I lose a lot of Zwift races. But sometimes those losses are… special.
Yesterday’s was the P Race–one of my favorite races because it ticks off all the boxes for me:
- Fairly flat: long hills don’t lead to good race results for bigger riders like me
- Double draft: this lets me sit in if I want, or push the pace over small rises to help drop a few riders
- Longer: typically 30+ miles long, and I like slightly longer races, especially on Saturdays when I’ll be riding for a few hours anyway
- Separate categories: I prefer races where each category competes separately. Never been a fan of trying to hold onto the A’s!
Zwift Companion said we were doing two laps of Watopia’s Waistband. This route is 15.8 miles long, but (and this is important) we also have a 1.4 mile lead-in to the start/finish banner. So the total distance should be 15.8+15.8+1.4, or 33 miles–which is exactly what Zwift Companion told me when I signed up.
[adinserter block=”1″]
Zwift had rolled out an update the day before which included in its list of bug fixes, “Fixed bug where event distance estimations were wrong.” I was happy to see this because there have been distance issues in lap-based events for a couple of months now.
I got on course, warmed up, then hit the start pens. Let’s race!
We began with 73 B riders, but that group was quickly whittled down to 50, then ~35. This race was feeling like most of my P Race experiences: challenging, but doable. While I was suffering after half a lap and wondering if I could do two full laps of this, I also knew I wasn’t on my very limit. It would hurt, but I thought I had a good chance of hanging with the front until the final sprint.
Knowing the distance bugs were supposed to be fixed, I had my eye on those numbers as well. I noticed that the distance completed on the top bar now matched the side bar (that’s good!), but I also noticed the distance to go seemed too low. Early in the race if I added the distance to go with the distance ridden it totaled 30.2 miles. Later on, adding those numbers totaled 30.5 miles.
30.2 miles total? 
And this is where I started questioning myself.
Was it 33 miles total? My glycogen-depleted brain wasn’t sure. Maybe they left the lead-in off of the distance we’d traveled? (A silly idea in hindsight, but race brain isn’t always the smartest.)
As we rode through the tunnel bridge toward downtown Watopia, I looked at the minimap. My distance counter said we had less than a mile to go, and I swear the start/finish banner in downtown Watopia had that blueish glow that you see when an event uses a custom finish line.
Was that our finish line? Were we not doing two full laps, but instead finishing here in downtown Watopia? My distance counter certainly indicated so, and the blue glowing banner reinforced it in my mind.
With 300 meters to go I jumped. Hard. Even with tired legs I gave it all I had, because I saw there weren’t many riders who had followed me. I crossed the line in 2nd, as the distance to go hit 0, having averaged 832 watts over 15 seconds.
But where was the finish screen? Nothing happened. And a few seconds later, the peloton flew past.
The race wasn’t over.
My distance to go now read over 10,000ft. Feet!? My legs were burning, heart pounding out of my chest, and the front was already 10 seconds up the road. My race was done!
I limped my way into the finish–demoralized mentally and completely drained physically. I probably wouldn’t have won, or even made it onto the podium. But I bet I could have finished in the top 10. Instead, I finished 38th (30th in the official Zwift Power results).
Don’t Make My Mistake
I’ve heard from lots of other Zwifters who have made the same mistake I made in this race. Here’s how to avoid it: know the race route!
Zwift events are run for a certain length of time, a certain number of laps, or a set distance.
Zwift races are always a certain number of laps or a set distance. (Time-based events are used for group rides and workouts.) Lap-based races end at the lap banner, unless you’re on a point-to-point route like Bologna TT or Road to Sky. I should have trusted what the app said: two laps of Watopia’s Waistband.
Next: trust the event distance. It used to be, the Companion app’s distance figures weren’t entirely reliable–I think perhaps they left off lead-in distances? I’m not sure. But I learned not to trust them, which is unfortunate. Because now they seem to be accurate! So pay attention to that event distance number when you sign up for the event.
One more tip: if you’re like me, you don’t make good decisions at the end of a race. Our brains aren’t quite getting all the glycogen they need, so while the legs are firing on all cylinders, the mind is not.
So take notes if you need to, indicating the start/finish point of the race and the overall distance.
It’s funny, but on Zwift, like outdoors, racing smart is really up to you. Sure, you can just hop in and race whenever you’d like–but knowing the course and devising a strategy beforehand will give you better results.
About the Bug
I’d be lying if I said I wasn’t disappointed that this distance bug still isn’t fixed. It’s been broken for two months.
That said, it sounds like Zwift’s QA team is pushing hard to get this distance bug squished, and that perhaps it’s only affecting specific routes. That’s good news!
(Kudos to the team at Zwift HQ, by the way. I know you’re juggling a lot of priorities and projects. When a company is growing quickly and building things out in many directions, stuff can get messy.)
Here’s hoping this fix gets rolled out very soon, since it’s affecting a lot of races and group rides.
More to Come
I’ll probably turn this into a series of articles, because I’m always devising new and interesting ways to lose a Zwift race.
Got inventive stories of your own? Have you lost your race by sprinting for the wrong finish line? Share below!
It was bad luck how the distance matched up pretty even with the Watopia finish banner, I was at the back of the bunch counting on the finish in the desert like a few weeks ago. I got nervous when the bunch started to sprint and when I crossed the 300m marker there was about 350m left on my counter. Luckily I was able to catch back up to the bunch and factor in the finish.
For sure.
I’ve actually done this once before, but in that first race, I didn’t go all in on the sprint, and the finish line was maybe 500km down the road. I ended up doing well there, because I had the legs to go hard and long.
But I gave it all I had to the line this time–nothing left at the end!
You should be making an article about the minimum 45kg for kids. My sons is 11 and weight 31kg which is great for cycling but he gets penalized by the 45kg minimum and don’t get the power output is putting on the pedals.
This is broken and should be fixed first as there is a lot of kid that are cycling on Zwift. They cannot enjoy Zwift as we speak because the basic speed don’t work for them.
Plus this has been requested and was promised to fixed by Zwift 3 year ago… this is in their top request by user.
Yeah, that’s a long-running complaint for sure. I’ll see what I can do about getting a post out about it!
Amazing how they can’t get basic stuff like this figured out but want e-racing to blow up and eventually become an Olympic sport.
Had the same problem, was looking at distance left and was like 300meters, while I knew the finish was not there. when the 300 meters passed it jumped to 1300 meters again orso.
This was for the volcano climb
A similar thing happened yesterday on the KISS 100 race, covering the 2 laps of the Dust in the Wind course. ZP and the companion both said 105 km, game said 93.1 (adding the total of elapsed and remaining in my Strava screenshot). I think 93.1 is the distance to the little peak half way up Titan’s Grove KOM on this route. Anyway, as the km’s were ticking up and down, I started to doubt myself in exactly the same way, and was preparing for an uphill sprint amongst our little splinter of the field. I was pretty sure it would be 105 km, but seeing the numbers going to zero had me doubting. Instead, at 0, the distance started ticking up (guessing this was the shortest distance to the desert finish line) and then eventually down. This is something that ought to work a bit better! There were quite a few non-finishers in this event (55 starters, only 17 finishers in B on ZP), I can imaging this contributed to at least some of the DNFs.
Eric, I am the community organizer of the Zwift P Race. Thanks you for the plaudits! It’s always nice to have the races that one organizes get a shout out. Nonetheless, I am sorry about the distance problem you and others had. I did not compete in it this past Saturday, so I did not experience it. Still, if I understand your article, it sounds like it was entirely due to some unresolved bug and not to something I inadvertently told them (whew!).
From time to time the distances and other aspects of the races I organize get messed up, but I think it is usually due simply to human error. For example, the P Race and the other one I organize, the MET-5 (5 a.m. U.S. Eastern Time on Mondays), are almost identical races, except for the time they are held. However, they differ in distance, too. But on one occasion recently, the shorter MET-5 was accidentally run the longer P-Race distance. That made my early Monday morning even harder than it should have been. Afterwards, I contacted Zwift about it, and basically what it came down to is that one person (I think) is programming a bazillion races and events every day, and occasionally that person understandably makes a mistake.
One time in either the P Race or the MET-5 Race we were supposed to go two laps of Greatest London Flat. But this was just after that route had been developed, when the meaning of two laps for that route was not clear. So, the race NEVER ended. People just gave up, one by one when they had had enough.
This is a separate issue, but I long for the day when the organizers of Zwift community races can access the parameters to make the races run so that one Zwift personnel doesn’t have to do it all. Or, perhaps there could be some automation to repeating races, like there is for repeating entries in my iCal, for example.
Hey Keith,
Thanks for organizing all the races, man. It’s a thankless job!
Yeah, my issue above sits squarely on Zwift’s shoulders. Hopefully they’ll get this bug fixed ASAP, as it’s obviously negatively affecting many, many group rides.
yes and the athletes that are going to participate in the e-sport Olympic are the kids of today.
Tour of London stage 1 was messed up by over 800 metres today.
Went to zero and we still had a climb to do n turn right down mall for finish sprint.
Tour of London had the same issue. I went full gas when the counter was getting close to zero and then the counter started going up. Got dropped from my group and that was the end for me. Eric Min was in my group so I believe he may have experienced the same thing.
EVERYONE is experiencing the same bug. It’s just whether or not they fall for it! 🙂
The counters in the side and the HUD still don’t match on events. Saturday’s BMTR Fundo was about 0.5 mile off, not nearly as off as the 100-miler the week before, but still off. Have to think it’s all tied to this other issue. The mileages synch up as soon as the event is over. The counter in the rider list is higher, and the XPs go off this count, but the event mileage goes off the lower HUD display.
Interesting… they’ve been accurate for me ever since Friday’s update. Are you sure you’re running the latest update, Joe?
I had a problem with the remaining distance at stage 1 of London today. From the moment it went from 0,5 km and started counting meters, it hovered around 500 meters for some time, down to 300-something and back up to 500. I had a feeling that the number was the direct distance to the finish as a straight line in the air, and not on the ground, because as soon as I was around the last bend, it was counting down correctly.
It seems consistent for particular courses. Worth compiling a list of the bug’s behaviour on each?
I think Zwift’s QA dept is already on it!
I was in that race too Joe and experienced the exact same thing, distance on rider list was 0.5 mile higher than on the HUD. Thought the race was going to end at a banner too, but that is my own fault for not knowing that. Luckily we didn’t have to climb to the top of Titans Grove for the finish. I was trying to hold people off and was gassed.
Agreed. Let’s put Eric Minn on the spot on this one. It’s like one line of code somewhere to change.
I was talking about the kids weight thing BTW
Thanks for organizing the P race Keith. One of my favorite riders too for all the same reasons as Eric but really like the segregated start for the categories. Now if we could convince the A riders who join every week to join the A group, it would be perfect.
I should be clearer…. There’s some regular riders who are A category riders per zwiftpower but join the B race. Totally not the fault of the race organizers, as there is no way to prevent this and there’s only a handful.
Glad I’m not alone then. Tonight I watched the distance remaining ticking down on the Tour of London Stage 1, could I make it sub 1 hour…. Last 0.2m to go, push push push… Then the ft left appeared, 2,000 and climbing… Over a mile later we crossed the line. I should have known the finish was The Mall, but hey.
I knew what you were saying, Tim, and I agree. It’s a perennial problem — folks not “noticing” that, in the B race, they just averaged 4.5 watts/kg and that they should therefore be in the A. I guess Zwift Power resolves some of those problems, but it does not do much in the heat of the moment.
Don’t feel bad, I blew up in the final 100m of a one lap race on Sunday. Legs felt pretty good … until I got up to sprint. I jumped first, hoping to catch the others of guard, then was astonished I was stomping the pedals as hard as I could and only putting out a bit over 500w! Gassed out before the line and got passed by pretty much the entire pack. Worst spring in the history of Zwift racing.
I know the tour of London is not a race, but you still can’t trust the event distance. Listed as 25.1, ended up being 25.9.
Agreed! Kids love this concept but get penalized.
I’ve been a victim of the distance bug, too. Very frustrating to blow up early and lose places. If course, I guess everyone is subject to the same bug, but it seems like such an obvious mistake… How did the coders miss that!?
Also, the update was supposed to fix the “slipped pedal” bug in the pen warmup. No change as of last night.
Yeah. Good to know he “will see about getting a post out ” about a three year old problem. Guess I’ll continue training on TrainerRoad while they are counting their piles of money.
You’re welcome to write one for us to publish, Geoff.
From a programming perspective, it should be a non-issue. The rider list already accurately shows distance completed. Remaining is course distance minus that value. Whatever they’re currently doing, it’s obviously over-engineered to hell and back again.
Never noticed this bug before, until yesterday when I did Stage 1.
The distance was down to around 200m, when it started counting UPWARDS! I watched it roll up to about 600m or so, before it stated going back down.
So instead of the event being 40.2km, it was 41.5kms.
Hi Keith, perhaps race organizers can help us all until zwift finally fixes this bug. Can you, and other race organizers, include where the final banner is within the course description? If “desert sprint banner” was included in the race description, Eric, myself and others would not have been tricked into thinking it was downtown watopia. Just a thought.
Same thing happened to me in a Central Park race yesterday.
It was a relatively flat course with high speeds and I was in a tight group of approximately a half dozen riders jockeying for position coming in to the finish. I saw the counter drop to 0.5KM remaining and thought, it’s time to punch it! Shortly after I did so, the counter jumped to over 1000m left (1.1KM) much to my dismay. Was able to hold on, but was wayyyyy into zone 5 and over 190BPM. :'(
>I had a feeling that the number was the direct distance to the finish as a straight line in the air
That is exactly it. Pretty ridiculous bug if you ask me, as it seems to be counting down correctly before the last bit.
Yep, experienced this exact issue as well during the Stage 1 London Tour. First time I have seen it, but I just started in Zwift last March and only race once a week If that.
I noticed similar on tour of London yesterday, the distance to go countdown suddenly started to increase having dropped to about 900 yards. It increased to 2000 or so yards and then started dropping again. Luckily I knew where the mall was plus I had nothing left for the sprint anyway!
If anything the latest round of fixes has actually made this one worse. In previous events (mostly the tours – I don’t race often) I have noticed that the event distance was out at the start but gradually fixed itself by the time the last couple of miles came around. However, stage 1 of the Tour of London was definitely screwed up … the distance wound down to less than 0.1 miles to go, then jumped to around ten thousand feet to go, counted down to around 1200 feet to go then started counting upwards, to around 1700 feet then counted down again to the finish.
Obviously, knowing the London course (both in Zwift and real life) fairly well, I wasn’t working off that, and the turn into the Mall is blindingly obvious but I have never seen behaviour like that before and can well understand why racers are very bothered by it.
Another of the joys of progamming … your code can pass in house testing but still not work properly in actual usage. Just like military planning, bug fixes often don’t survive first contact with the enemy (or the user in this case).
I rode the Stage 1 the Tour of London and noticed the distance bug. I realized that the distance tracker near the end of the race is based on the straight line distance you are from the Finish Line and not accounting for the distance via roads. When you ride toward the line, it decreases, but when you make the turn away from the line, the ticker starts going up until you make the final turn back toward the finish line.
It seems like the distance is as the crow flies and not accounting for roads. This seems to only be a factor at the end of the race/event.
**I haven’t tested this theory on other courses/routes. This was definitely happening in London though on Stage 1.**
You are correct, Austin. It’s straight-line based!
Felt your pain. I knew the finish would likely to have been on the Mall so held out until after Trafalgar Square roundabout, and then went for it. My heart did sink when the countdown changed as I was going hard to keep with the pack I was in!
TDL Stage 2 has it also but I was ready. Taking a loop at the map on the companion app and a look a the veloviewer distances of the intended route made it pretty clear that I’d not be crossing the finish line in .2 miles. I held back then gassed it coming out of the subway climb with good results.
I think many of you are WAY to upset about how well our fake virtual bike races are going.
Friend with cancer = actual problem
Being upset about having to ride further in a fake bike world = not a problem
Having been a ‘victim’ of this before, I put special attention on the counters in today’s race (3R volcano climb, 13:40 CET). (And I’m usually in chasse patate anyway, plenty time to watch the counters). The race was advertised as 22.4 kilometer long, though this route is 22.8 according to strava and Zwift insider. Adding the distance so far and distance to go, also suggested 22.8 ( and then 22.9 amd even 23 for a bit). By the time we already had 20 done, the distance to go was 2.8, count down from there was accurate. Perhaps the error occurs when counter sets in the advertised race distance but the actual route distance is different?
I’ve been 205+ pounds for a few months now, currently riding at 206 in Zwift. Hills are definitely tough, but what’s tougher is heat in Houston, even just a bit. To give you an idea, I also run half marathons, and winter weather (at least what passes for winter in H-town) is a big help. I can train at a 5:20 to 5:40 pace per kilometer in the fall lead-up to winter races, and then race at a 5:00 pace if race day is chilly. And it often is in those morning starts in December and January. So my half marathon times are around 1:45 for the most part, which ain’t too bad for a 40-something runner!
On Zwift, it’s no different. If my power meter calibration app temperature reports 20° C or higher, I’m strictly a B rider, but if it’s in the teens, I can ride with the As in a flat race and then peter out relatively slowly on the hills. Maybe my power meter accuracy fluctuates a bit with temperature, even with daily calibration (which is what I do, unless it’s more often than daily). But a lot of it is the simple fact that at higher temperatures, my heart needs to work a lot harder just to cool off my large-ish body. And at lower temperatures, it doesn’t have to work quite so hard, which means more energy converted to watts for longer periods of time.
What’s comical is for me to enter an A race on a warmer day. I can still hang for several kilometers, but if the race is longer than 15 km or so, I’m getting dropped badly, even on a flat course. Since Houston temperatures fluctuate so much in the winter, and since I ride in the garage without heat or A/C, some races get funny fast. But since I have to work so hard just to stay middle of the pack, it’s still a good workout. And that’s what I’m after.
In summer, though, it’s always hot. And I’m strictly a B rider then.