Zwift introduced varied rolling resistance in October 2019, adding a new level of variation to the performance of virtual wheels across various riding surfaces. Smart racers should take care to think strategically when choosing frames and wheels for mixed surface routes like Road to Ruins.
Here’s a complete rundown of all the rolling resistance numbers and resulting wattage affects for Zwift wheelsets.
(Note: we update this page as new wheelsets are released–see the changelog at the bottom for a list of changes.)
Zwift’s Road Surfaces
Zwift worlds use several different road surfaces. Here’s the complete list from fastest to slowest:
- Pavement
- Brick (example: part of downtown Innsbruck just before the Leg Snapper)
- Cobbles (example: Italian Villas)
- Wood (examples: Watopia fishing village piers and bridges)
- Ice/Snow (found only on the Radio Tower climb, as far as we know)
- Dirt (example: Mayan Jungle)
- Grass (found only on Repack Ridge, as far as we know)
Zwift has built the game so each wheelset can have its own Crr (rolling resistance) value for each road surface type. So a set of TT disc wheels may roll super fast on pavement, but terribly on dirt. Just like you’d expect outside.
When the Crr increases, two things happen in game: your speed drops, and resistance increases on your smart trainer.
Zwift Wheel Categories
Right now, there are three categories of wheelsets in Zwift:
- Gravel: includes only the “Zwift Gravel” wheelset
- Mountain: includes only the “Zwift Mountain” wheelset
- Road: includes all wheelsets except the “Zwift Gravel” and “Zwift Mountain” wheels
See the full list of Zwift wheelsets >
Each of the wheel categories has its own Crr values for each surface in Zwift. This makes each type of wheel perform differently across different surfaces.
Example: on pavement, Road wheels have a Crr of .004 while Gravel wheels have a Crr of .008 and Mountain wheels have a Crr of .01. This means Mountain wheels will roll slower than Gravel wheels which roll slower than Road wheels.
Wheelset Crr Values
We’ve done some testing to determine the Crr of all the wheelsets on each of Zwift’s surface types. Here are the current numbers:
| Surface | Road Wheels Crr | MTB Wheels Crr | Gravel Wheels Crr |
|---|---|---|---|
| Pavement | .004 | .01 | .008 |
| Brick | .0055 | .01 | .008 |
| Wood | .0065 | .01 | .008 |
| Cobbles | .0065 | .01 | .008 |
| Ice/Snow | .0075 | .014 | .018 |
| Dirt | .025 | .014 | .018 |
| Grass | .042 |
You can see why the jungle dirt feels much tougher on a road bike than a mountain bike. Because it is! The rolling resistance is nearly double!
It’s also worth noting that the Crr value for road tires on pavement (.004) is quite low. We’re virtually rolling quality tires on fresh tarmac. Living the dream!
Wheelset Wattage
If you know the Crr value of a particular tire/wheel, you can extrapolate the wattage required to overcome the rolling resistance at a particular speed. This is commonly done outdoors, and in our tests, it seems to work with Zwift’s physics as well (nice work, Zwift programmers).
Doing this tells us how much of our power is going toward overcoming rolling resistance. In turn, this tells us how many watts we can save just by moving to a wheelset with lower rolling resistance.
Here are those values, assuming a 75kg rider on a 7kg bike traveling at 40kmh (24.9mph).
| Surface | Road Wheels Wattage at 40kmh | MTB Wheels Wattage at 40kmh | Gravel Wheels Wattage at 40kmh |
|---|---|---|---|
| Pavement | 36 | 89 | 72 |
| Brick | 49 | 89 | 72 |
| Wood | 58 | 89 | 72 |
| Cobbles | 58 | 89 | 72 |
| Ice/Snow | 67 | 125 | 161 |
| Dirt | 223 | 125 | 161 |
| Grass | 375 |
Now things are getting interesting. So if you’re in a race on Watopia tarmac traveling at 40kmh, and you’re on a road bike while the guy next to you is on the Gravel bike, he has to put out 72-36=36 watts more than you just to overcome rolling resistance.
But if we move to the Jungle Circuit, suddenly that Gravel rider has the advantage, able to keep up with you while doing 223-161=62 fewer watts! (Of course, in the jungle nowadays we don’t usually go as fast as 40kmh, so this difference would decrease a bit.)
The numbers will get wild once we have more wheelsets with different Crr’s available. Pull out your calculators, kids!
Note: the wattage numbers in these examples don’t take into account additional differences that could occur based on varying rider weights, heights, drafting status, bike frame used, etc.
Fighting Resistance
Of course, rolling resistance isn’t the only thing you’re feeling on your smart trainer. In fact, it’s typically the least of the three factors which determine overall trainer resistance. Just like outdoors, we know that wind resistance and/or gravitational force are usually the biggest things slowing our avatars down on Zwift.
Changelog
- Dec 8, 2019: added gravel wheelset numbers to the tables
Great information. Thanks for the work.
Is this correct that a gravel bike needs almost 100W extra over road bike on the ice/snow surface? Zwift physics are just stupid 🙂 I would like to see this irl where the road bike hits the ground even before he can reach 40kmh 🙂
It’s correct, BUT keep in mind the only ice/snow surface in Zwift is the radio tower climb, which nobody will do at 40kmh. At a much lower kmh, that wattage difference is also much lower.
Is it just me or does Zwift need a complete overhaul of their 1 to 4 star ratings for frames and wheels? The information provided adds no value.
So why would anyone choose to drop 400,000+ drops on a gravel bike? Surely Zwift needs to correct this.
How about NYC’s glass roads? CRR = 0.0001?
Great information. Now I’d love to know the additional wattage needed to overcome the Tacx Neo roadfeel! 🙂
Does it mean that vEveresting on RadioTower are too hard/to long time compared with old records??
In real life rolling resistance decrease when tire size increase, not the other way around.
So gravel tires (with small threads) should get lower Crr values than road tires.
But at the same time, CdA (effective drag area) increase, and by a significant margin.
Combined with the less aggressive geometry of gravel bikes, resulting again in CdA increase, means typical gravel setups are slower than pure road setups at high speed (ie on tarmac), simply due to aerodynamics.
So either Eric’s calculations are wrong or Zwift physics model is wrong.