[Closed] Zwift Physics

They don't weigh 45kg either.

Haha

'your mate' got pwned in a sprint by a 10 year old 😉

Nah. Apparently weight doping in Zwift is very common. GOK why.

But the 200W power is verified, isn't it? And if Zwift has recorded this person as having 200W when they had more, then they would be under-performing on Zwift not over performing.

Surely it should not matter what they weigh on the flat?

Some geeky analysis here: Loads of info here - and other pages on this site.

May explain what's going on - or confirm someone not being totally honest.

Edit - though their FTP of 200W doesn't mean they couldn't have a great sprint and were putting out 1500W for a few seconds.

Most zwifters interested in racing use zwiftpower for results - a third party website that filters out most of the casuals / cheaters / out of cat riders. Your friend should look at that if he is interested.
V light riders are handled a bit differently, don't recall how exactly, but I think in this case of obv bollox they'd just get binned out of the results.

It's all W/kg, so 300/73 = 4.1 w/kg, 200/45 = 4.4 w/kg, hence the lighter (sandbagging) rider will be faster.

He is annoyed and thinks the physics model is rubbish.

FTP has [I]some[/I] relevance to how much power someone MIGHT put out for any given sprint, but there are obviously many other factors.

The physics model is actually really good. Loads of articles about exactly how it works.

Surely it should not matter what they weigh on the flat?

Is your friend upset about being beaten on the whole race or just the sprint?

Also a 45kg rider should be tiny, so will be much more aero than your friend. Zwift accounts for this so heavier and taller riders are slower than lighter shorter ones for the same relative effort

If you race smart and can read the road it's quite easy to sit in the pack and expend little energy compared to the group .

The same way it is in real life.

Up hills it's watts per kg on the flat it's all about watts. As people have said FTP is more about sustained power whereas a sprint is just putting out maximum power for 20 seconds.

It’s all W/kg, so 300/73 = 4.1 w/kg, 200/45 = 4.4 w/kg, hence the lighter (sandbagging) rider will be faster.

Only on climbs.

If you race smart and can read the road it’s quite easy to sit in the pack and expend little energy compared to the group .

This is plausible. Mate is a very experienced triathlete and tester (and very good) but he's not a road racer.

Surely the ftp is irrelevant in a sprint? He can check on strava to see how much the guy put out in the final sprint probably.

Apparently weight doping in Zwift is very common.

The whole idea of defining categories by w/kg totally ruins races for me. If you have a low ftp but high w/kg (ie are light) you compete in the higher cats and stand absolutely no chance

I'm thinking of saying I'm far heavier than I am so I can compete in the lower cats without getting disqualified. It won't enable me to win as folks in lower cats still have higher outright power, but i can at least be competitive.

Case in point, I competed in a cat c race the other day, finished the race at over 4 watts per kilo, with average of around 277 watts. Everyone that beat me put out far higher watts average, yet only one had a watts per kilo over 3.5.

Aren't there hillier races that you can choose to favour your skills then?

The whole idea of defining categories by w/kg totally ruins races for me. If you have a low ftp but high w/kg (ie are light) you compete in the higher cats and stand absolutely no chance

I’m thinking of saying I’m far heavier than I am so I can compete in the lower cats without getting disqualified. It won’t enable me to win as folks in lower cats still have higher outright power, but i can at least be competitive.

Case in point, I competed in a cat c race the other day, finished the race at over 4 watts per kilo, with average of around 277 watts. Everyone that beat me put out far higher watts average, yet only one had a watts per kilo over 3.5.

just like real life then ? light low FTP riders get humped on the flat and win in the hills.... How ever in zwift you must look for the hilly races as most folk like to race on the flat as they find it easier.

This is plausible. Mate is a very experienced triathlete and tester (and very good) but he’s not a road racer.

saw this the other week on club event at critcity

guy who has been club and scottish TT champ tried to power off the front of a double draft race with a lap to go .....all we had to do was keep in his draft to the end and then apply the afterburners.......even i beat him.... hes a considerably stronger rider but didnt race smart.

45Kg is conveniently the minimum weight you can input, so the other guy is probably cheating. As others have pointed out, FTP is not peak power, what really matters in a sprint. And they may have been a better drafter hence saving energy for the finale. Zwift is a great mirror on society, give people a chance to cheat on a faceless crime and they will fill their boots.

just like real life then

Yep..and I have no issue with that. The issue is that you don't define real life race cats by w/kg do you, so in real life if you don't have massive power you can compete in a lower cat. Picking that option in zwift gets you disqualified, or at least disqualified in the zwift power results.

Cats by ftp would be a far better way of defining race results. Or even better, some kind of points system.

Do we know which race it was? we all wanna have a nosey at the results on Zwift Power!

So apparently he drafted this person all the way round. She has a real Strava profile and is a runner apparently with very little cycling on the board. She was putting out 220W against his 307W.

He is using a dumb turbo and a power meter though - that might have something to do with it.

EDIT without a speed sensor

Your mate has roughly the same numbers as me, and I've done a fair bit of zwift racing, if that guy genuinely has an ftp of 200w. There's no way he should still have been in the bunch at the end of the race, regardless of what he put as his weight, your mate and the rest of the bunch weren't pushing hard enough for the rest of the race. Been in a few races like that, I hate them, as like your mate I'm never going to win in a sprint against heavier riders either in real life or in Zwift, it's up to you to play to your strengths and ensure either you finish in a select group in which you can win, or that everyone has had the sprint knocked out of them in the previous hour 🙂

So apparently he drafted this person all the way round. She has a real Strava profile and is a runner apparently with very little cycling on the board. She was putting out 220W against his 307W.

Something very wrong there, 220w should have seen them just about able to draft a 307w rider, not the other way around.

He is using a dumb turbo and a power meter though

power is power.

so long as its not Z power - which is not power just an estimate of power.

Yep..and I have no issue with that. The issue is that you don’t define real life race cats by w/kg do you, so in real life if you don’t have massive power you can compete in a lower cat. Picking that option in zwift gets you disqualified, or at least disqualified in the zwift power results.

Cats by ftp would be a far better way of defining race results. Or even better, some kind of points system.

no you define them by race placings.... which is great when you A. Have a limited number of races to be done , B. A fixed calander and C a limited number of potential entrants. Much easier system to fiddle as well .... if i want to race but only be a cat 4 (if i was mental) just sit in the pack or even control on the front then sit up at the end.

For comparison this is last nights ZHR, the three in the image are all cat C, that's me in the middle.

There's a big difference between power but relatively little time.

I'll let others decide if volcano climb is a flat or hilly route

Need the link on ZP for the race results if either of them are on it.
If it's the Zwift results on their app then it's pretty useless.

Volcano climb is flat. Climb not steep enough or long enough to unship C cat races with extra weight like me compared to 70kg riders. (within Cat limits) My extra power keeps me in.

BTW the 3 of you are cheats if you were in cat C ;0)
Edit Ah I see CatD is CatC now. Makes sense...

in a 41 min race ?

Yes. I may not have won the races I've done over that but it was only down to the sprint.
A Cat might be a different story but drafting is in good effect on a lot of that climb.
I've lost the wheel over the top before now and been severly punished on the downhill though.

in a 41min race

Well I wanted sub 40 but couldn't manage it.

BTW the 3 of you are cheats if you were in cat C ;0)
Edit Ah I see CatD is CatC now. Makes sense…

Ah, yes, they had us all down a category, so I race B but was in C for whatever reason.

but drafting is in good effect on a lot of that climb.

Not with no draft its not 😢

If FTP were the sole determinant of race results, virtual or real, you wouldn't actually need to have a race. You could just roll up, throw down a card with your massive wattage on it Top Trumps style, and then go to the pub (well, you could have done). It would work in real life too, Rohan Dennis would just be awarded everything - as happens on Zwift anyway 🙂

throw down a card with your massive wattage on it Top Trumps style, and then go to the pub

That sounds like a ruddy great plan.

Once the pubs are open.

Rohan Dennis would just be awarded everything – as happens on Zwift anyway 🙂

Was watching a funny comment war on cycling weekly Facebook page about that how the top zwift riders would have pumped Rohan.....

Then when you looked at FTP and w/kg.....they wouldn't even have been in contact at the line.... he was going 5.9w/kg for an hour.

He is 73kg and has an FTP of about 300W. He stayed with the bunch in his race, however he got beaten in the final sprint by someone with an ftp of 200W who weighed 45kg. Surely this should not be possible on the flat?

Correct

He is annoyed and thinks the physics model is rubbish.

Did your mate check what the winner's power output on Strava was in the sprint? Zwift's physics is pretty good so they would have have had to out power him on a sprint.

With a weight of 45kg and a fantastic sprint, chances are they were cheating - either keeping their average power down in the race which is quite easy with an unrealistically low weight and good drafting, then using their real power just on the sprint. It's also possible they went for an unrealistically low height, which reduces drag on drafting and sprinting. Or they are a Japanese lady pro cyclist who just likes to bimble along at 200W most of the time.

Why do people do this - GOK.

I had it recently in a Crit... But then i felt better as his description was he was U14.... So may well have been 45kg after all.

He found her on Strava, she's a runner and not a very regular cyclist by the looks of it - or doesn't record often. Pretty handy as a runner but even then not as fast as my mate who is handier still.

I think you should just link your zwift power so we can see the race in question.

Sorry I mean their

Is the ftp a number that the user enters? Or do you have to do the test in zwift, which presumably you could soft pedal though

You could soft pedal the test but it will update it as you complete rides so to keep it artificially low you'd have to b soft pedal for ever, which probably defeats the point somewhat

Looks like there is a market for a turbo with built in scales, I'm off to the shed to get inventing.

Looks like there is a market for a turbo with built in scales, I’m off to the shed to get inventing.

A splash screen on load saying "it's a ******* game, if you want real world physics go outside, or play tiddlywinks" would be cheaper and easier

Games have rules otherwise they are pointless. This is not an unreasonable question at all.

Of course, but they're not the same as real life, I mean, when did you actually get to buy a hotel in Mayfair for a few hundred quid?

Looks like there is a market for a turbo with built in scales, I’m off to the shed to get inventing.

but actually yeah... integrating a load cell sensor with a wifi connection into one of these rocker plates that are all the rage nowadays probably wouldn't be that hard... it would use peoples' proper "all up" weight as well (inc. kit etc) rather than their theoretical minimum weight (i.e. dehydrated when you've just got up, after you've had a big poo, like how I weigh myself 😂 )

Is the ftp a number that the user enters? Or do you have to do the test in zwift, which presumably you could soft pedal though

Zwiftpower members can manually enter and adjust the FTP in their profile, but your category is based on the average W/Kg of 95% of your best three 20min efforts (just races and TTs, not group rides?) in the last 90 days.

You can adjust your weight via your profile at Zwift.com and then click "refresh profile" to sync that stat to Zwiftpower, no limits, but changing your figures wildly and rapidly while racing and TTing will often get get reported on https://zwiftpower.com/viewtopic.php?f=11&t=5612 and possibly result in sanctions.

So taking my own modest profile figures https://zwiftpower.com/profile.php?z=588304

I set FTP to 280W after the Newbury TT on 6th May, after a best 20mins of 295W (using my 4iiii for power) https://zwiftpower.com/events.php?zid=704776 , which is now 5 weeks ago, so I need to try and do another effort in the next week to see where I am now. 3.59W/Kg 95% figure when I was ~77.8Kg that day.

But my best 3 efforts in the last 90s days are from when I was lighter (as low as ~76.3Kg) and a bit stronger (~294W in early March and been steadily declining until that Newbury TT effort), so Zwiftpower is working on 280W and 3.61W/kg 95% figures...

Even though I was ~79.5Kg when I weighed myself a few days ago, but that still puts me very safely in just below mid Cat B category at ~3.5W/Kg.

--------------------------

In other news, been trying a few of the new "Zwift Racing" workout plan events recently, some have been quite fun... But don't think if you run a race sim workout at the venue of that sim, the intervals will sync with where you are on the track, I tried the New York Grand Central at the venue and it was mad doing my biggest efforts on the KOM descent! 😆

Crit City sim was quite fun, lots of very short intervals, not so great for getting ERG to power match (it does recvommend turning ERG off) but fun nonetheless.
https://whatsonzwift.com/workouts/zwift-racing/

integrating a load cell sensor with a wifi connection into one of these rocker plates that are all the rage nowadays probably wouldn’t be that hard…

The thing there would be your bike weight would matter, your wheel base would effect it and so on, you'd just fiddle the calibration if you were bothered enough.

You've got to assume everyone plays by roughly the same rules (like PEDs and so on) but that's a very different issue to the physics of the thing.

The issue in the OP isn't (solely) one of weight doping, as its "flat" it's really how does 200w beat 300w, and that's down to not understanding the way zwift works rather than specifically any cheating. There's no moaning in the OP about being able to take the bends on the alpe for instance at 70mph on the way down. The physics are the same for everyone.

The thing there would be your bike weight would matter, your wheel base would effect it and so on, you’d just fiddle the calibration if you were bothered enough.

I would assume anyone willing to go to those lengths would be obsessed with super-accuracy and definitely NOT cheating

You've not done much online gaming have you?

The physics do overly favour light riders on the flat.

Although it might be the other way about in that it over penalises heavier riders - you see it in the TT races on tempus fugit (flat, everyone on a tt bike, no drafting) there are riders on similar time with a huge spread in power. The last TT i did there was someone who did over 460w and he finished a few seconds ahead of someone who was under 300. Both took about 23 minutes...even if someone has a poor aero position in real life on their TT bike then 460w will be significantly quicker than 23 minutes for a 10 mile TT.

To be fair the 460w is probably a miscalibrated turbo anyway, but just using that as an example.

You’ve not done much online gaming have you?

Does IDKFA work in Zwift? What about IDDQD?

460w will be significantly quicker than 23 minutes for a 10 mile TT.

Quickly dumping the numbers in to bikecalculator.com (which tends to be about right for me in terms of on screen speed to power)

@5w/kg
That 460w rider weighs 92kg and will do 42.7kph on an absolutely pancake flat route so should cover 10.23 miles in 23 minutes.

The 300w rider at 60kg does 37.2kph on the same (zwift puts me at 36kph at 300W:80kg on the standard bike)

On that same 10.23 miles they'd be at 26.4 minutes so over 3minutes slower.

Zwift adjusts areo for weight and height.

According to ZI:
There's 30seconds difference between the zwift TT bike and the cervelo p5x on tempus fugit at 300W
53sec between the standard wheels and the zwift 858 super 9s.

So that's almost half the time difference under an identical rider and the speed increases aren't entirely cumulative, wheels & frame upgrade is better than simply wheels + frame upgrade.

Also https://zwiftinsider.com/how-height-affects-speed-in-zwift/

"a 5′ tall rider will be ~2 minutes faster than a 6′ tall rider for every hour of riding on a fairly flat course"

Weight on flat routes (2 laps tempus fugit) makes a big difference
https://zwiftinsider.com/rider-weight-speed/

82kg @ 300 watts: 52 minutes, 17 seconds (39.7kph)
75kg @ 300 watts: 51 minutes, 26 seconds (40.4kph) – 51 seconds faster

(that's 3.6 seconds a kilo on a single lap)

So...

On the same 10.23mi,with the best bike and wheels vs standard setup the 300W rider, assuming 1' height difference would be 75 seconds slower, not allowing for cumulative effects of kit, any gradients, or any difference in rider weight.

If the weight relationship we're linear (I doubt it is) the 60kg vs 92kg rider would be 115sec faster for weight, so 40 seconds quicker over all with best kit.