Zwift unfairly favours lighter riders on flat courses (data included)

It is common knowledge that climbing favours lighter riders and flat riding favours bigger riders.
This is because lighter riders have a better power-to-weight ratio and bigger riders have better power-to-CdA (power to air resistance ratio).

However, in Zwift, it is all about power-to-weight.
Below is a plot from a Time Trial on Wattopia’s waistband. As you can see the speed of rider closely tracked their power to weight ratio while absolute watts had a very weak relationship.

Obviously, Zwift is overpenalizing weight rather than tracking the real world physics. I would like Zwift to be fair to everyone.

Zwift does have to make some assumptions on CdA given whatever height and weight the rider gives it. Would be interesting to know what those are.

ZwiftInsider did a few tests a long time back for both weight and height, but maybe there are better tests out there.

So what’s hard to tell on the graphs above is which dots represent which weights and heights. Might be fun to use the colors of the dots in these graphs to represent either weight or height in a color scale (From dark red to dark blue to represent light to heavy or something like that). Then you could go online on some biking speed calculators and use those same values to plot various power to speed curves based on those with the lines being the same color as the estimated weight (or height) to see how that would look.

1 Like

The fastest 16 or so riders are easy to pick out on each chart, although of course not from each other. But there’s really only three significant outliers in those ones, the three around 43mph and at or near 350w. All between 4 and 5 w/kg, but it looks like three of them are significantly heavier than the rest.

I’m not sure this shows a significant unfair advantage, but I’d be interested to see more. Weight isn’t inconsequential even on the flat. But I (bigger rider) have thought that CdA isn’t right relative to how I descend for example IRL. I don’t really think I’m as big of a sail as Zwift thinks I am (tall and skinnier).

1 Like

Which is exactly what it’s trying to do.
It’s not trying to match real world physics; it’s trying to strike a balance so more people can ride relative to each other.

Those graphs are fun to look at… but what’s the context? What are the values? Who is who? How tall are they? What are their actual weights?

You’re missing a lot of visible data allow anyone here to come up with any conclusions.

Zwift has to make a lot of assumptions either way, do speeds get a little wild when you get people outside of the bell curve of “normal” ? Yes, it does, which is when it’s easy to spot cheaters or folks who have abnormally low weight. I will agree on that point.
But folks within the average curve of 60-80kg, things seem fair to me; I get crushed on the flats, and dropped on the descents, and they for the most part keep up just fine on the inclines.

But once you start looking at people in the 50kg and below range? Things break REAL fast.

There has been enough bot testing done by zwiftinsider to show that heavier riders go faster at the same w/kg in all circumstances, assuming the same height and equipment. This lines up with the real world, and I don’t think it’s correct to say that light riders are given an unfair advantage. I think the underlying issue is that zwift unfairly favours short riders, many of whom are also light.
The CdA estimate is based on height and weight, as that’s the only data zwift has to use. No idea what actual formula is used, but the model seems to be very upright.

1 Like

I expect that when you make a “line of best fit” in excel you get a very high/trust worthy R2

Heavy rider may be
Are they carrying fat which don’t help
Are they caring muscle which can help
Photo of person would help

• I understand that Zwift has to make assumptions, my point is that those assumptions are miscalibrated.

• Since a lot of people want more context with the data, I can you link you the dataset through another channel (zwift forum doesn’t allow me to share links)

• I did some actual research on this, see my blog here:

• It is true that heavier riders do have to put out a lower watts/kg. But in reality, this effect is far greater. Firstly, because the bike makes up a larger proportion of the system weight in smaller riders. For flat courses, differences that I typically see on Zwiftpower are about 0.1 watts/kg for 10 kg of body weight difference (when finishing equally), and this is simply too low. Under the assumptions of zwift, the influence of power to weight on a time trial seems more similar to a real world Alphe D’huez climb than to an actual Time trial.

• I made a graph of wattage normalized to length:

[image]

Speed as a function of watts/cm

Maybe I just struggle at reading dot charts when I’m not the one making them and don’t recognize what the data is or how it’s being produced… if those unique points are all unique tests/riders, etc.

Have you also considered that maybe all of your data points are from different in-game equipment? ie: are all of those dots based on the exact same bike and wheelset in game?

As said; we already know that Zwift biases more towards w/kg than raw watts, but raw watts still wins regardless. It’s never going to be accurate to real world, and is in fact designed to not be like the real world.

If you want real physics, there are other choices out there that attempt to offer such.

But last I checked, all of my heavier teammates can continue to crush me on the flats, and still can outclimb me… the lightest person on my team. The only reason why I end up with the most points is I’m the only one with a sprint and have a reasonable Vo2. All of which lines up to what the Insider testing proved from a while back which has been mentioned several times now.

Your scatter charts are interesting, but they are still failing at telling any meaningful story without knowing all of the details.

It’s also worth noting that Insiders’ tests were with steady state bots.
You’re saying this data came from iTT tests by unique racers? What’s to say the speeds and watts are in the same location on the course? (Which is not flat by the way).

As someone mentioned above, it is not meant to replicate “real life”.
Last time I checked, Zwift is still a “game”.
In fact, if you look at the source of Zwift’s home page, the word “game” is in the code 39 times.
My workplace uses a proxy server, and zwift.com is blocked because the site is categorized as “games”.

Unfortunately I cannot link here, but each dot respresents an average speed for a rider on a Time Trial in zwift racing league. Materials are different but will produce negligible differences in this chart.

Also, there’s difference in draft, timing, normalized power, and so on. Even with all these factors you get this very tight relationship between w/kg and speed.

I understand it needs more context but these provide better and more objective evidence than your personal anecdotes.

Wattopia’s waistband only has 95 meters of elevation over 25.5 km. If we can’t agree that’s flat, then - with all due respect- there’s no point in discussing this.

Again, the funny thing is. When I receive a dataset like that for my work. I’m very happy.

This is a completely predictable linear line. Again ask excel to plot the line and show us the R2.

We have a discussion about an almost perfect dataset

1 Like

So would I like it to be fair for all but the pace dynamics are bad for lighter riders who have to do much higher w/kg just to get through near the front of a robopacer group.

While a bigger rider just stands on the pedals with 400w and rockets through to the front (quoting a post elsewhere).

But that is expected on flat correct? On flat it’s known that watts (and CdA) should matter more than w/kg.

1 Like

We bigger riders can do that, but we sacrifice by not being able to ‘dance on the pedals’

i’m in no real position to express an opinion one way or another on topics like this, but light and heavy people perceive things entirely differently, when it comes to sports. at least from my own conversations with guys who more or less share my attributes at different sizes

I’m still trying to figure out how, as the title suggests,… lighter riders have an advantage on flat ground? Hills yes, flat no.

W/kg has no meaning on a flat road. On a flat road it is all about raw power.

1 Like

Indeed Gerrie, you get the point.