One popular question from wheel-on trainer users in the Zwift Riders Facebook group is:
“Why don’t my Garmin computer’s distance/speed numbers match Zwift’s?”
The answer to this is quite simple, although it can be hard to wrap your head around if you’re new to virtual cycling.
Simply put, Zwift calculates speed (and therefore distance) based on power, while your Garmin calculates speed based on wheel revolutions. Because of this, your numbers will never match. And that’s OK!
More Details
Let’s imagine you have your Garmin turned on while Zwifting. Here’s how distance computation is happening on your Garmin and Zwift:
- Your Garmin is picking up each wheel revolution via the speed sensor on your back wheel and multiplying the number of revolutions by your tire circumference to compute speed and distance.
- Zwift is taking the power you’re putting into the pedals and feeding that into its algorithm to determine how fast you go in the Zwift environment given your weight, height, virtual equipment, current road gradient, drafting status, and more. Read “How Does Zwift Calculate My Speed?” for more info >
Note: if you don’t have a smart trainer or power meter, Zwift is estimating your power based on known power curves for particular trainers (we call this “zPower” or “virtual power”). In this case, Zwift does use your speed sensor, but it is converting that wheel revolution data into power numbers. Read more about Zwift’s virtual power calculations >
Mind *blown*!
I see the same thing happening with a kickr core. Social ride recorded in Zwift is 55km, same ride according to Garmin watch 45km. Power numbers and duration are if course the same. avg speed and total distance differ. Issue can’t be wheel revolutions witj the kickr core, can it? Same sensors are connected to both. Probably the drafting effect in Zwift is not picked up on the Garmin?
My Garmin 945 used to be about 15% slower on average compared to zwift speed depending on the route. Last few rides it’s now saying 60%-70% slower or 8mph-10mph. Zwift is still fine and the power is accurate. I don’t get what happened. Haha. Time to reset it and see if I can fix it
Same for Neo 2T and 945. (Not sure it is 15% all the time but it is slower). It is slower if I use 100% trainer effect or not. I guess the main question is how Tacx (or any other company) calculates the speed it transmit to the cycling computer. I can’t find the answer to it.
“while your Garmin calculates speed based on wheel revolutions”
But in direct drive there is no wheel and in smart trainer the speed is taken from the trainer not the wheel (and today many do not have wheel based speedometer and just use the GPS).
So why there is still a (BIG) difference? How the trainer calculates the speed. For example Neo has the data on rider and bicycle weight and it generates the load (sloop) and calculates the power. So calculated and transmitted speed can be very similar to Zwift w/o maybe drafting effect.
Garmin’s speed correlates with the cadence,while Zwift’s speed correlates with the power.
At the end of the day, Zwift is a game. All I know is I ride my bike loads more than I would outside. I do more miles and burn more calories no matter how you wish to measure it on Zwift compared to outdoors. I only free ride so far and I love it.
Outdoor comparison isn’t the issue here. Many are using Garmin’s fitness load and training status metrics with indoor cycling at the same time as Zwift to maintain accurate stats with Garmin. When Garmin and Zwift have a 5-10% distance variance it causes confusion.
Sadly, Garmin doesn’t use Zwift data to generate training status so we have no choice but to track with both zwift and Garmin at the same time.
My issue is that my zwift power shows about 1/2 what my Garmin 1030 plus does. I have stages power meter both left and right. When pairing with Zwift for power, sometimes it give both cranks, sometimes the other. How do I get the two to match closer?
No! It’s not OK! E=MC2. A given mass pushing a 52×15 @200w @95rpm, the object is moving @26mph. Not 7mph.
Regardless of weight, height, equipment, gradient, drafting, heart rate, wind or any other external parameter.
E = mc2? You mean f = m x a.
Thx for the explanation – while I was rather “shocked” that it is that big. Yesterday completed my first 100 kilometer Zwift (using Elite Suito) and had the Garmin running as backup: Zwift = 101.1 km vs Garmin = 80.3 … anyhow I really enjoyed what I did!
Can you explain with a wheel off trainer how the two speeds would compare – Garmin watch to Zwift calculated speeds? Where or how is the Garmin device getting it’s speed data?
Your trainer will broadcast its own calculated speed, which is totally different from Zwift’s. So that’s what your Garmin watch is picking up.
every person here seems to be getting less on garmin, which would be fine but mines teh opposite, i came here looking to see that zwift was 15% lagging behind my garmin, now i feel i cant trust the data zwift has on me at all.
I have just got myself a Garmin Edge 530. I used it for the first time with Zwift today. My average speed on the Garmin was 77mph (I wish) and that’s despite it dropping out every so often. It’s paired as a Speed/Cadence sensor with my Elite Zumo, but the cadence never dropped at all and Speed was rock solid in Zwift. Power and Cadence seemed to match pretty closely between the Garmin and Zwift.
I’m hoping it was just a first time glitch and next ride I’ll have better results, but not holding my breath….
Reading the numerous entries on speed calculations and power calculations and still don’t understand why I can barely get over 3 mph on a flat. I am on dumb rollers with a Garmin edge 530, speed and cadence sensors, and a heart rate monitor if that makes a difference. The weird thing is that the speed sensor shows the same speed as my Garmin head unit on the bluetooth setup screen, but then when I try to ride, it can be off by more than 10 mph on the flats. I’m guessing from the reading I’ve done on this site… Read more »
Hi, I think there’s some confusion, but Eric got the answer in his post: Your Garmin is picking up each wheel revolution For direct trainers (I ride on a Suito), you dont ride with the bike wheel, but there’s still the trainer wheel! And yes, this is much smaller than the bicycle’s. So there goes Roberto, trying the Suito for the first time, and riding at 500 kms/hour. Yes, my Garmin sensors were configured with the wheel size of my bike (2096mm), and therefore every time the small trainer wheel (84mm) was spinning, it obviously had more revolutions, and sending… Read more »
I’ve been using a power meter for several years now and have a pretty good idea of power output based on perceived effort. Zwift consistently shows a lower power output (and consequently speed) than I know I am putting out. How can I adjust this to correct it? Elite direto smart trainer.
I used to have an Elite Direto, and it was very accurate, giving similar readings to my Powertap pedals and Tacx Neo T2. I’m no expert, but have you made sure that you have picked the FE option at the pairing screen. I remember when I first got my Direto I didn’t and it caused problems.
This makes very little sense if you have speed data available. If you’re driving a “wheel” (virtual or otherwise) through your real geartrain then the wheel revolution speed is your speed. Speed is speed. Period. That’s what speed is. All the Zwift “physics” can be used as they should be – to determine the resistance to your motion. E.g. if you’re going up or down a hill, if there’s headwind, etc. But in all cases: wheel revolutions = speed. It’s a direct and CORRECT measurement. I don’t understand the enormous mystery and difficulty in this world when it comes to… Read more »
Yes exactly. Wheel revolutions x circumference = speed. How can Zwift speed differ from trainer speed, when you’re actually pedaling? Different when you’re free-wheeling of course e.g. downhill without pedalling
That’s not the way Zwift works, if you use the default 50% trainer effect, as I do, if you go up a hill and maintain the same cadence your speed will drop which doesn’t happen with a real-world bicycle drive train, Same cadence, same speed…so Zwift is not using trainer wheel speed directly. I have found that speed/ distance correlation varies with the hilliness.
Thanks! I’ve read your article as well as the comments and -in spite of magnitudes and experiences- you are correct. My question is, if you’re using Strava/TR or other third party data bank which data should we upload Garmin/Kick?
I kind of understand the theory, but my garmin speed sensor reads nearly 10miles extra over a 25 mile zwift ride. One must be closer to correct than the other?
This may be true, but when I a producing 190 watts and my Zwift speed is four mph, either something is wrong or the algorithm is useless. Happens every time I use Zwift. Fortunately, I only use Zwift for FTP tests once in a while so not that big of a deal. If it was more accurate or if it also picked up my Garmin speed sensor, I would probably use it all the time.
If you’re on flat ground in Zwift doing 190W and only going 4mph in Zwift… that means you’re not doing 190W. 🙂
How are you detecting your power, and what does Zwift show on screen for your power numbers?