Longtime Zwifter and Zwift Insider reader Spencer Jones recently messaged me with an interesting chart he’d put together. And it was no small task!
Spencer says, “Using your website, I pieced together what I think are all of the Zwift releases of roads and plotted vs time.” I double-checked Spencer’s work, and it was quite thorough. There were a couple of Zwift construction project data points that needed to be tweaked a bit, but overall, he did the heavy lifting.
Here’s the final chart:
Zwift Road Construction Timeline
A few notes:
- Each point includes a letter denoting which map the roads were added to. Mouse over that letter for a note explaining what was added.
- Technically, the first data point should be reduced by around 4.7km, since Watopia replaced Jarvis when it launched.
- We know of at least one world, Gravel Mountain, that was created but never released to the public. This world is not included in the chart.
- The 2020 Richmond point is for the Richmond Refresh. In one sense this didn’t add new roads to the map, but in another sense it did, as it let us ride the existing roads in both directions instead of just one way.
- The 2023 Repack Rush point is interesting, because it’s the only time when the total length of roads decreased, since Repack Rush replaced the slightly longer Repack Ridge.
Tracking the Trends
Looking at the chart, you can easily see there have been a few instances where there were sizeable time gaps between new road releases.
But what’s more interesting is the distinct falloff in construction rates. From the launch of Watopia in early 2015 to the launch of Neokyo in late 2021, Zwift averaged ~52.7km of new roads added per year. But since then, Zwift has averaged just ~22km of annual construction.
(In fairness to Zwift, the chart above doesn’t include Climb Portals, and Zwift has launched over 420km worth of Climbs via this feature since mid-2023. I’ll do a separate post charting the growth of the Portal library.)
Your Thoughts
What do you think of Zwift’s new road release rate? Should they invest more heavily into new roads, or should they invest elsewhere since we have enough roads to comfortably hold our community’s interest and avoid overcrowding?
If you’re a “more new roads” proponent, what would you like to see built?
Share your thoughts in the comments below!
Interesting. Does Zwift have any information on ridership on the road segments? A great deal of my riding is leading group rides over about ten courses in rotation, so I don’t get out much.😜
I wonder what the graph for users over time looks like.
New road distance is great to see, but is not the only measure that can be used. When a world has more roads, it is easier to add more routes with less actual new road. The recent France update is an example of this. It would be interesting to see the same chart for total route length.
My point too. Not much of new France was actual new roads. Does this graph add all the new route distances or the amount of actual new road not on a previous route? 2 very different things.
This graph is for actual new road. Not routes on the same roads.
I disagree about the usefulness of that metric. Deca Dash is over 50km of “route” for only a few km actually created. The New France routes are the same. Add a few km then make a couple hundred km of “routes” that never actually get used.
Though maybe it would be a worthwhile shaming-metric to point out how little was actually created.
Definitely agree – I feel like more routes over the same roads is entirely unsatisfying. It’s the same ol’ cracker with ‘NEW’ on the package.
First, excellent job putting this together! Is there a list of roads or expansions that never made it to production? Also, do we know if there will be a road that goes up to the planetarium in Watopia? I belive I have seen some road there. But to go up there.
No list that I know of. Gravel Mountain is the only world (and roads) I know of that Zwift created but never released.
Is gravel mountain ever coming out? Need to know if I should bother upgrading my Crux or not.
I don’t think it ever will, Mike. It was sort of an experiment, and if they haven’t released it yet… I don’t think they ever will.
I remember that there also was a video (or just screenshots) of Eric Min riding a 500m long velodrome.
Ah, yes, I forgot about that!
I’d love to see France get some more love. So much potential to not continue building it out. But, I’d also be happy with new roads in Watopia or Makuri.
Although I always appreciate new tarmac, a new world excites me least. As Jad Delokk mentioned, adding to mature existing worlds gets us more per kilometer built than a new or small world.
BTW Eric, is the timing of this post a coincidence…or do you know something we don’t? 🚴
When will Paris be connected to or inserted into France?
Indeed, that one is a bit of a head scratcher
I definitely know some things the general Zwifter does not… 😀 but as far as I know, there’s no new roads coming in the next month or two at least…
Thanks for ruining my summer. 🙃
But I’ll keep my fingers crossed for September.
Yeah, I mean what we have at the moment looks pre-built to add Tro Bro Leon style roads
Expanding Watopia is key, nothing touches Watopia
One of the Zwift staff said they were up against memory limits in Watopia on older devices so future expansions are most likely going to be elsewhere.
This doesn’t take into account the construction of Climb Portals, which seems to coincide with the slowdown in construction of regular road the last few years. A complete analysis of how much newly available road is constructed would have to include all the Portal roads.
While new roads are great, I think new routes are a better way to go. Rode London Sunday morning to tick off The London Pretzel, it was so quiet, appreciate it probably wasn’t on rotation. More worlds would lead to more dilution. I like
Riding Watopia as it’s always busy so just keep tweaking every few months.
One area that could be expanded is the Yorkshire dales, not far from Harrogate , with climbs like Fleet Moss, Park Rash and Rosedale chimney.
Most of the IRL-based worlds are obvious candidates for expansion. Bologna is presently an out-and-back and the Giro dell Emilia race every year does a circuit up the San Luca climb but descends the backside. NYC could extend to a Hudson River route. Innsbruck 2018 had a “long lap” in addition to the “short lap” although that climb was probably too steep for Zwift, but the area is full of options. Paris could add the Montmartre loop used this year, and go to the Tour Eiffel. Then France (not IRL) could have an alternate climb to the Ventop route, following the train.
But if you want unlimited new routes, I’d recommend checking out BikeTerra, which has scenery not as rich as Zwifts, but has user-designed routes. It is sort of picking up where RGT left off.
Great information. So, I do like expansion, but I would love to see investment in more realistic world in general. I think the graphics are amazing now, but think if we could get to the level of Unreal Engine. Maybe Zwift will eventually migrate over.
The pace over the last years in combi with new routes and portal is ok. The more (and faster) you expand the more “unreachable” it feels to tick it all off for new users (unleash +100 levels to keep carrots for the die hard riders). More important maybe: better downscale again to two maps open at once to keep the densitiy high enough. High densitiy of users is the core of the zwift experience if you ask me. Nothing so boring to ride up an almost empty ven top of Yorkshire)
I really want the European worlds merged into one mega world forming a European watopia.
Having London, Innsbruck and France as one world presents so many cool opportunities.
Imagine how fun it would be to do a lap around London, fox hill included then go finish off the ride with a trip through Paris rubaix?
Or do a painful trip through Innsbruck and finish with a flat circuit around the champs-elysse.
This just seems like a big missed opportunity to me.
“EUtopia”! <3
Going from London to Yorkshire would be awesome too (especially if they expand Yorkshire to include an area like the Dales)
Better still, link up the 3 British maps (London, Scotland, Yorkshire) as one map, and the 4 European maps (France, Paris, Innsbruck, Bologna) as another. Huge numbers of new route possibilities without any of them getting so big as to end up being completely empty.
Great idea but it’s probably scuppered by the Strava integration. You would have huge distance jumps between the routes that would wreak havoc on Strava.
Can’t believe we still don’t have Mt Fuji in Makuri! Climb portals are great but no subsitute.
Mt Fuji total rides 28,000
Tourmalet total 53,000
Willunga total 88,000
Just an example.
Ventoux 400,000 rides
Alp Huez 4 million rides
Nice stats! Where did you get that?
What needs to be added to reflect the full truth, Alpe du Zwift and Ventoux have been available for much more time. So maybe the rides per year might be giving more useful insights.
i definitely want to see more new roads as the primary objective, after almost 80k miles on zwift, i could definitely use some new scenery
Do you think that since there is a “southern route”, that there may be a “northern route” in the works?
One that climbs up the backside or the Alpe?
I agree with the “merge worlds” camp.
I’d like to see Crit City brought into Watopoia.
And merge some of the European maps together.
Seems like an easy way to create some interesting new routes without having to develop new roads.
Strava integration would be an issue with this idea, unfortunately.
Id love to see real offroad singletrack. Climb a path up the rock behind the Tempus Fugit pen, cross that huge bridge and head into the woods of Titans Grove.
I would prefer for them to put more effort into refining the system (particularly with racing and scoring of points), and content that keeps us interested in the platform. I think a single annual road expansion that is significant would suffice.
Alpe du Zwift is the best road in Zwift. I’d love to see more iconic climbs fully realized, not just in the climb portal.
I hope for new roads with every update, but in the end the choice would always fall on “Tempus Fugit” anyway.
I would love for them to have a route to connect Paris to the rest of France. They could do it via tunnel like the Makuri Islands. The same could be said for Yorkshire, London, and Scotland as well as Richmond and New York.
What happened to Bologna and Italy?
That could add one kilometer, put a velodrome in Watopia, France, and London. (and to those who say it is impossible, trainerroad has a great one and bkool has several.)
Given the current access of only selected climb portals, even when using the hack of changing the date on the computer, I think it’s fair not to include the portal KM’s here. But it’s an interesting visualization to what people have noticed and vocalized on the zwift forums, or other social media sites.
Does the slowdown in road building inform us about staff or finances at Zwift? Since elements (trees, rocks, buildings, etc.) are reused, I thought that the rate of building would increase, not decrease. The graph is counter-intuitive. I think it deserves a company response.
No, it really just highlights the state of their programming. They likely aren’t using advanced automated tools—probably because developing them can be expensive—which means new features are being added pretty slowly.
This is great but there is more to Zwift than km’s. I’m late but early to the game since I can track my time to some of these additions. But to Zwift’s credit, racing and workouts have advanced, mission badges added, new challenge beyond the Everest challenge, XP added, XP challenges, Zwift Academy, Baseline….
This must be length of routes, not roads, right? Because no way France has more km than Watopia, amiright? Tire Bouchon does all the old lowland roads in both directions in 60km, Ventoux is 20km, and the new roads are < 9km.
I think you’re misreading the chart. This is a chart of total road length over time. It doesn’t show France as having more roads than Watopia…
Yup. My bad.
I really like Richmond, Yorkshire, Innsbruck, London, New York Central Park (ground level), as being models of real courses. I’d like to see main/”zwift-able” features of World Champs, Olympic or famous World Tour courses more often (e. g. the Wollongong city lap, Rio de Janeiro, Paris as an extension to Champs-Elysees,…).
The climb portals are OK but due to the graphics no real replacement of the modelled courses.
I think even more interesting than new roads would be a course creation tool for rides and races, something like a course import to Zwift from zwiftmap.com!
agree – being able to plot a route and then ride it would be a real transformation
I think Zwift should add an ‘Out of this World’ world. Eg. Mars or Moon or something similar. This would prove to everyone that Zwift is more than an alternative to IRL, it can be a unique riding option.
Olympus Mons KOM with 26000m elevation gain, but at 2/5 of the gravity… 😆
A brand new 500m velodrome with lots of exciting new race formats would go a long way toward easing the frustrations many users are feeling right now.I don’t agree with Eric Min that steering is absolutely necessary for a velodrome to work well. Indievelo’s velodrome proves that. What really matters is having a solid in-game ranking system based on age and weight.
As I have noted previously in Zwift Forums – to optimize Zwift HQ’s road designer efforts related to what constitutes the most flexibility for it’s customers – just as the ‘bypass’ connector that was added several years back which bridged the two road up to the Radio Tower area instantly created more possibilities for route expansions – hence, the best ‘bang for the buck’ at this point for Zwift in road construction is to add three specific new ‘connectors’ (see attached image). The addition of these three new connectors will provide an order of magnitude of additional options and creative route possibilities – including something for climbers as well as new extended routes for inclusion in robopacer rotations (to give greater diversity to the usual ‘flat-route’ treks they currently cycle-through)… While purely based on kilometers of new road – these connectors are not monumental additions, however the depth of possible new routes that become opened-up across Watopia are exponential!
Eric: Do you have images attachments disabled? I cannot post my illustrative “3 Connectors” map
Thanks – it is now showing up!
I guess it’s more efficient to create small bits of new road that can get used in multiple new routes. Creating the scenery is probably the most time consuming. They found a way around that with climb portal and for the rest seem to use the strategy of creating shorter connection roads to make dozens of new routes possible. Be interesting to see the stats on route kilometers and route per road km.
Trouble is that they’re kinda running out of ways to add small parts to Watopia. Maybe connect the start of Epic KOM to the newish coastal road coming out of the desert.
Don’t think there’s any signs for a Rwanda world ahead of the world championship, is there?
You make a good point, that climb portal added new km too. Climb portal is great and you need to consider it, having a look at “new km added”. A 1:1 ratio on “climb portal km” vs “regular km” might not work though.
Thinking of RGTs magic road, it was automated. Just add a .gpx and the algorithm does the rest. I guess for Zwift it’s the same. (If not, they made their lives much more complicated than it could be…)
So in the end, I guess climb portal adds a lot of opportunities for zwifters, but Zwift could and (imho) should be getting back to the ~50km per year rhythm.
All in all I think it’s pretty shamefull how little roads zwift actually creates.. Perfect example is Jarvis Island, which was highly anticipated but in practice is just a standalone 5 km of road where all you can do is ride loops. I think the absolute HIGHEST priority zwift should have is creating more training and racing roads in such a way it will substantially expand the island and routes. It seems zwift doesn’t have the manpower and/or will/incentive to do this unfortunately.. Even by creating relatively small segments of road you can expand watopia substantially: for example, by connecting jarvis island to the jungle or the foot of the alpe, this would open the possibility of new bigger routes, instead of just making (quite useless) standalone routes of under 10km…
It’s a tricky balance, to be honest. With the Watopia + 2 guest worlds & 2 climb portals we’re well aware that Zwift prefers their worlds to be alive with riders. Having tried various other platforms where that wasn’t the case, I do concur with Zwift’s approach there.
Adding worlds means more time between availability.And lets be realistic, there are plenty of worlds where people go only when they “have to”.
So additions to existing worlds makes more sense then. But it also had to add something. Roads for the sake of roads will only have a limited attraction for badge hunting purposes and it’s a shame to put in all that development work only for the community to ignore it after 3 weeks.
The Grade was something new and it made sense to place it on the side of Epic and tying it to the jungle in the west and the desert in the north east gave a lot of new route options.
The new roads in France were pretty much a Paris-Roubaix promo event that made an otherwise dull France world a little more interesting.
@Eric Schlange Is New York missing from the chart or is my eyesight going?
Good catch! Yeah, it somehow got dropped – I just added it, and tweaked the average annual road length in the “Tracking the Trends” section so it’s accurate…
Hi Eric as Jaws is 50 years old do you know if any sharks are in the watopia waters?I often look for fins when crossing water
There are dolphins and a big whale, but I’ve never seen a shark…
It would be interesting project to link the different worlds via bridges….without the need to ‘jump’ from one world to the next, and do away with the world rotation