Zwift’s beautiful new Makuri Islands world has been getting rave reviews for its landscapes, route layouts, and familiar Japanese feel. Zwifters are embracing the opportunity to earn new route badges and do some exploring, too: over 59,000 people rode the Flatland Loop in the first 6 days following the launch!
My Ride
After wrapping my head around the new world and its routes, I decided to try devising a route that would cover all the roads in one ride. The route I came up with seemed pretty efficient, covering all the roads in 31.8km. With semi-tired legs I just chugged along at a high zone 2 pace, averaging 221W (2.7w/kg) for the 64 minutes and 28 seconds it took to complete.
The Challenge
Can you do better than I did? There are two ways to beat me:
- Ride all the roads in less than 31.8km total distance
- Ride all the roads in under an hour
If you do both, you’re extra-special!
Give it a try, then post your results below. When you comment, please share at least these two key bits of info:
- The distance you covered in order to ride all roads
- The time it took to complete your ride
And before you ask: I won’t be linking directly to my ride – can’t make this too easy for you. But you can find it on Strava or in my Zwift account if you look around just a bit.
Prizes
There are no prizes. Nothing physical, at least. You’re doing this for glory, fitness, and the fun of a slightly nerdy challenge.
But the top man and woman will get their name (and hopefully a cool pic) posted on Zwift Insider’s social media accounts. And that’s better than a kick in the pants.
Winners will be chosen on Monday, June 7th. So you’ve got almost 2 weeks to plan your route, give it a go… then give it another go if needed. See you out there!
I was thinking of doing this today (before reading this post) but realised that I couldn’t be bothered to think through what turns I’d need to make and I just hoped Zwift would create a route for it at some point!
I agree. They should have added that route with the initial rollout. That way there would be a route that takes longer than 25min to complete.
Even better they could let people create routes manually for others to upload and ride. How cool would that be?
Unfortunately I’m just lacking the drive and attention span to ever do a route that requires any kind of manual turning.
The problem is they are clearly planning on an expansion, at which point the omni route will become an albatross, and they don’t appear to ever delete routes.
You got my attention. I’ll give this challenge a go today
How exactly do you go about choosing your own route Eric? I have only ever stuck to the set routes.
Pick an existing route and then at a junction manually turn a different way so that you ride your own custom route.
Thanks for clarifying Colin. That simple…..well I do already do that. Somehow I got the impression that Eric was designing a new route of his own and then following on-screen rather than making choices at each junction.
Unfortunately that’s not possible but it would be nice if you could do it.
Sounds like the perfect opportunity for someone to create a Rebel Route… 😎
Ha, I did this “just for fun” on my second time out in the new world, about five days ago. I didn’t pre-plan anything though, just wandered around until I’d covered them all. Did the KOM in both directions on purpose, and tried to find the “reverse Castle KOM” too, but it doesn’t actually exist. My stats were 55km in 2:01, (weighted) average power 197 or 2.3 w/kg. Not gonna win any prizes with that, but it was satisfying. 🙂
https://www.strava.com/activities/5335012527
Need to optimise.
I hate to break it to you, but it looks like you missed a bit of road (highlighted in yellow). Gonna give this a try tomorrow on the distance piece, but I know I won’t finish in an hour.
Every route I’ve tried to devise here, there’s always some little bit of road that I have to go back and cover…
Yippee, time for some Rebel routes!
Doing a U-Turn seems like cheating to me. 😀
Maybe somehow I altered my route, but last week I picked the 2 Village, 3 Village and Sea to Tree but ended up not actually riding them. I have many route badges so the process isn’t new to me.
Anyone else have this issue?
And to actually comment on this post, this sounds kike a fun ride. Not that many kilometres distance or metres of climbing. I may give it a go if the weather keeps me inside this weekend.
Interesting challenge! But the bar is a bit high for me 😅
did it in Around 31.1km with no u-turn and and a few less climbing Meters.
https://www.strava.com/activities/5361452216
Your route is excellent! It seems you can cut 300m off this by starting at the intersection just past the start line and cropping the initial portion out of the ride. This would make the shortest route 30.8km.
Interestingly, I cannot generate an equivalent route to this using a directed postman algorithm. The shortest I can achieve by modelling the route with 18 vertices and using distance between vertices as weights for 25 edges is 31.3km (https://www.strava.com/activities/5386227837 I made a small route error that cost me a 200m U-turn). I suspect that it isn’t possible to measure accurately enough (0.1km resolution) to get the distance weights correct and the route is very sensitive to initial conditions.
Also observing that none of the routes posted here cover every piece of road surface. They all leave out some portion of the roads at T-shaped intersections, as do U-turns.
31,3km with no u-turns. 1:03:22 in a recovery-pace 🙂
Didn’t think the navigation through, but maybe it wasn’t that bad?
Will try it again at full speed for a better time!
Question: if I copy one of the routes shared here (via Strava in a fox file) could I re-ride that same route in Zwift?
*gpx file
Challenge accepted! Did it today with the map printed out on paper in front of me, no u-turns and within the hour. Created a private Strava segment (see screen shot) to find out the exact time and distance: 31,62 km – 53:56. You’re free to check my Strava and Zwift accounts!👍
Might try this! Just to be clear, is this roads only, ie not dirt tracks? Including dirt tracks seems to make it quite a bit more complicated, and you’ve said roads in the article, but just wanted to be sure 🙂
Gotta ride the dirt too! It’s ALL rideable roads/paths on the map.
You can Submit this as a map speedrun on an official zwift speedrun category on speedrun.com
Did it in 31.37 miles last Sunday. took longer though: I was cruising.
https://www.strava.com/activities/5343868847
Made it in just under an hour, but I’m pretty sure I could have mapped a better route. Started as recovery then transitioned to a bit of a push after the KOM.
Aww crap, I went down an Eulerian traversable route inspection postman problem graph theory wormhole in order to find the shortest route and didn’t even get a sensible answer. See https://www-m9.ma.tum.de/graph-algorithms/directed-chinese-postman/index_en.html and http://www.mymathsteacher.com/discrete/postman.pdf and see if you can work it out!
Yep, that tool has a problem with directionality. I don’t that you can create undirected edges with it. You can add both directions, but it looks like it tries to traverse both. Right graph theory, but another tool is needed.
I’m old, fat, and slow, so one hour was out of the question. But on the distance, I got every road covered in 19.54 miles/31.45 kilometers. Strava map and link: https://strava.app.link/cPTsDuMlBgb
Wrong map. Here’s the right one.
https://www.strava.com/activities/5367198010
Missed out by 2 minutes but was 2k longer than shortest
“Do a ride on a certain route, but I won’t tell you the route” – that’s Zwiftspirit at its best…. And by the way, even if I’m part of a minority, the new map might by beautitiful in terms of graphics, but it is totally overrated otherwise…
It’s not a certain route though, it’s a challenge to find the optimal route. If everyone would head for Eric’s route, that objective would never be met.
Jarvis was 2 miles wasn’t it?
It was closer to 3 miles: https://zwiftinsider.com/jarvis/
Is it possible to get al bages in one ride?
Like other route badges… you can only get them by starting a new ride and selecting the route for the ride. So you could get them all in one SESSION, but each route would be a separate activity in terms of Zwift/Strava.
I love this. A great suggestion! 31.36km in 56min dead with 4 bike changes. Should probably have made that two, one at the base of the climb and one going into the city. Now to find a shorter route…
https://www.strava.com/activities/5371598619
meh— y’all have fun with this. I’ve done this with Watopia, every iteration, and this one is not too much of a challenge. This latest Watopia update with the little shortcuts near the bridges and the maya bridge really [REALLY] made an efficient WATOPIA WHOLE loop tough.
For the shortest spanning distance goal, would it be fair play to get to a starting point by sending a domestique to a cleverly derived location and starting off with “ride with”? I’m not certain that would yield a shorter full-coverage route, but it seems like it might…. maybe that’s the unlimited division. On further reflection I guess the same effect could be achieved by cropping in Strava, eliminating the need to have any friends.
I say go for it. Extra points for creativity! 😀
Did it this morning 31.3km in 1:15. 🙂
I used Tobias Baur’s route which is about 31.5km and neatly groups virtually all the dirt roads together minimising bike changes.
Directions are R,L or S which are Rightmost, Leftmost or Straight on at a cross-roads. R or L may actually be straight on depending on the junction geometry.
Start on Flatland Loop route on a road bike.
RLLR back to the banner
LLRR to the next banner
LSRS through Crit Village
R – change to gravel bike
LRRL to the Spirit Tree
LRL – change back to road bike
LRRRRL – village circumnavigation
LRRRL past the paddock
R – Stop at Crit Village
31.7km 55:21 on Zwift, 55:06 on Strava.
https://www.strava.com/activities/5375944985
That’s a solid ride with great instructions; just complicated enough to keep focussed. I saw 19.4 miles to the last R at the junction just before the return to Crit Village. 1hr 4min for me on borked legs (should have been a recovery day, natch)
https://www.strava.com/activities/5391895572
https://www.strava.com/activities/5380577785
not technically under the hour due to some technical difficulties with a friend, but short none the less.
So, to get technical, and go down the Alice in Wonderland of rabbit holes, the way to solve this is using the Chinese Postman Algorithm. First posed by Chinese Mathematician, Kuan Mei-Ko in 1962 to calculate how a postman could best choose a route to visit every point on the map in the minimum time (distance). But, and it’s a huge but, there are 14 odd nodes in the route (a node is a road intersection, and there are a total of 14 nodes but you have to add one node on the climb section to make a proper route, giving 15 nodes). To solve the algorithm you have to pair up all odd nodes so that the sum of the distances is minimized, which you then add to the total road distance (I calculate 27.7 km) to give you the minimum route distance to cover all roads. And this is where the brain starts to hurt. For 14 odd nodes there are 135,135 ways of pairing them, and I won’t have time to do that before the 7th June, lol. However, if you look at the map and start to look at the sections of it, the quickest route does, I think, pop out, and several here are doing just that route (@Peter Bevan). Repeat sections total about 3.25km, giving a total length of 30.95km (https://www.strava.com/activities/5371598619). Would love if someone could find a shorter route…
Last thing to consider, do you change to an MTB after the Crit Village rather than a gravel bike? And do you ride the rest on a TT bike or the fastest Road Bike you have? I did mine on the TT bike and MTB. Going to try another combination to see how it goes.
Obviously not the best route choice, including adding a couple of hundred metres from a double u-turn when I thought I’d gone the wrong way but hadn’t. Made it just under an hour…
This was fun. My first attempt was to wing it. Not surprisingly, this was not the best solution. It ended up being 23 miles and change. For my second attempt, I printed out the map and plotted out what looked to me to be something that might be pretty close to the best route. The result was 31.29 km. Not bad! Is it the best possible? I don’t know but I pretty happy with it. Thanks for the challenge.
I didn’t bother trying to do it under an hour. I’m forty years and many pounds away from my racing prime.
Oops! Here’s the link to the ride: https://www.strava.com/activities/5396934752
Although a different route, your route repeats the same sections as the route most are doing but your ride comes in very slightly shorter, about 60m, but that is probably due to the very slight difference in the finish point. A route to try nonetheless!
I was a bit longer than you, but I also did a fair bit of overlap at 32.5 km—probably a more efficient way to cover the whole map, I’m excited to try again soon.
Had another crack at it, using James Thayer’s route. The exact same distance as before, 30.95km, but a bit faster at 51:47. One bike change, an MTB before going into the forest climb section.
https://www.strava.com/activities/5412962028
Got it!
Can’t beat 31.8 km,but equalled it on a different route tou you Eric. If it had spawned me closer to the farmlands turn, might have just sneaked it.
Tried copying other peoples routes, and after cropping on Strava to remove lead in and overshoot, got down to 31.01km and just over 56 mins…
31.6 km and 53:25 min. I could crop it a. It yo save some time and distance.
For an alternative challenge: Are there any nice Makuri Islands routes that do NOT mix up the surfaces? So a full tarmac road so you aren’t dealing with changing rolling resistance and effort based on the surface within a route?
I know some people must like/love it, but personally I generally hate dealing with mixed surface routes in Zwift, especially where a section is too short to give an overall benefit from a bike switch but long enough to be annoyingly more difficult on a single specific bike type.
Would’ve loved it if they’d added more fully dirt/gravel routes for mtb/gravel bike use/fun, and of course more all-tarmac routes, but a load of mish-mashed surface routes doesn’t ‘float my boat’ I guess you’d say
I’ve marked the dirt roads on Makuri on the attached. There isn’t a route through all of them, with the only real practical route to avoid bike changes being to do repeats of the hill climb.
And you can happily do a variety of short or longer routes on road surfaces only, without having to go on dirt.
Thanks for the dirt highlights map though, when I get a chance I may try and at least figure out a couple of nice routes that minimize the dirt while maximizing the tarmac 👍👍
flatland loop has all tarmac
Yeah, I’m looking to sort some longer or more varied routes, now I know where all the dirt is I’ll see if I can create any fun rebel rides 👌
At last. 31.4 and no u-turns. Simple things.
After several attempts… 30.65km in <1h. Thank you all for the inspirations for the route! Let me know if I missed something, or else I'm done. …Watopia anyone?
https://www.strava.com/activities/5471017520
I finally got around to it and did 31.3km in 52:57 and then I continued til I had done all roads in both directions https://www.strava.com/activities/5917819004
So I did this including the new coastal roads – but very inefficiently as I didn’t really plan it out! Took 3.5 hours! https://www.strava.com/activities/8234981870