In an earlier post, we dug into how rider weight affects speed on Zwift, specifically calculating how much each kilogram of weight slows (or speeds) a rider on Zwift.
Many readers asked us for similar data based on watts per kilo. That is, if we hold w/kg steady, how much does each kilogram of body weight (and the corresponding additional wattage required to maintain the same w/kg) slow or speed the rider on Zwift?
First we summarize the data, then we discuss conclusions. Enjoy!
[adinserter block=”7″]
First, a Disclaimer
Zwift is a rich simulation of outdoor riding in which many variables are taken into account to provide a realistic experience. What we’re trying to do here is show simplified test results to illustrate basic cycling physics which apply in Zwift and outside.
There are many, many variables which these tests do not include: varying rider heights, drafting, bike weight or aerodynamics, rider posture, etc. Our goal here is to focus on w/kg. But in practice, there is a lot more going on!
All of our tests were done using a single rider in isolation – so no draft effect. This rider was 183cm (6′) tall and used the Zwift Aero bike with 32mm Zwift wheels.
The Data
Climb Test
Climbing tests were completed up Alpe du Zwift, an 8.5% average gradient which is 12.2km long.
Power to weight ratio (w/kg) is a very good predictor of speed on steeper climbs. For example, we see here that two riders of wildly varying weights (50k to 100kg) are separated by only .6kph when they both hold 2w/kg.
- When holding 2w/kg: every kilogram of weight + 2 watts makes you ~0.012kph faster
- When holding 3w/kg, every kilogram of weight + 3 watts makes you ~0.016kph faster
- When holding 4w/kg, every kilogram of weight + 4 watts makes you ~0.022kph faster
- At 75kg, each additional w/kg increases your speed by ~3.55kph
Why is there a speed difference at all, and why is that difference greater at 4w/kg than 2w/kg? Read “Conclusions” below for further discussion.
Flat Test
Flat tests were completed over two laps of Zwift’s Tempus Fugit route, a 0% average gradient which is 34.6km long.
Speed varies significantly on flat roads between riders of different weights who are holding the same w/kg. For example, the speed difference between 50kg and 100kg riders both holding 2w/kg is 3.8kph. And like the climb test showed, this speed difference is even greater at higher w/kg: the difference between 50kg and 100kg riders at 4w/kg is 5.5kph.
Notice also how the speed gap between 2w/kg and 3w/kg is wider than the gap between 3w/kg and 4w/kg? Again–it’s just physics! As you speed up, each additional kilometer per hour requires more added power than the one before.
- When holding 2w/kg: every kilogram of weight + 2 watts makes you ~0.076kph faster
- When holding 3w/kg, every kilogram of weight + 3 watts makes you ~0.084kph faster
- When holding 4w/kg, every kilogram of weight + 4 watts makes you ~0.11kph faster
- At 75kg, each additional w/kg increases your speed by ~4.4kph
Descent Test
Descent tests were completed down Alpe du Zwift, an -8.5% average gradient which is 12.2km long.
Here we see the largest speed difference between rider weights, and this makes sense because gravity is now working in favor of the heavier rider.
- When holding 2w/kg: every kilogram of weight + 2 watts makes you 0.206kph faster
- When holding 3w/kg, every kilogram of weight + 3 watts makes you 0.21kph faster
- When holding 4w/kg, every kilogram of weight + 4 watts makes you 0.214kph faster
- At 75kg, each additional w/kg increases your speed by ~1.95kph
Conclusions
Our Zwift test results don’t tell us anything an experienced cyclist hasn’t already learned outdoors. Still, let’s discuss some of the findings and how they’re applicable to riding and racing on Zwift.
Don’t Attack on Descents
Look at the speed difference between each w/kg band on the climbs or flats versus the descent. Notice how the speed difference is much greater on climbs and flats. This (along with safety concerns outdoors) is why you don’t see riders putting in big digs to attack on descents: the speed you gain just isn’t worth it!
Example: on the Alpe descent, if a 75kg rider bumped up their wattage from 2w/kg to 4w/kg, their speed would increase by 3.9kph, or 5.7%. If they had put in the same effort on flat ground their speed would increase by 8.8kph, or 28.2%. More than double the difference! But get this: put in that same effort while climbing the Alpe and the rider’s speed increases by 7.7kph, or 92.2%. This is why racers attack on climbs–because you can put massive time gaps into the competition.
[adinserter block=”1″]
Heavy Wins
As the road gets steeper, riders at the same w/kg will see their speeds coming together. On a steeper climb like the Alpe, speeds stay quite close together, varying at the most in our tests by just 1.1kph (between 50kg and 100kg riders both at 4w/kg).
But why is it that when two riders are holding the same w/kg, the heavier rider will always be faster? There are multiple reasons, actually–but one big reason is what we’ll call “true w/kg”–that is, including the weight of the bike in our w/kg number.
Suppose we have two riders, 100kg and 50kg, both riding at 3w/kg. But let’s say they’re on 9kg bikes. If you add that bike weight to the rider’s weight and calculate the true w/kg, you get this:
- 100kg rider + 9kg bike @ 300 watts = 2.75w/kg
- 50kg rider + 9kg bike @150 watts = 2.54w/kg
So even though both riders are holding 3w/kg, the heavier rider is holding a higher true w/kg.
Another reason heavier riders go faster is that, unless you’re climbing straight up (which is impossible), your effort isn’t only lifting you up the hill–it’s also driving you forward by overcoming the forces of air and rolling resistance. Heavier riders are putting out more pure watts than lighter riders, meaning (in simple terms) there are more watts available to overcome air and rolling resistance after the lifting is done.
But Light Really Wins
But let’s not misinterpret the data. While the heavier rider always wins at the same w/kg, what we see in reality is that it’s really tough for a heavy rider to hold high w/kg. Making 400 watts of power is much more work for a 100kg rider than making 200 watts is for a 50kg rider.
So while our theoretical 100kg rider would be faster up the Alpe at 4w/kg than the 50kg rider at 4w/kg, there are very few 100kg riders who can hold 400 watts all the way up the Alpe! By comparison, there are many lightweight riders who can hold 4w/kg all the way to the top.
At the upper echelons of our sport, there are riders who can hold 6w/kg+ for over an hour. But those riders never weigh 100kg! The biggest are perhaps 80kg (Cancellara), but most are closer to 60-70kg (Froome, Dumoulin).
But Heavy Wins (Until the Climbing Starts)
On the other hand, if we look at Zwift’s standard w/kg-based race categories in light of the data above, it’s not hard to conclude that lighter riders face a significant disadvantage in flat races. Yes, you won’t find many heavy riders able to hold 5w/kg to win A races, but you will find plenty of heavier riders who can hold 3.1w/kg, meaning they could race as a C and really put the hurt on lighter riders. (In a flat race, a 100kg rider at 3w/kg will travel over 3kph faster than a 60kg rider holding 3w/kg.)
This is why many flat C races are won by racers weighing 90-100kg. Check out the category C results of this recent race, for example.
So what’s a lightweight D, C, or B racer to do? Skip the flat races. Enter events with significant climbs, where speed differences between weights are much smaller! If you’re able to hang with the heavier riders on the flats, then punch it up the climbs at a higher w/kg than the heavy riders can sustain, you can create the winning selection.
Longer-term, we’d like to see results-based categorization take the place of w/kg-based categories. But until that happens, lighter riders will need to race smart on the flats, attack on the climbs, and perhaps lobby for more climbing races.
Larger Variance at Higher Power
As watts per kilo increase, the speed difference between riders of varying weights also increases. So we see up the Alpe the 100kg rider is .6kph faster than the 50kg rider at 2w/kg, yet at 4w/kg the 100kg rider is 1.1kph faster. Almost twice the speed difference. Why is this?
Again, our “true w/kg” idea explains most of this difference:
- At 2w/kg, the difference in true w/kg between a 100kg and 50kg rider is 0.14w/kg
- 100kg rider + 9kg bike @ 200 watts = 1.83w/kg
- 50kg rider + 9kg bike @ 100 watts = 1.69w/kg
- At 4w/kg, the difference in true w/kg between a 100kg and 50kg rider is 0.28w/kg
- 100kg rider + 9kg bike @ 400 watts = 3.67w/kg
- 50kg rider + 9kg bike @ 200 watts = 3.39w/kg
Your Thoughts
Got questions or comments? Share them below!
I weigh at times between 190 and 210 lbs. Climbing is an albatross. Not only am I heavy, an unseen factor is my 32″ weightlifter thighs consisting of energy gulping, fast fatiguing, fast-twitch muscle fibers. My climbing has to be deliberate, controlled and steady. It’s not the speed factor that controls pace, it control of fatigue that controls speed. Just did the Apple at 2.0 work, 1.5 rest, steady pacing. Thanks for your info.
I normally love your stuff, but I think there would have been better choices of route than Alpe de Zwift as it contains so many flat sections that it will skew the data. You really need to rerun this but on a much more consistent climb without all the regular flat sections. I don’t know which hill would be best, as my Zwift knowledge is nowhere nears yours!
I dont think any hill on zwift will have a consistent grade all the way. I guess the alpe taking so long does help to iron out any anomalies.
“So many flat sections”, you say? Hmmmm… https://www.strava.com/segments/17267489
Show me a more consistent climb on Zwift, and I’ll eat my bike shoes. 🙂
Fair enough, I stand corrected and apologise. I’ve always felt like it had quite a few flattish sections. Maybe they were so well needed when they arrived that they were very noticable!!!
Think I must have missed all of the flat sections on Alpe du Zwift…….did I take a wrong turn?
I think you might be confusing the climbs. The two big ones on Zwift are the Epic KOM and the Alpe du Zwift (which is a copy of Alpe du Huez). The Epic KOM has many flat sections while the AdZ does not and is a consistent climb.
I am posting while climbing Alpe du Zwift and can assure you the flat stretches are extremely short and far between..
There are some “flatish” sections on the Alpe; however, they are short and primarily going around the switchbacks. Other than that, some occasional grade drops to 4-5% can FEEL flat when you are otherwise climbing at 9-10%!
Hi Sheridon,
The test was over 12.2kms at an average gradient of 8.5%. Whether that is made up of 18% pinches and some descents isn’t really relevant (although, in reality The Alpe is a consistent climb with not too much variation). Perhaps after a 10% segment you are finding the 6-7% gradients feel relatively flat? Reminds me of climbing the Stelvio, when the gradient flattened to around 4-5% I was like, wow this is awesome, some respite but then I’d look at my Garmin and realise I was actually still climbing.
Oh to be climbing the Stelvio again..
Suppose you can ride 40 kph on the flats. Wind resistance is proportional to speed cubed. So if you’re climbing at the same watts, but going half the speed, that’s around 7/8 of the power going to climbing, 1/8 to wind (there’s also rolling resistance, but neglect that for a moment). Now the climb gets steeper, and you can go only 1/3 has fast. So now 26/27 of the power goes to climbing, 1/27 to wind resistance. In each case wind resistance is small, and climbing is most of the power. So until your speed gets to well over half of the speed you can hold on the flat, grade fluctuations won’t affect conclusions much. This crossover is around a 5% grade: 20 kph @ 5% = a VAM of 1000 m/hour. L’Alpe only very rarely drops below 5%.
Good article, thanks! I guess this also shows that if you get dropped on the summit of a climb then digging deep to get back to the bunch on the descent is even harder!
Good point!
all depends on your descending skills as most descents aren’t straight. Or on zwift which wouldn’t matter 😂
Nice article. I would really like to see results-based categorisation or at least something better than what they have now. I did the 2020 Tour of Watopia as my first foray into racing. It was a lot of fun and I now plan to do more. I’m a newbie so I’m kicking it with the guys in Cat D. Every stage I’ve entered so far has been won by guys doing 4-4.5w/kg. In one stage, the D winning time was 3 minutes slower than the A winning time. I’m still overweight and underpowered, so it doesn’t really bother me now, but I would like to think as I improve that one day I might be able to fight for a win. You should never say never, but I highly highly highly doubt I will ever get to anything close to 4w/kg. People should genuinely select the appropriate category, but given they don’t, there should be a better system that does it for them.
And please, different colour jerseys for different categories so we know who we are racing.
Hi Bruce, did you check the Zwift Power results? I used to get upset with people pushing unrealistically high watts in races. But now I try find a realistically strong bunch to ride with and then check the results in Zwift Power after the race. The fliers in the race usually don’t have any results in Zwift Power, so I take that as my real placing, rather than the placing in the Zwift Companion app.
Thanks, this is very helpful for folks like me who are new to this and trying to work out how best to apply their limited w/kg!
With respect to your campaign for better race categorization I’d like to understand how that would affect those of us near the bottom. Being near the bottom end of the C group (in your Forever Losing category) I rarely do well enough in a race to pick up points. But I usually find some riders around my w/kg level and have a good race. If you categorize based on race results does that mean I get lumped in a category with the bottom end of B, C & D who all have no race points but w/kg all over the map?
Results-based categorization, if done right, should make categories feel “tighter”–more competitive. There will always be riders at the bottom (and top), but results-based would allow those at the very bottom to eventually race a lower category, and those at the top would get upgraded. No more of this “stay at 4w/kg forever and keep destroying the B cat” stuff!
For me, as a 70kg rider who can typically push out 3.3 to 3.4W/Kg for an hour or so this is just depressing, although it confirms what I have been thinking for a while. Most Zwift races tend to be shorter (sub 1 hour), on flat courses, and a rider like myself at the lower end of cat B stands no chance, especially with the tenancy for starts to be maximal efforts by heavier, more powerful riders. Recently I have been riding Cat C races, knowing full well I’ll be DQ’d by Zwift Power, but at least having the ‘enjoyment’ of racing in a peloton.
IRL I know that if I can ‘hang on’ for the first hour (IRL races tend to be longer), the pendulum swings in my favour, as I know I have the stamina. I’m beginning to think that I should be looking elsewhere for my virtual bike racing. I have considered ‘Reverse weight doping’ as suggested in a recent Zwift Insider post, but this just seems kind-of wrong to me!
Quite why Zwift is hung up on the ‘Watts per Kilo’ thing is a mystery to me!
I am in a similar situation to you. I have just been upgraded from C to B on zwift power, as my 20 minutes w/kg is 3.25. I weigh 71kg. I was mediocre in C races…..which are usually won by much heavier riders with a lower w/kg. Now I’m way off the pace in B races.
The situation you describe here is what would happen in reality on the road in a flat race. You and others are thinking of W/kg as a metric which defines that you should be doing the same speed as heavier riders with the same W/kg – but in reality pure W output is much more relevant on the flat, even more so in a peloton. I weigh 87kg and my 20 min FTP is 272. My W/kg is 3.2 which is lower than yours, but if I can hold 270W then isn’t it correct that I would be faster than you doing 230W? In group rides, W/kg is only relevant when climbing and that’s when its much tougher for heavier riders, as Eric explains above.
I’m in your situation but at the lower end of C. Its going to take me a long time to get anywhere near the top of the group. In the meantime my strategy is to let the top riders charge off out of the gate and try to link up with a group that’s slightly above my w/kg. If I can hang onto their wheels until the race settles down I find I can be competitive and make my own “race within the race”.
I know this is super late, but 6 months ago I was in your boat and now I fairly regularly podium in C. What made the difference for me was that initial effort. I found that instead of letting the fliers fly away, if I kept up with them for the first 1-3 minutes, the pace settled down quite a bit, at least until the last 300 meters or so.
Richard,
My weight and FTP are almost exactly the same as you so i feel your frustrations.
I don’t even bother racing in zwift because of this issue. I am really hoping that Veloton becomes a real thing.and provides a more realistic option for those in our situation. It certainly looks fantastic if they can finally release it.
Looking at these results a slightly different way…..
On the flat, a 50 kg rider holding 3 w/kg is putting out 150 watts and traveling at 33.8 kph. While a 100 kg rider, also holding 3 w/kg, is putting out 300 watts and traveling at 38 kph.
So the 100 kg rider is putting out twice the power (100%) but moving only 12% faster.
That doesn’t seem realistic.
Speed isn’t linear to power. The faster you’re moving, the harder it is to go even faster.
True.
In this test the rider is the same size and weight, so we can just look at the change due to increasing velocity from 33.8 to 38 kph. Power to overcome aerodynamic drag goes as velocity cubed. I make that to be a 42% increase, or 63 watts.
Let’s also add in the increase in rolling resistance due to increasing mass from 50 to 100 kg. I make that to be 23 watts.
150 + 63 + 23 = 236 watts.
Where are the other 64 watts going?
A reasonable Crr for Zwift is 0.4%. Calculate this for each rider and I get:
50 kg: 18.4 watts
100 kg: 41.4 watts
subtract that off and I get:
50 kg rider: 131.6 watts (against wind)
100 kg rider: 258.6 watts (against wind)
Then I can calculate effective CdA, assuming air density = 1.2 kg/m³:
50 kg: 0.265 m²
100 kg: 0.366 m²
The ratio is 1.38.
This suggests the cross sectional area is relatively close to the square root of the mass, assuming equal height. If you think of the body as an ellipsoidal cross-section of constant eccentricity, then both the width and the depth scale proportion to the square root of mass, at constant height. If you assume width matters but not depth, then this is this matches the data fairly well.
In the real world doubling your watts would make you about 30% faster. If it takes you 180 watts to travel at 30kmh then you’d need about 360 watts to travel at 40kmh. However, in Zwift I suspect the rider weight (along with rider height) is also used to calculate the drag coefficient. A lighter rider has less aero drag so not only are these tests changing the watts they are also changing the aero drag which helps the lighter rider partially overcome their lower watts.
So: 33.8 kph * 1.3 = 43.9 kph
That sort of feels right to me….
Weight isn’t static in these tests, though. It’s not like we’re using the same rider and just going from 150 watts to 300 watts.
Zwift uses weight + height to compute a CDA for each rider. So the 100kg rider will have a higher CDA than the 50kg rider. My guess is this accounts for much of the “slow down” for the 100kg rider.
According to gribble
100kg at 300w is 38kmkh
50kg at 150w is 31kmh
23% Faster
I think this is a great, and difference in speed at various weights is not unique to Zwift. The same phenomenon happens outside. My question is how closely does this scale to what happens outside? Is Zwift more or less generous to pure watts vs wt/kg?
No doubt, Zwift is about w/kg…even in the flat
In real world it’s more about pura absolute watts in the flat
That’s not really true. See for example:
Heil DP (2002) Body mass scaling of frontal area in competitive cyclists not using aero-handlebars. Eur J Appl Physiol 87(6):520–528
I got highlighted in one of your articles!!! I personally prefer big boned. I always knew there was an advantage for being heavier just not how much.
Another way of seeing this is by looking at the zwift rankings for C. All but one of the top 10 are over 85kg.
You should provide a power vs speed graph. Wkg doesn’t normalize the data as cda becomes a variable independent of gravity.
Great article Eric! I have a question that is a little off topic. I race as a B and weigh 83.9 kg . I typically can hold 3.5 most races but I’m curious when Kiss has a 100k race on Sundays I can’t hold 3.5 for that long typically for that long of a race I would be right around 3.1 to 3.2 . Do I still race as a B and get smoked but if I race as a C for the long race do I get Dq by ZwiftPower?
You would race your zwift power category. Which sounds like is B. All other riders should be in a similar boat and see a fall off from their 20min or 60min power.
So I’m a heavier rider, actually 100 kg exactly. On the flat and on the climb I’m doing 2.3-2.7wkg, not much difference. The lighter riders keep up with me on the flats By just putting Out 0.2-0.3 wkg more but on the climbs they push out 4-4.5 wkg. Why don’t they go at 4 wkg on the flat too since they obviously have that Capacity? Is it because they are lazy, whinging bastards..?🤓👍
They just attack where they have the best chances to succeed. It’s non sense for them to attack an heavy rouleur on the flat.
Try to drop them on the flat, slide at the rear of the pack, build up your speed as you move to the front and use the slingshot effect to escape.
What is your power profile?
https://www.trainingpeaks.com/blog/power-profiling/
If you are interested in results, try to pick race courses (including TT), that best suit your profile.
Ha. They’re just being smart racers. 🙂
Thanks Eric. Nice analysis (as always!). I am curious to know what a similar analysis would look like for Zwift dirt/gravel on a road bike. I assume that is also a good place to attack, and gives an added advantage to riders able to put out more absolute watts rather than w/kg similar to IRL?
Innsbruckring is really the case study. As a larger rider I can hold onto a pack all day on the flat but once w hit the city climb I’ll out the back. However if I can sag-climb and stay close enough to the little ones, I can put in a dig maybe 20m before the right hand turn at the top. This gives me the separation to keep them out of my draft on the descent and then I’m gone. I can usually completely recover by the next little bump. Then I can ride that down and if I can just hold threshold for the 2-3min(?) it takes to get to the finish I’ve soloed away.
You seem to have forgotten about the Age of riders.
As you get older, you do get slower.
I can’t compete against a rider weighing the same but is some 20 years younger.
I’m blowi g out of my bottom trying to keep up, let alone passing him/her. 😋
I’m old at 75, in Category C, 71 Kilograms body weight, 5′ 6″. I started Zwifting 8 months ago in January 2022, started racing in Cat D, but quickly bumped up to Cat C. over the last 3 months have gotten quite a few 7th place through 2nd place finishes in races. I do mainly flat course races such as two lap Tic Toc races, as I don’t do well in longer steep climbs. Have gotten two 3rd place finishes in Zwift Insider Tiny races by just racing in the 4th of 4 races. Now my Zwiftpower race ranking is down to 283, ranking me 155th in all Cat C riders. Of course I won’t be able to keep up this success for too many more years, as it it inevitable as I approach 80 years of age that my stamina and speed will decrease.
Great article, Eric. I really appreciate the science and writing you do.
Thanks, Ric!
Do you know if this graphs along the same lines or is there a blip every 25kg? The reason I’m asking is that at nearer 130kg than 100kg these numbers generally don’t work for me, I’d like to understand if I can just take these numbers and grow the graph out to show what I “should” be expecting and measure that against what I see on ZwiftPower to measure my performance.
Just trying to get the courage to ride Alpe but by my reckoning I’m looking at nearer 2 hours than 1.
I think the graph would be pretty steady if extended to 125kg. I would just extend the line out from what you see…
Thanks Eric.
Things will continue to favor heavier riders unrealistically until autobreaking in sharp corners are implemented. You see less heavyweight wins across categories for outside races because 1) repeated accelerations, say in a crit, will wear out/drop 100 kg riders faster due to simple physics, and also 2) 100 kg riders rarely can sustain an aero position as in Zwift…
I’d have to disagree. As a 85+kg racer, a lot of being efficient comes down to simple racecraft which Zwift will not be able to implement. Sticking in the top 5-10 of a crit pack has no braking at any corners. Hanging in the middle or back is a huge waste of effort with the yo-yo accordion effect. Also “100 kg riders rarely can sustain an aero position” is the same as saying 65kg riders can’t sprint.
Autobraking is a bad idea.
I can tell you that as a former Cat 2 road racer in Colorado who has achieved all of my best results through criterium racing, auto-braking doesn’t happen in real life unless you need to work on your racecraft. Zwift is a game and it will be impossible for them to implement the skills needed outdoors to transfer to indoor results.Ride at the front of the pack and every corner is smooth.
Maybe “autobraking” is the wrong word. I’ve also got plenty of criterium racing experience, and although you should absolutely be minimizing any braking during a real life crit, you DO slow decelerate coming into corners and accelerate out of them. It is not a constant speed. As a lighter rider, the more corners in a crit (as well as gradient changes), the better, as I felt that the repeated accelerations (small though they might be) generally did not fatigue me as much as it did the heavier riders.
Great article. This is something I’ve always wondered about. 🙂
The downhill weight advantage makes me think that’s the ‘real’ reason for the zwift supertuck, it just levels the field that little bit. I use it every chance I get and it really helps keep me from blowing up when trying to hang with powerful heavier riders on downhills.
It helps me recover after nearly burning myself out trying to stick with lighter climbers on the ascent.
Great post!
I am one of those lightweight triathletes (61kg, 288 FTP, 4.7w/kg) who has zero chance to compete on flat courses. If I want to be able to stick with the front group on a flat race, I have to sign up for category B.
Sure. And if I want to hang with the front group in a climbing race, I have to sign up for category C! We’ve all got different strengths.
Eric,
How happy would you be with zwift if you had to cheat (sign up for the wrong category) in order to be remotely competitive on flat and rolling races in zwift (most races on zwift)?
Great analysis, would be interesting to see *how well* this represents real world physics, I guess you could put numbers in to bike calculator
The following paragraph taken from the above blog doesn’t seem to make sense to me as it applies to zwift. You say the following Eric..
“So while our theoretical 100kg rider would be faster up the Alpe at 4w/kg than the 50kg rider at 4w/kg, there are very few 100kg riders who can hold 400 watts all the way up the Alpe! By comparison, there are many lightweight riders who can hold 4w/kg all the way to the top”
Because of the way zwift categorizes riders in races, that doesn’t matter. If there are no heavy riders who can hold 4 w/kg for a little under an hour, it doesn’t matter. It doesn’t matter because those heavy riders won’t be racing against those who can.
It seems as if the A+ group is almost the most fair then if it is a bunch of people of similar light weights producing high watts battling against eachother.
It is all of the other categories which are unfair to lighter riders. For all of you heavy riders who complain about how hard it feels to keep up on the climbs, that’s what it feels like for light riders all the time except climbs. On the climbs, it’s close to equal.
what about the idea of race speeds being determined solely by w/kg? Eg, each w/kg = a speed. As you produce more w/kg, one’s speed increases linearly? As long as categories are determined by w/kg, isn’t that the most fair way of doing things? It might not be the most realistic way of representing speed or translating how two riders would race against one another in “real life” outdoors. But, in real life, outdoors, those two riders wouldn’t be in the same category. For example, imagine a 50kg and 100kg rider who can each hold 3w/kg for an hour with max power of 500 and 1000 watts respectively lining up against eachother at a crit race. It would be ridiculous..it’s not realistic. However, that’s what happens in zwift.
My point there wasn’t about race categories – it was to say that, although it looks like the heavy rider will always win, it’s harder for heavy riders to hold high w/kg… which is why lighter riders generally win, when w/kg is the factor (eg, up climbs).
Race speeds determined by w/kg? Nah. Zwift needs to stick with real cycling physics. They’re certainly not going to change things just because the race categories aren’t set up quite right.
W/KG = watts per kilogram. more kilograms = more watts. more watts = more speed. It’s always been right there in the category description, takes some people years to understand it, if they ever do, some still think dropping weight is an advantage.
Thanks Eric.
As leader of a ride with a 2.0 – 2.5 W/Kg ride I really appreciate this data. It’s interesting and I will try to pass it along.
(I plug this site every ride because it is a wealth of all knowledge Zwiftful)
Well done Eric. So the chart shows two riders of the same age, with the same lifestyle. However, there are very few such couples. Performance at different ages cannot be measured.
I have a question not directly related to this, but about racing and maybe w/kg. Whenever I enter a corner with a pack, it seem I usually (frequently? – way over half the time) am the one who is moved to the outside of the turn, has to take the longest path, and loses a couple of meters.Do you know how it is determined, who gets the inside of the corner? It doesn’t seem to be order in the pack – maybe w/kg as you enter the turn?
That’s how it works for everyone. Everyone’s avatar takes the bad line! I wouldn’t worry about it too much. But if you really want a better line–go hard so you’re at the front of the pack going into the corner. Should improve things.
Zkus Zwift fórum. Někde jsem o tom už četl
According to gribble
100kg rider at 200w does 32kmh on the flat.
50kg rider at 100w does 26kmh on the flat
So both 2wkg including 7kg bike weight.
why are the numbers so much bigger than your Zwift numbers?
Because Zwift’s physics are not realistic and doesn’t fit with the facts on the road… it’s all.
I think that Zwift is clearly NOT realistic on the flat.
I think it’s not considering the effect of inertia generated by weight,as in the real world, when you are moving over 40km/h is easier to mantain speed for the heavier riders, so they accumulate more kinetic energy that help to overcome wind resistance (as the main opposite) with rolling resistance too.
Gravity is not the main thing on the flats, so… w/kg is only important maybe in accelerations.
Real world experience in ITTs reveals that lighter riders (weights around 60kg) moving even 6W/kg on the flat… have no chances VS heavier riders (75-80kg) moving “only” 4.8W/kg.
This is what everybody familiar with real world competition and numbers know.
So, if Zwift guys think that 20kg of weight or 10cm of height will worse CdA to the extreme of make that “compensation”, I think it’s clear that they are in a mistake.
This is EXACTLY why I hate climbing! Give me the desert any day! (Yes, I know…lose the kgs….but I love food, beer and wine too much!)
Interesting again and best explanation so far on being beaten nu heavier riders. Just too bad there”s a lack of true hilly races. Having a top 20min wattage in your cat hardly ever matters.
On side note, after reading this; would this also mean the 50kg rider has less advantage of a TT bike on Zwift, compared to the 100kg rider? Since the weight penalty of the added weight has a higher penalty for the 50kg rider, given the “true wkg”?
I’m a ‘bigger’ guy in cycling circles- 90kg at just over 5’10″/1.79m.
This is my experience… out of the seat I can generate much, much more torque than the small guys. Put me on a steep section and I can grunt up it, lighter guys have to spin up.
So long, shallow climbs? I suck. Steep, short sections? I love ’em.
As to whether it’s unfair that a heavier guy with equivalent power-to-weight travels faster. HELL NO!
I’m heavy because I carry more muscle.
Because I choose to participate in other sports the penalty is I need a bigger cardio engine than the guys who are 75kg because they only cycle or run.
Otherwise I could simply train less hard, lose 15kg of muscle (a sixth of my weight), let my power drop by 50w and I’d have the same w/kg!
Trust me, it’s easier to maintain a 250w FTP + skinny arms than a 300w FTP with 140kg bench press…
Power & Aero RULE
Completely agree with these conclusions. I’m a 67KG, 180cm rider with an FTP around 4W/KG. Even so I’m always struggling to hang on to the lead group in flat B races. However for routes that finish on a climb, assuming I can get to the climb near the front, things are very different and I can ride most people off my wheel.
It would be great to see a few races split along both weight and power to make things fairer, maybe not all the time, but occasionally.
“But let’s not misinterpret the data. While the heavier rider always wins at the same w/kg, what we see in reality is that it’s really tough for a heavy rider to hold high w/kg. Making 400 watts of power is much more work for a 100kg rider than making 200 watts is for a 50kg rider.”
Not arguing with the fact that 100kg riders don’t win mountain stages, but why would it be harder for a 100kg rider to maintain 4w/kg than for a 50kg rider ?
Because there are physiological limits. On average, a 100kg person doesn’t have twice the lung capacity, heart stroke volume, muscle mass, etc of a 50kg person. Our bodies don’t scale like that.
Loved reading this bit of data – since the data is to hand it would be very interesting to plot similar graphs showing plots of pure watts or speed… so say I’m 80kg, I could use the data to see what wattage I would need to put out to keep up with someone who’s 65kg pushing 250W 😀
What about crit type races that involve lots of accelerations? Does the lighter rider benefit enough from having to accelerate less mass every time to overcome the pure watt difference?
I didn’t read all of the comments, but my feedback would be it isn’t the absolute speed difference which matters (km/hr), but the fractional difference which matters.
So for example, what is the fractional change in time from the fractional change in weight.
On a “pure climb” with zero bike/clothing weight, a 1% increase in mass makes you take 1% slower. If you now add bike weight, a 1% increase in mass makes you (to first order) 1% / (1 + f) slower, where f is the starting fraction of weight from the bike + stuff. When wind resistance is added, it’s more challenging, because now you need to know the relationship between body dimensions and wind resistance. For example, one model is that wind resistance is proportional to height times the square root of BMI + a component for the bike. So this gets more complicated.
But the key result is the same as yours:
1. Light riders @ same W/kg have a disadvantage on climbing, due to the bike weight
2. Light riders @ same W/kg have an even bigger disadvantage on flats, since wind resistance isn’t proportional to weight generally.
3. Light riders have a physiological advantage, because they can produce more W/kg, generally, except for very short durations where muscles, rather than cardiovascular capacity, dominate.
https://www.zwiftpower.com/events.php?zid=920829
Hey Eric … I wanted you to have a quick look at this Tempus Fugit Time Trial I was just in …
I finished at 26:00, with a 346 watt average … I had a fairly steady pace, and probably ranged between 320 watts and 380 watts for the whole race. I was riding the Specialized Shiv Disk TT bike.
If you look the at 5th place finisher, with a time of 25:45, he only averaged 224 watts.
Now, I know there’s a large weight difference (111 kg for me, vs 55 kg for the 5th place finisher), and his w/kg was 4.1 vs 3.1 for me, but I was under the impression that on a fairly flat course like Tempus Fugit, my 346 watts should be more than enough to beat his 224 watts, no matter what the weight discrepancy was.
Is this a normal occurrence, getting beat on a flattish time trial despite putting out 122 more watts than the other guy, or does it look strange? Not accusing the other guy of doing anything wrong, just curious about the physics involved.
Yes… would like to hear your thoughts on this as well, Eric. This is a pretty absurd outcome, on a pancake flat course like TickTock.
*Tempus Fugit
Perhaps a much smaller rider as well ? Should compare heights too…
When will Zwift consider some age group activities? I am 65kg with a current 191 http://ftp…at age 75+ it would be not much fun if I were to be interested in more than just getting a good workout. No way I have the same capacity as some 30/40/50/60 yo…tia
I only started Zwifting early this month but am just a boy of nearly 60.
Like pbernstein above, I’d love to see age categories too.
Currently, I’d be quite happy to race but once I’ve rid myself of my Christmas belly I’ll be sitting on the wrong side of Cat C. I’d be floundering less than half way into a Cat B free for all. I’m perfectly content to enter real world CX races for that sort of humiliation. .
However, I’m not unhappy. In Zwift, I’ve discovered something which has enhanced my riding experience beyond anything I could have imagined.
The age category thing will come eventually. The game will have to bring it in to keep the egos of the current crop of youngsters happy as time begins to catch up with them.
In the meantime, I’m going to enjoy exploring the world of Zwift, get the Alp done in under an hour, get my Tron bike and just have a ball.
It would be interesting how the aerodynamics is calculated. Is a 100kg person just a “scaled up 50kg person” when it comes to front-facing surface area. Personally, I’m drop-shaped 😉
Watts per kg works well on steep climbs as a predictor of speed, whereas just watts works for modest downhills. Over an average course of up, down, and flat, presumably the best predictor of overall speed would be watts divided by weight to the power of n, where n is between 0 and 1 – any idea what n should be?
At 58kg, and after reading that, I feel I should get, at least, a 33% discount on Zwift.
From this very interesting article you can see that Zwift is a place much more suitable for heavy people or at least those who are 70 to 80-85 kg.
For us unlucky light cyclists there is little left but Alpe, Radio Tower and Ventoux.
Little bad since the whole thing is to be considered as a training platform since, fortunately, on the road the values between cyclists change a lot from what you see in the game.
I’ve been noticing that on a flat I often pass riders who are putting out half a watt or more per kilo. I just started paying attention to these number but I also just earned the Tron bike so I attributed, incorrectly, the difference to the bike. But I weigh 120kg so my total watts are much higher than most light riders. Going down hills the difference is ridiculous but then it’s the same in real life. Climbing is equally ridiculous but the other way. There’s nothing like working your ass off and getting passed by absolutely everyone including runners. Anyway thanks for this article. It cleared things up a lot.
This is a great article, explains why sometimes I’m going 3.5 up a climb and get passed by someone going 3.2. I’m always like, “how is that possible?”
Do you know at what grade it becomes more level at? Because, obviously, I see that occur at 3-4% a lot more than at 10%. My rough estimation is that it mostly equals out around 6%, but I don’t have any way to verify.
I find the w/kg so puzzling. I weight 48g. When I’m on a group ride, I always have to ride about .5 w/kg faster to stay with the yellow beacon. So if the beacon is riding 1.5 I always have to ride at 2.0. And if I fall off, forget trying to catch. Impossible.
It’s just physics, Deborah. Same holds true outdoors. When you’re as lightweight as you are, you’ll have to hold a higher w/kg than others, particularly on flat ground where watts is a better metric than w/kg.
In bike racing, typically lighter riders get dropped on flats, heavier riders get dropped on climbs.
Lighter riders win is on the climbs because they can typically sustain higher w/kg than heavier riders, since (for example) 5 w/kg for you would be under 250W, but for my friend Boone would be over 500W!
The advice on riding smart as a light rider doesn’t really apply in Zwift, only IRL.
A strong light rider using his “advantages” will be DQ or upgraded to higher category because the need to exceed the thresholds. While it may work against a new no-fit heavy rider with no clue, but against a heavy rider with some average game-sense and average fitness it doesn’t work, more than once per 90 days, since it will make you category enforced to next higher category. Sad
I get what you’re saying, but it DOES apply in Zwift, to some. Depends on how light you actually are.
If you’re light enough to not exceed the absolute wattage thresholds (which are there to help light riders) then you can exceed those w/kg thresholds but not be forced into a higher category.
For example: I race against a strong kid (14 years old, I think?) who only weighs 35kg. He holds some crazy w/kg in B races (4.6 or so), but that just lets him hang in with the pack with a lot of work, and sometimes take a win.
Nice article! But.. I’m curious: How did you get this data? I mean you wrote itt was a single rider (fixed weight) yet the x axis is with respect to the weight of the rider.
By changing the rider weight and re-running the tests. 🙂
How can you beat the heavier riders up the climbs in Zwift races when their w/kg are the same as yours due to Zwift categorisation? You can only beat the heavier riders that have a lower w/kg than you, whereas the opposite is not true.
Great article, thanks for all your tests and clear and simple explinations