For many cyclists, Zwift is their first experience of riding “with power”–that is, seeing a wattage reading for their effort. And they have no idea if what they’re seeing is way off or absolutely accurate, and if that number means they are super strong or very weak.
We feel your pain, and we’re here to help!
Today’s Accuracy Outlook
Smart trainers and power meters are improving in terms of accuracy and reliability as the technology matures, and I think the vast majority of Zwifters enjoy acceptably accurate power numbers today.
But we still see riders on “classic” (or “dumb”) trainers which use Zwift’s virtual power algorithms to calculate power numbers–and those numbers may be inaccurate due to simple mistakes like under-tightening the roller or having low tire pressure. Additionally, other riders may be using uncalibrated trainers or trainers with known accuracy issues.
How to Test
The only way to test the accuracy of your power numbers precisely is to compare them with a power meter you know is completely accurate. This article doesn’t cover this sort of testing.
Instead, let’s discuss two simple ways you can learn if your power numbers are at least in the neighborhood of being accurate.
Test #1: The Outdoor Speed Test
The easiest way to approximate your FTP is to look at your solo outdoor efforts on a flat stretch of paved road. Ideally, you would do a one-hour out and back effort on the flattest roads possible with as little wind as possible. Flat roads are essential for this because when the road is flat, your weight doesn’t affect your speed much once you’ve accelerated to “cruising speed.”
Sidenote: I actually enjoy doing these rides a few times each year, I call them my “one-hour challenge” rides, and it’s a fun test to see how much stronger I’ve become since my last attempt. Here’s one I did July 2018 targetting 24 miles in the hour, and I made it by just two seconds according to my trusty Wahoo Elemnt.
Even if you can’t do an all-out solo hour on flat roads, you should know by now what kind of speed you can hold by yourself outside on a flat road with no major winds for a longer stretch of time (20+ minutes).
Here are rough estimates of the power needed for a solo road cyclist to hold various speeds:
- 25kph (15 mph) – 92 watts
- 30kph (18-19 mph) – 143 watts
- 35kph (21-22 mph) – 212 watts
- 37.5kph (23-24 mph) – 254 watts
- 40kph (24-25 mph) – 301 watts
- 45kph (28 mph) – 415 watts
So if you’re unsure if your power numbers are accurate in Zwift, just find the speed above that you can hold outdoors on a flat ride going all-out for 20+ minutes. Then see if that wattage matches what you see in Zwift when you’re riding alone and giving it everything you’ve got for that same amount of time.
[adinserter block=”1″]
Disclaimer on the Numbers
- Of course, the numbers above assume you aren’t extremely lightweight or overweight, or extremely short (or tall). They will be most accurate for a person in the 60-80kg range, and of average height.
- If you’re sitting in the draft on Zwift, your speed will be much higher than if you’re riding alone. Best to test while riding outside of the Zwift draft.
- The numbers above are an average from several bike speed calculators to check for accuracy. Each calculator computes speed a bit differently, but if you want to calculate numbers for your own weight/height, we recommend giving kreuzotter.de a try.
- Speeds can vary greatly if you are in a more or less aero position. The numbers above were done assuming riding in the drops on a standard road bike setup (not TT).
Test #2: The Sniff Test
The sniff test is much simpler than the outdoor speed test. But it’s also less scientific.
If you aren’t a trained athlete (let’s say you’re in your first year or two of serious cycling) and you are of average size, chances are very slim that your FTP is anywhere above 300 (for males). Chances are it’s close to 200, and maybe 250 if you have good genes or have been training for several months. Many riders will be significantly lower than 200 at this phase, especially if you are an older rider. And that’s OK too!
The important thing to understand is that you certainly won’t have an FTP above ~325 at this phase in your training, which is why any time a newbie rider posts about such an FTP increase in on Facebook, folks laugh it off.
It doesn’t pass the sniff test.
Pro male riders often have FTPs in the mid to high 300s, but these are genetically-gifted experts who have been training hard for hours per day, for years on end. If you’ve only been riding a year or two and you weren’t a serious endurance athlete beforehand, it will take you at least a few years to get into anything touching pro territory. (And even then, once you get there you’ll realize you’ve got a long way to go.)
Got Problems? Here are the solutions.
If you believe your power numbers are inaccurate, there are several ways to get them fixed:
- If you are on a classic trainer, make sure you’re following the manufacturer’s setup instructions regarding how tight the roller should be against your wheel. This is essential, since Zwift’s calculated power numbers assume you are following these instructions.
- If you are on a smart trainer, make sure your trainer is properly calibrated. Using the manufacturer’s suggested calibration procedure (usually via a simple phone app) is the way to go.
- Consider an upgrade! Today’s direct-drive smart trainers bring a high-quality Zwift experience within reach of more riders than ever before. See our Top-Ranked Turbo Trainers >
Conclusion
That’s it–two ways to quickly see if your numbers are at least somewhat accurate. I hope this helps some folks out!
Comment below if you have questions, or if this post has helped you figure out your power situation.
Something to consider. Roller on wheel trainers give you a mighty advantage. If you want closer to a real world outdoor experience, get a direct drive. It feels more like riding outside, especially if you crank the trainer difficulty up in Zwift.
Regarding the pro-FTP comparison: one has also to take into account how very lightweight pros usually are – we have to be talking W per kg… even if you reach absolute numbers in the 300eds, depending on your weight that can still be far far off…
I have some problems with this article. Here’s what I found:
Pros have FTPs in excess of 400w varying by weight and height. Tom Dumoulin has stated in a podcast(live slow ride fast dating to the 1st of January 2020) that his 20min power output is 465W. Apply the usual formula of FTP=20min*0,95 and you get an FTP of 441,75w. So 442W if we round it up. This is WAY higher than what is stated in this article(mid 300s). It would be ignorant to call Tom Dumoulin an average pro. But the sheer difference with the number in this article is too large. I can confirm that mid to high 300 FTPs are common under elite level racers and world tour level Colombian climber types. Average world tour and continental pros usually exceed that. Unless you take the absolute bottom end of the spectrum and focus on the 160cm tall climbing gods. As you stated, the sniff test is inaccurate, but the numbers used in it are non-representative for the pro athletes. Having said that, the estimates for the newer riders are somewhat realistic, and maybe slightly optimistic. But they are fine.
The speeds are optimistic to say the least for that power. Even on good asphalt and with no wind, I have to exceed the numbers put in this article. I also ran them through calculation software that can be found on google(go to numbers used at the bottom of this comment), and it came to the conclusion that for all the speeds provided in this article you would have to exceed the power numbers they were linked to. For 25kph you have to do 99W(difference of 7W). For 30kph you have to do 159W(difference of 16W). For 35kph you have to do 251W(difference of 39W). For 40kph you have to do 348W(difference of 47W). For 45kph you have to do 485W(difference of 70W). There differences are rather large especially at higher speeds and exceed what I would call ‘rough.’ And I therefore believe the numbers in this article shouldn’t be used for the test described as test #1. Which isn’t a very good test to begin with anyways, as the scenario is fairly unrealistic. You will always have wind, and finding a straight road like that is difficult, especially if you live in Europe or any other densely populated area.
Numbers I used: weight of 70kg(my weight) and 8kg for the bike(roughly the average weight for a decent road bike.)
CdA of 0,360. Source: https://www.adaptivehp.com/blog/2018/1/7/aerodynamic-efficiency-in-road-positioning the test subject was wearing aero clothing hence why I upped the number slightly. The numbers I used to get to 0,360 were a Cd of 0.75(which I found to agree with most of what I could find) and a frontal area of 0.48(which I also found to agree with most of what I could find online). The assumed position is the baseline positon of sitting on the hoods. You will find that your CdA goes down as you sit lower on your bike or use more aerodynamically efficient clothing you will go faster for the same power output(logically).
A drive train loss of 3%. Source: https://ridefar.info/bike/cycling-speed/mechanical-resistance/ we also don’t all maintain are bikes to the optimum. So I figured 3% was a fair if not slightly optimistic number.
Rolling resistance coefficient of 0.0042. I used a comparison tool(https://www.bicyclerollingresistance.com/road-bike-reviews/compare/continental-grand-prix-4000-rs-2018-vs-specialized-s-works-turbo-2016) and took the pressures at 5.5bar and took a nice round mean of the two different tires.
This was the calculation software I used: https://www.gribble.org/cycling/power_v_speed.html
I totally agree…pros reach way over 300watts I’ve seen pros on Zwift doing into the 500s not only mid 400s….Zwift is way off of being accurate…I’m.pushing 423watts of http://Ftp…I‘m.a bike tourist use to moving close to 300 pounds of cargo and bike….300 watts on a light road bike is nothing to me….this article as you stated in your comment is way off! LoL
Halo, i agree with you, but…. I weight 94 kg… I am amateur cyclist 41yo, riding for 3 years. My FTP Is 420watts(different methods approved). The point Is, that what matters most in Speed above 30km/Hours Is Wind resistence… And the difference in someone 70kg in aero position And 95kg 1.9m height on hoods Is massive when Cda may go from 0,2 to 0,45 which changes everything. The power
necessary to rise the Speed Is related to the 3rd exp of the Speed. Physics can’t be ruled out And beaten…. Wish you good luck And endorphines. Jan
I said, “Pro male riders often have FTPs in the mid to high 300s”. Which is absolutely true. Most pro riders don’t have 400+ FTPs, although SOME do.
The point I made, as I wrote, was “If you’ve only been riding a year or two and you weren’t a serious endurance athlete beforehand, it will take you at least a few years to get into anything touching pro territory.” And I stand by that. Show me someone with an FTP of 350+ who has only been riding for a year or two and wasn’t a serious endurance athlete before that… and I’ll say you’ve got a rare genetic freak on your hands, and “The exception proves the rule.”
As for the rest–we can quibble about speeds all day long. The point of this post was to give people ballpark numbers to look at it, and I think it does just that.
On your first point: I am sure I mentioned that in my original comment and explained what the levels of those riders usually are and what I have found with riders in my area and punt them into context well enough. Hence why I stated that they are my findings and so yours may differ. The definition of pro varies from country to country and from person to person. But I see UCI continental as pro and baseline elite as semi-pro. All of which is explained in my original comment. World tour pros exceed the benchmark set in the article, but that varies with height and weight. Which I also mentioned in my comment.
On your second point: I mentioned that in my comment too. I even stated that your numbers are optimistic but that they are fine. The reason as to why I went so far with the speeds – which is probably why you mention the 350W ftp example in the first place I assume – is because you did the same thing. I wanted to show my findings for those speeds because I know they are wildly off from my own experiences as an experienced cyclist and an aeronautical engineering student. We both know, and any logically thinking cyclist would know, that a beginner athlete will not have a 350w FTP unless there is some dodgy stuff going on.
Hi Eric, i am that athlet you ask for…. See my response no. 30875. Zwift Is an amazing Tool, i Wish you And your team all the best And thank you. I understood the importance of power measurement in cycling immediatelly , the simplicity And versatility And validity of watts Is game changing even for hobbies. Best Regards Jan Večeřa , Czech republic
I have a kickr snap. I had done the short ftp test in November and got an ftp of 183 which was a little disappointing given the riding I had been doing all summer and fall, but I was pretty tired when I did it.
Come December and the tour of London events, I was treating those like races, going all out and I saw an ftp increase to 274.
I took my bike off to lube the chain at the beginning of the month and saw could barely maintain 150 Watts for any period of time. I had done several spindowns in this period, but that was Wahoo’s suggestion.
After factory spindown, and a new ftp test, I am showing 223. I think I have been consistent in the tension on the wheel and tire pressure. I liked the idea of the 275 ftp, but with it moving all over the place structured trainings have been tough. Any ideas about why my reported power has been all over the place? I don’t have a power meter to refer to, but this has made getting one a priority. 😀
Wheel-on trainers generally suffer from poor accuracy, because things like tire inflation, cold calibration, and tire pressure affect things greatly. That said, the KICKR Snap is the best wheel-on trainer around, in my opinion. If you follow Wahoo’s instructions you should be able to calibrate it in a way that it delivers pretty consistent numbers.
Nothing is accurate on Zwift since no braking is needed…in real life we would all be dead riding as we do on Zwift lol
Because of the fact there is no braking on Zwift at all no measurements will even come close to the real world….this is nothing more then a video game for athletes so we dont get bored while training alone at home! Stop trying to make people believe something Zwift can never be! Zwift will never be an accurate measurement of power or speed ! LoL Keep dreaming! LoL
I completely disagree.
You said Zwift will never be an accurate measurement of power… but Zwift doesn’t measure power. That’s what power meters do, and they generally do it quite accurately.
As for speed: cycling speed is simply a matter of physics–which can be replicated quite precisely indoors.
Sure, there’s not braking on Zwift–but the point of this post is to head out and do a longer effort where you don’t need to touch the brakes. You can absolutely ride outdoors and compare that speed to Zwift.
I wish there was a high-accuracy power meter we could rent and use concurrently with whatever power meter we use for daily rides. Then we could compare meter-to-meter, and check accuracy. Next would be a power meter standard such that our daily driver power meter could be re-calibrated against the reference meter so we would have accuracy from then on.
This could actually be quite a brilliant business model..
Thanks for the article, interesting read. As a newbie to zwift, where can I find graphs like the ones in the banner (power over time especially?)
Thanks
Those are Strava graphs, @guest_30224.
My virtual power on indoor dump trainer is 60 watts higher than my crank-based power meter on road. My indoor trainer is calibrated for an average rider weighting 74kg riding on a 1% gradient. I am 65kg. To me, it is like riding on a 2% gradient using the indoor trainer. My indoor trainer speed hardly goes above 32km/h when I am putting out 250w. How to explain the big difference in wattage between indoor and outdoor riding?
It means Zwift’s power curve for your dumb trainer isn’t accurate and/or you don’t have your dumb trainer set up per the manufacturer’s specs.
If you have an accurate power meter, why not just use it instead of the dumb trainer’s numbers?
Okay, 270 inside is a struggle, 300 outside is easy. Done the calibration etc….. Just put down to zwift. I can’t get to max power inside, but can outside. Okay, speed… 90km outside is do able, 80km inside is big wattage. On a Tacx flux2 with all updates done.
Interesting Stuart – my FTP using a Stages outdoor is 340….indoor using Zwift on my basic rollers it is 330 approx (acceptable) – however using my Stages as the power source on my spanking new Neo 2T with same bike it’s more like 300w….do I have a problem with the “Smart” trainer?
P*ssing me off, as of course it is screwing up my downloaded Training Peaks workouts to Zwift that my coach sets…I’ve done two TT “races” now…one on the Neo and one using my rollers….of course my performance was far better on the rollers but using the same power source (btw I gave up on the Tacx power source)….
300w on rollers I can talk (HR 168) – HR on Neo @ 300w 185 – so unrealistic. Have I got a dodgy not so Smart Trainer? Thx
your power meters are probably single sided units so will be well off. @ keith your neo should be more accurate than your pm, i suspect stages will be off by 10% as that is a single sided unit. why does everyone always assume the lower figure cant possibly be correct?? some neo 2T have know to be off tho so could be a combination of the two. only way to know for sure is to source a set of dual reading pedals and do some comparisons to those. for clarity my single sided inpower crank gave me 25 to 30 watts extra, my ftp at the time was about 240-250 watts.
I have just upgraded my trainer from a wheel on Elite Qubo to a Kickr core……my last ftp test was 178w and was quite hard work, yet now riding on the kickr core 178w seems a doddle……was my previous trainer just wholly inaccurate and now I am experiencing a true power reading?
As long as your KICKR has been properly calibrated, I’d trust it over the Elite for sure!
I got a reality check when upgrading from wheel on JetBlack Z1 Pro to a Kickr 18. Now I have to accept the fact my power output is about 20% less (from ~210 down to ~170) than wheel on. Doing more analysis on outdoor rides and “Strava Power”, however its calculated, appears to be more in line with what I’m seeing on my Kickr. Not looking forward to my next FTP 🙂
Hi quick one for you,
I use a spin bike with power meter pedals, I’m 6’2 – 95kg
i did the short FTP test and finished with an average speed of 28.8km/h and 189 average watts 130m elevation gain.
my mate has a smart trainer
he is 5’11 – 80kg and did the same FTP test with an average speed of 31.8 and 148 average watts
im wondering how he used less power to go faster.?
He’s significantly lighter than you, and also shorter. That means he’ll always go faster with less watts… indoors and outdoors!
I average about 215-235w during a short training session of 50 mins. But today I was posting 900w. I don’t understand why as I’ve done nothing to my equipment. I’m thinking of getting a new speed sensor. My turbo trainer is a manual one. Welcome thoughts on what is wrong.
Could be this: https://zwifthacks.com/why-are-tacx-trainers-suddenly-boosting-your-power-and-how-to-fix-it/
Thanks I’ll give it a try
I’ve been using my Tacx blue motion turbo trainer for the past month along with the Zwift app. Ive been getting about 350 watts average but today all of a sudden it’s up in the 1000’s . I have not made any changes to my turbo trainer. Does anyone know why this is happening?
This: https://zwifthacks.com/why-are-tacx-trainers-suddenly-boosting-your-power-and-how-to-fix-it/
24 mph for an hour? That’s useless. How about using examples that relate to normal people?
Ummmm… I literally included 6 examples directly below the 24mph reference, at various speeds…
Wow! You got a rough ride on this article Eric. Today I rode a road bike at 250 watts for 1.5 hours over rolling terrain at 35kph. Your numbers look good to me.
Thanks for the article. I found it useful.
Eric love the soeed and wattage estimate and as ‘mid-sized’ racer (70s kg?) those numbers power to speed seem awesome. On road bike im drops I do 22.5 mph and with aero helmet speed suit and aero bars at least 23.5 (lol) at 280 ish watts?
Great ballpark test with adjustments fir size/weigjt as you said.
Another (similar?) reality check is if doing ride with others who are experienced and on climb where you settle in 8+ min but can be less especially if multiple climbs on a longer group ride. It becomes VERY CLEAR where you are vs other in watts if about same weight or w/kg!!! If others are tuned into their oower can triangulate yours to a reasonable range to compare to testing. People naturally push hard on climbs and reality is hard to not see!!! Ditto a particular climb you can compare your times to others who know their power AND see if you get faster on same stretch of hill.
(easier to push for 5-10 min on hill than long on flats solo and easier to find good stretch of road). ‘
As Eric said if riding around and going hard on flats you really can see usually around same speeds taking tail headwinds a kittle up or down you can get as Eric /headwind over time into account can start seeing ‘I can hold about X speed solo.’