set of very approximate approximations based on weight and height. And that’s the source of most debates and data on discrepancies between Zwift speeds and real life.
@Robert_C
CdA is only one of the metrics Zwift could ask for more rider parameters like shoulder width, if you want to be more realistic then you would need a few Cray installations to do Realtime CFD of the draft and pack dynamics.
I think having Realtime data to the server and have every rider seeing the same thing is more important at this point.
If you have CdA (standing, on the hoods, in the drops, in a TT position - an hour in a wind tunnel should cover it ), you don’t need anything else aero-wise. Yes, failing that, the aero model could be improved with additional parameters; but we already have major debates with two (height and weight), imagine 3 or 4… It would probably be cleaner if they took out height and just assumed a fixed relationship to weight. Would remove one cheat. And ensure you don’t extrapolate those CdA = f(weight, height) too far off the source data. I have a hunch that’s also part of the problem.
If you get in all the way into drafting and pack dynamics, well yes, sure, pretty heavy CFD would be needed. But while this is a common complaint area, at least the weirdness of drafting and pack dynamics are common to all riders, while border cases of height/weight drive a lot more debate about “Zwift speed is not realistic”.
By the way, Crays are not great for real-time processing.
For me personally I see Zwift as something like cycling, I don’t compare my Zwift speed to outdoor speed. I compete against a set of people online if they are faster on Zwift than IRL then let it be we are in the Zwift world and in Zwift world they are better and I will train and get better than them that is my aim. Once we go out door things change I will be faster and stronger than without Zwift.
I just want things to be repeatable and feel almost like racing/riding my bike.
It will not be like riding your bike if the fundamental physics in Zwift is different from IRL. Yes, it is better training maybe, but when more and more people experience these anomalies and question them, I think it is time for Zwift to address these problems. Right now there is a competitor that is better in this area and their users highlight this as one of - or the biggest advantages over Zwift. If it feels wrong, it doesn’t matter if we get a new Tokyo world. These things are fundamental.
I think what Gerrie wants is repeatability within the Zwift game. Not necessarily an accurate physics model between Zwift and IRL. The way the code is written at this time, that is probably not possible so just make it the same every time we log on and the same on all of the routes.
But being old, tall and heavy is a bummer on Zwift especially on long climbs. The Radio Tower is an absolute killer for me.
It’s a good job I don’t do it in real life as if the bike wasn’t connected to the trainer I’ll fall over I go that slow.
Original Poster is correct in his interpretation IMO.
Zwift should look at it , if not to make a more realistic simulation but maybe more so as recent issues around weight and disclosure highlighted they have to be pretty sensitive and careful here . If Zwift are as is stated “punishing” weight and as a result users embark in unhealthy weight loss regimes as a result that could be quite problematic for the platform in the “wrong” hands.
I don’t think these issues are inconsistent. They behave the same way all the time so that is good, but that does not help anyone. I would be very surprised if the physics is different from route to route.
I think the source of the discrepancies and the excessive emphasis on WKG are deliberate, and of commercial nature, rather, to cater to regional markets where people are by nature on average lighter (Asia being ofcourse Zwift’s largest growth market). It levels the playing field a bit.
What you describ is a différence of real life and zwift over 100W, it’s to much i’m surprised with “only” 15kg.
If your friend weight 75 and you 90, in real life you must doing 50/60W more on climb for follow.
The faster you go the more power you have to get out than your friend on climb
If you are better on climb is that you have much more power than your friend and should outrun him easly on flat road because weigth is less restrictive here.
From what you say you should have better w/kg result than your friend
Therefore you should easily support the difference of watts output on zwift for follow your friend without burning yourself.
), you don’t need anything else aero-wise. Yes, failing that, the aero model could be improved with additional parameters; but we already have major debates with two (height and weight), imagine 3 or 4… It would probably be cleaner if they took out height and just assumed a fixed relationship to weight. Would remove one cheat. And ensure you don’t extrapolate those CdA = f(weight, height) too far off the source data. I have a hunch that’s also part of the problem.