Right now my Zwift weight is 5 - 10 lbs heavier than my actual weight. I’ve been dieting and also lost some weight from Covid. I’m actually being very conservative with my weight for a couple reasons. 1 being I absolutely don’t want to cheat (owing to the pandemic I can’t actually access a scale right now). And 2 being that I want to “cash in” on my weight lost when my training progess plateaus… kind of for the mental boost. I’m curious to see how much better I do in the racing once I figure that out.
As far as the talk about the “weight advantage”, I noticed that before the new year, I was noticeably faster on the downhills. That advantage is all but gone now it seems. Have they messed around with the height physics or something?
yes, it’s one of the caveats of being bigger… as a quick example a 65-70kg guy can expect to burn about 800-900 more calories than me (55kg) in a 2 and a half hour race, assuming we finish in the same group. power meters are reasonably reliable for calorie calculations as kilojoules can be converted into calories burnt rather easily.
just thinking about it logistically, eating 900 calories in 2 and a half hours while cycling sounds like a nightmare at the best of times to me
i mean, there are things bigger riders can generally do that lighter ones can’t, also. i can race an A1 zrl points race and mop up a bunch of FTS and FAL points but i can’t contribute to an A1 TTT. similarly, my irl 10 mile TT times are ■■■■, even at 4.5wkg+.
But making myself heavier in Zwift slowed me down similarly to real life… I must just be weird. Everybody in her makes out that if I get heavier, I get quicker!
Hi Tim. If you could let me know how you increased your in real life muscle mass in proportion to the amount you increased your weight when you did your ‘weight test’ in Zwift can you let me know? I can’t be the only one that would like to be able to do this and you sound like you’ve mastered it by how confident in your test you are…
Would your cars top speed be higher or lower if you had a couple of extra passengers in it?
Would your cars top speed be higher or lower if you put a bigger engine in it?
adding weight is very dependant and how and where you add weight.
Most race routes are flattish in Zwift. I was very successful cruising Road To Sky and typically won by several minutes, but the main reason for the success was the lack of competition - from other cruisers! A heavy cruiser wouldn’t be disadvantaged against me up AdZ. It’s just that I wouldn’t end up in a situation where I couldn’t follow him without risking an upgrade. It would be a competition about who would cruise more efficiently, while leaving the legits way down in the ascent. On a flatter course that heavy cruiser (or just a plain heavy guy) would be untouchable, like in the race I wrote about.
The important point here is that this is not just natural differences like the ones in RL racing. A light guy has a hard time against superior muscle volume on flat course outdoors too. But the difference isn’t as extreme and above all the light guy isn’t forbidden to compete. The point is that it is the system that creates this extreme and unsurmountable inequality, not DNA or Newtonian physics.
That you found that the winners in that race moved up to cat C two years later I don’t find that interesting. They were cat C’s back then too. And a long time has passed, a lot of time to improve further and make racing in cat C more interesting (by not having to be pack fodder anymore).
There’s also another aspect. I remember even further back when I published this post. At the time the WTRL TTT’s with their mixed teams were new and there had been a lot of cat specific TTT’s. Although I never participated in any of them myself, I was provoked by winning teams in cat D with entire teams where all members weighed exactly 100 kg. Maybe it was an accepted convention to adjust team members weights upwards like that, but it was still exploiting. But once this behavior was in the spotlight it quickly disappeared, and it wasn’t just because of WTRL. Many of those guys apparently went straight into a “diet”, lost several kg’s quickly and also accepted the upgrade to the bottom of cat C, where they have stayed since. I could give you names, but there is only really one name that matters here: Zwift Inc. That’s the real badguy.
You mustn’t forget that back in 2020 when I wrote those articles, cruising was not an accepted concept. The vast majority of racers were completely unaware of the phenomenon, wouldn’t believe that Zwift had faults and argued against me, just like you do today, every time I tried to draw attention to the phenomenon. Which was the reason I started to cruise myself - just to prove it to them that cruising was possible, real and even the optimal way to race in Zwift (still is if you play it right). Once the realization crept up on people, it was no longer as much fun to cruise. I think many of the cruisers back then didn’t see themselves as cruiser or wanted to. Cognitive dissonance. What to do? Well, either you keep underperforming for the win, knowing you are now the badguy. Or you accept the upgrade.
Weight is trickier though. If you weigh 106 kg there is probably (although not necessarily) a little overweight in the picture. And it’s HARD WORK to get rid of if you try. It’s a struggle and there are no evidence based weight loss methods (other than, to some extent, gastric bypass) with positive long-term results. And what are you going to do while fighting those extra kg’s, if you decide to put up a fight at all? Avoid racing in Zwift because you are advantaged? Of course not, and you shouldn’t have to. But races should still be fair.
You are right, but you are partly speaking from an outdoors perspective. Yes, big boned guys will be faster on the flat outdoors too and they should be. And lighter riders need to at least try compensate by adding leg days to their training regimen (what makes you think they don’t already?)
But you have to separate that from the artifacts that the idiotic W/kg cats create, which make it logically impossible to compete against a substantially heavier rider. I am not an enemy of Newtonian physics or the human genome. I am an enemy of performance based cats. They are an abomination and shouldn’t be allowed to exist because they kill the spirit of sports.
You are still missing the key point, Tim.
Yes, obviously, if you double your weight (in Zwift settings or in real life) you will be a lot slower uphill. You will also have a harder time accelerating up to a good speed to keep on the flat. If we are talking RL kg’s, then with time your muscle volume will increase a little. It has to or you won’t be able to walk. Your body has to adapt to the increased pressure. But it’s not going to compensate for a massive weight gain like that. So no, you are not going to win more races by putting on weight. That is not the point.
The point is that every cat B-D is a span. you have guys at the bottom of the cat. They can be light, they can be heavy, but they are never going to win a race. Then there are guys at the top. If we consider the old categories, which were more transparent (just for simplicity/sake of argument), in e.g. cat C you have a performance ceiling at 3.2 W/kg. You can’t go above it or you’ll get DQ’d or upgraded or both. There is no rule saying it’s forbidden to cruise a race but the ethics of these cats is they are only supposed to contain guys who, at most, can do e.g. almost 3.2 W/kg in a 1 hr race. If they are any stronger, then they should race in cat B instead.
Now, up at the top of cat C there will be riders who can do 3.19 W/kg. Races under a performance based system will always tend to be hard efforts, hard as in close to 3.2 W/kg. Why? Because it’s how you win. If you can do 3.19 and the next two guys can only do 3.10 but have a stronger sprint, then your only way to win is to push them to their limit and beyond well before the finish. You need to make use of that 0.09 W/kg advantage you have and make it count. So yes, most races in cat C will show a podium where people were pretty close to the limit of 3.2 W/kg.
Some of the riders at the top of cat C will be light, others will be heavy. There will be both kinds of riders, and all at the top can get close to 3.20 W/kg. So one rider at 100 kg is nearly able to do 3.20 W/kg, right? That means he averages 320W, quite a lot. A guy at 70 kg near 3.20 W/kg is only doing 224W tops (70 x 3.2). So on the flat he can’t keep up with the heavy guy without going over limit and get upgraded - by the logics/ethics of the system he then would no longer belong to cat C. So he is not allowed to compete with the heavy guy.
Uphill both riders are doing nearly 3.20 W/kg because they are top guys in cat C. They are not competing against the losers at the bottom of the cat, they are competing against each other. And both keeping 3.20 W/kg means they are both climbing at basically the same speed (the heavy guy has a slight advantage because of the bike dead weight but it’s a small advantage). No one is advantaged or disadvantaged. But in an iTT on Road to Sky the heavy guy will win easily. The climb they both do at the same speed but the heavy guy can go faster in the approach to the climb. If the light guy tries to keep up, he will get upgraded.
That is the point. We are only talking about racing at the upper end of the cat, right beneath this artificial performance ceiling that you are not allowed to go through. It’s only those guys up there that are in position to win anyway. And among those guys, the heavies are advantaged. And they are already there, right beneath the ceiling. If any of them would double their weight they would no longer be and they couldn’t win. But they didn’t.
i’ve never felt disadvantaged by my weight. quite the opposite honestly. i will personally argue in favour of lighter riders being given a small amount of leeway on zwift for the most part, which they historically have been, because the fact that most don’t do S+C only amplifies the difference between a fit heavy guy and a fit light guy, but it’s not the case when you have someone who is both light and strong
these guys agree with me (thank you for the link dave. lol)
Yes in the sense that absolute strength (raw Watt) means a whole lot in both RL cycling and Zwift, but even so Zwift seems to have backed off from the initial intention to lean the new model heavier towards Watt rather than W/kg.
No in the sense that the actual weight component matters too because it is the divisor in the expression W/kg. It is what allows you to keep a higher Watt in a lower category, regardless of whether you can or not, whether it is easy for you or not.
Remember, for some people with higher than average weight, the people in the top of a cat, it will be easy or at least manageable to keep Watts that put them close the the W/kg limit. It is those guys who have an unfair advantage against lighter riders.
What happens mid-cat is not interesting or particularly problematic. Those guys, light or heavy, won’t win anyway and they are not bumping against the artificial performance ceiling. It is when you are close to hitting the ceiling that the problem starts. So the problem isn’t your weight. It is the ceiling. There shouldn’t be a W/kg ceiling. W/kg is the wrong way to divide cats. It should be something else - past results.
You are not being penalized (quite the contrary) and I’m not suggesting that you should be. I just don’t think that you should be given an artificial advantage on top of the natural advantage of being fast on the flat. So how do you solve that? Multiplying a heavier riders race time by 0.9? Nooo, that would be penalizing. You solve it by killing the W/kg thing, the performance based categories, and by a move to results based categorization.
I think you should reread my first reply to you slowly and carefully. I can get the impression that you don’t want to understand and that you try very hard not to. It was the same last time a few months ago.
Saw in another thread that you considered yourself a borderline case. And then looking at you on ZP, oh my, aren’t you a bad example?
You seem to compensate your very low weight, relatively speaking, with very high short-term W/kg, again relatively speaking. Your variability index is high by comparison. And you obviously accelerate much faster than others.
I also notice, when looking at a few races, that the variability index seems to have been generally higher overall in cat B races than what I usually see in C-D. Meaning the races are not the pedal to the metal from start to finish as is often the case in the lower categories. They run more TT like, although in group. Which is the same as squeezing against the ceiling throughout the race, problematic if you’re light. It’s probably influenced by the abundance of sandbaggers in the past and cruisers (they are more numerous the lower you go in cat since there are more potential riders like that above the racing cat in question. Oh, and then there’s the higher variance in weight in lower cats. You don’t really see people at 110 kg in cat B. The weight funnels, closer to the presumed optimal range up in cat A, so probably less conspicuous.
Anyway, in a nutshell, you’re not very representative. Most very light riders in B-D don’t sprint at 14 W/kg. And it’s not because they forgot leg day, just the same as most riders will never get to cat A and not for lack of commitment. DNA has a way of getting in the way for all those grand plans. And so they don’t win any sprint finishes, unlike you. And they shouldn’t, that’s not the problem. The problem is the W/kg ceilings.
There was actually a guy not unlike you in my sample in the OP above, a successful very light guy. I didn’t discard him as an outlier, I kept him. The results still showed an advantage for above average weight… We’re in the land of averages and tendencies, not the land of individuals and anecdotal evidence.
Andreas, I think your schtick is way over the top. I very much favour results-based categories, but your ranting makes me look like I’m with the crazy side. For a bunch of reasons (speed of convergence and avoiding the ‘glass-legging’ @S_A_Cestria_CC talked of, I think having both a power component and a results component to the categorization system is important.
Also
you have the TTT thing Bass Ackwards big time. The TTT’s had been using the mixed-cat classes for a long time in July 2020, same-cat TTT’s basically started with ZRL in September 2020.
“cruising” as you term it, is sandbagging: (from Mirium-Webster)
It wasn’t a unique and new concept you came up with in July 2020, you just decided that the existing name wasn’t good enough because people were also using sandbagging to refer to entering a lower category than they should.
You’ve got this wrong too. Watts required to go a certain speed in Zwift on the flat follow same form as IRL:
Watts required = A * weight * speed + (B+C * weight + D * height) * speed^3
Rolling resistance + aero resistance
I’ve used this extensively for TTT power target planning for teams that have very different sized riders. Has held up pretty well.
Using the coefficients I have: A 92kg riding at 300W will go the same speed on the flat as an 80kg rider producing 275W. Some of your base assumptions in your year-and-a-half long tantrum are incorrect.
In short, I think you’d have more fun if you put all this effort into something else.
How could it possibly be important to have a power component to categorization when exactly every other sport in the history of mankind has managed to avoid it? You’re stuck in the short tradition of Zwift and can’t let go because it scares you, but you need to broaden the horizons. There is a reason no other sport has ever used power metrics to divide participants in pens. And that reason is not the low availability of power meters in the lower classes in the 19th century, I promise.
It’s absolutely true that there were single cat TTT teams in spring 2020 and that some of them obviously faked their weights. Like I said, I could give you names but I won’t point fingers, especially not in the wrong direction. W/kg cats made them happen. Without them we would never have been in this mess to begin with, so the fault lies with Zwift.
I’ve been through this so many times before… The reason I didn’t want to group cruisers with sandbaggers but wanted a unique name for the cruisers was that people couldn’t distinguish between the two because they didn’t understand that cruisers existed. All they saw was cat B’s racing in cat D, something cat enforcement can now stop. Calling cruisers sandbaggers made people completely miss the point over and over. Smack them in the head with Merriam-Webster if you like, I don’t care.
Partly correct. I didn’t even come up with the term myself. I st0led it and have tried in vain to find the guy who first used the term because I wanted to credit him, but Zwift deletes threads in this forum so it’s long gone. It wasn’t discussed here or elsewhere back then though.
You can deny all you want that I drew attention to it back then by being just as noisy as today and that the phenomenon is now a rather well-known fact, at least among the “savvy” (whereas people vehemently tried to deny or argue against it back then). But it won’t change a thing.
That’s a weird formula you are using. Or is it just the formatting that was a little off? RL cycling is usually modelled with the formula I show in this blog post (which is about how to cruise the first crucial 20 min in a Zwift race, any intentional cruisers knows this).
All else equal, someone doing 300W on the flat is going faster than someone doing 275W. That’s a fact.
Then you add in example weights I didn’t use myself to try make a point, but in doing so you miss my point. My point is that in real Zwift races as a light rider you frequently run into situations where you can’t keep up with a heavier rider (both of you in draft for the sake of argument) because you break the performance ceiling doing so. This is very noticeable when cruising since you are actively monitoring your performance and actively avoiding the ceiling.
The scenario in the race report I linked above is a very common one when cruising. You gauge the opposition before the race and write down names to watch during the race. Names from ZP of other likely cruisers plus the guys with substantially higher weight than yours, guys who can still get close to the performance ceiling. The two types are usually your only real competitors.
Then at start it would be best for your trailing average to keep an even pace but you nevertheless have to get in the frontmost-likely-legit group (given that sandbagging is allowed), i.e. the frontmost group that has riders in it that ZP won’t DQ. You can never catch them later.
Then during the first 20 min you regularly face the tough decision whether to stick with that group or to drop in order to not get DQ’d/upgraded. You don’t know at that point whether the breakneck speed will be enough to DQ the heavy (someone identified as a threat pre-race) in that group, but you become more and more certain that sticking with the group is going to force you to go through the ceiling. So you reluctantly drop. Then if you’re really lucky that heavy will get DQ’d. But I have been in many races where he didn’t (although I would have had I stayed in the group). So he was untouchable. Get it? You are in a race against someone you are not allowed to compete against! You are not allowed to beat him! How insane is that?
If you had been racing legit against him, assuming you had the fitness to barely stay with his group, this is where you get a WKG.
Don’t try to deny this. This has been a reality in Zwift racing. The new model changes and obfuscates things a bit, but we still have performance ceilings (although somewhat blurred) in cat B-D. Light riders still have a disadvantage because of the ceiling. And you can still cruise, although it seems not everyone can anymore, you need to be a certain type, have a certain profile.
The problems will never go away unless we kill the performance based cat thinking. It should never have been in Zwift racing to begin with, and the time to get rid of it is NOW.
Yeah, wouldn’t we all? You think I’m in it because it’s fun somehow? I don’t think all the fact resistant people are in it for fun either.
Zwift racing sucks and someone had to step up and point it out in front of all the fanboi yeasayers. That’s all there is to it. Now even Zwift (Flint) admits as much, whether it’s just oil on the waves or more. Things move zloooowly around here. (Slow company or slow minds in the community, I don’t know, don’t care.) But they do move. We wouldn’t be where we are today, you wouldn’t have cat enforcement, if there weren’t for a handful of annoying b******* like me. You can write that down too as another fact. Yet another fact is that the new model tries to address issues I and a few others have pointed out in the past while still trying to save the Newtonian cats (can’t be done but… “A” for effort), another sign that discussions like this one do matter. In the long run. And I’m not going anywhere, sorry.
Adreas, I’m not going to try to refute everything - I don’t have time to go through the volume.
TTT Coffee classes were the same in spring 2020. You implied they changed, they’re the same as now.
The power required for the speed formula I formatted using the INPUTS THAT ZWIFT USES - weight and height - they use weight and height to calculate a frontal area, so lighter riders have an advantage on the flat in watts required.
Why use power -
First because it’s how the game works. The game physics runs directly from the power meters. You’d never use that IRL because they’re all different - the numbers from everyone’s powermeters won’t match their IRL performances - but they will match their Zwift performances, because the simulation takes the power number.
Second because the population size is so huge - Whatever results-based system you use, there are relatively few people who move between race times because of where people live in the world. It’s an easy leveling / correction factor. I raced a lot in different time zones, and when I race in the “corner” zones (Americas West and APAC to use the ZRL zone names) I’m often far higher ranked in the ZP ranking system than faster racers who race a lot - because I race in a “richer” pool sometimes.
If you would run these tests for the A-category, you would of course come to exactly the same conclusions (only the differences would probably be larger). So I am not sure what the point of all this is…
Out of curiosity, @Andreas_Traff , how frequently do you race? Can you share your zwiftpower profile? I’d like to understand where all this frustration is coming from.
This is quite the paradoxical statement. Unlike every other sport, Zwift can use a power component. They have far more data at their disposal than IRL sports.
IRL racing in the UK would be a lot better if it could use a power component. Instead the entry level of IRL racing is packed full of incredibly talented riders, to the extent that someone with Cat B Zwift metrics would be blown out the back within minutes. You simply can’t promote enough riders through the ranks based on results for the entry level racing to actually be entry-level. Other sports haven’t avoided it, they just take place in a completely different environment. You need to think outside the box a bit, which is ironic seeing as your rants seem to take place on cloud 9.
RL cycling definitely could. But they shouldn’t because it would be idiotic.
You ought to know well by now that I have been advocating power metrics for initial seeding of new and returning racers. But you cannot use that to create cats and you can’t use it to dictate in what cat someone should stay in after initial seeding.
A parable for people who still don’t get it:
There’s a fairly popular online computer game called DotA2, or Defense of the Ancients 2 (Dota). It doesn’t matter what it’s all about. Let’s just mention it has a pro scene and that it is the most well-paid e-sport of all. The top pros make the same kind of money as mid level pro golfers. So obviously there is money and sponsors and whatnot involved, so for Dota to stay an e-sport it has to be really sports like. Skill-based, fair, etc etc. From the grassroot level and up to the elite.
Dota is a 5-man team game. You play against other 5-man teams. It has a ranking system which governs who you get to play against if all you want is to just play a ranked game. To get into tournaments you need to meet certain criteria (such as e.g. a certain rank).
If you are a new player and want to play ranked games, you don’t have to start from the very bottom and work yourself up. You can get an initial seeding which, in most but not all cases, will be a decent approximation of your current skill level. This seeding is you get through 10 trial games with and against ranked players, where you are measured in several ways. The publisher (Valve, owner of e.g. Steam) used machine learning to come up with an algorithm that estimates your player skill and hence a reasonable start rank. On a group level I’m sure it’s good enough for its specific purpose.
What goes into this algorithm is a secret but it’s likely many things. One funny detail is known though. One factor is mouse click rate. (Compare it to the power measures Zwift uses in the new model.) They have found a positive correlation between how often you click the mouse button in the game and your success as a player. You really don’t win any games by just clicking the mouse (you need to know exactly where to click and when among other things), but it’s just one of those things. There is a positive correlation. So they use it for the seeding, to some success.
Then once you have played your trial games you get an initial start rank and the game will match you against similarly ranked opponents. The algorithm thought “OK, here’s a guy with such and such measures. It most closely corresponds to a player with the rank of 1650, so let him start there.” And if it got it right you will stay there early on. If it got it wrong, as it sometimes does, you will drop or rise early.
Now, here is where Zwift and Dota diverge. Zwift will keep these measures to separate players, dictate who gets to race who, who are in best position to win a race. Dota takes another approach - the standard one. Past the initial seeding ONLY game results matter, in an ELO style way. You win, you gain rank. You lose, you lose rank, adjusted for how highly ranked the opponents were. It works pretty well. And it’s fair.
Imagine Dota would instead keep mouse click rate as a key measure and divide players into pens based on that. ”Nooo! You are not allowed to play in this pen! You click the mouse too often! Cheater!” That would be confusing a mediator with outcome, if we use the language of scientific method. They are NOT the same thing. They are fundamentally different. It would be idiocy to mix them up.
The same kind of idiocy as keeping W/kg ceilings, soft or hard (post-race DQ’s or no). I can’t put it more nicely than that. Not at this point.
