Cycling is a voyage of self-discovery. While it teaches us a lot about how we think it also teaches us so much about our physical selves. How much recovery do I need? How strong am I at different types of efforts (sprints, endurance rides, etc)? And how do my physical strengths rank against others?
Thanks to the universe of data amassed by ZwiftPower, Zwifters have a powerful, free tool at our fingertips to help us better understand our physical selves. But before we dig into that, let’s lay a foundation.
“Rider phenotypes” categorize cyclists based on their maximum average power over different time periods, also known as “critical power”. The time periods vary from book to book or coach to coach, but one popular model uses 5 seconds, 1 minute, 5 minutes, and FTP. A rider’s power numbers may be placed on a graph, creating something like this:
chart from Training and Racing with a Power Meter, 3rd Edition
In Training and Racing with a Power Meter the authors define the following phenotypes:
- All-Rounder: generally even across all four time periods.
- Sprinter: downsloping plot with best power in the 5s range
- Time Trialist, Climber, or Steady-State Rider: upward-sloping plot with best power in the FTP range
- Pursuiter: an inverted V-shaped plot with the peak in the 1-minute range
What’s Your Phenotype?
Here’s the fun part: ZwiftPower displays critical power charts on the profiles of each ZwiftPower user, allowing us to see where our strengths and weaknesses lie. Just view your profile page (here’s mine) and look at the graphs directly below your profile pic. Here’s what mine looks like today:
Mouse over each bar for a little more detail, including a percentile ranking against the overall cycling population using Cycling Analytics’ numbers found here. (This percentile grade is split by gender.)
The bars on the left display your power numbers in w/kg, while the bars on the right display them in pure watts.
Your best three bars in terms of population percentile rank are color-coded – these are your strongest points. Orange is your best, dark blue second best, light blue third best. So in my profile, my best is 20-minute wattage, followed by 5-minute and 1-minute.
A Word about Data Quality
It’s worth noting that your phenotype charts will only be as good as the data fed into them. ZwiftPower uses your last 90 days of group ride data for its phenotype charts, so if (for instance) you haven’t put in a hard 20-minute effort in the last 90 days, your chart may show you as a weak Long Climber or Time Trialist even though you may be very strong. And keep in mind this is only for events set up on ZwiftPower, which means your free rides or personal workouts won’t be included.
If you are regularly participating in all-out Zwift races of varying lengths your charts should be quite accurate.
Using the Data: Racing
A quick glance at a Zwifter’s chart can tell us what sorts of race situations suit them best. In my chart above we can see that, first of all, pure wattage is my strength. Power to weight (w/kg) is not! So I’ll do better on flatter courses, where wattage determines speed more than w/kg. Since my 20-minute power is my biggest strength (69.98% rank), I’m looking for flat TTs, or road races where the pace stays high throughout and there is a lot of attrition as riders drop off.
Now look at my weakest bar: 5-minute w/kg of 4.46w/kg (49.22% rank). My chart says that if a race involves a substantial climb in the 5+ minute range, I will struggle to stay with the front riders. And guess what: that’s true!
Contrast my chart above with the current world top-ranked Zwift racer, Stefan Kirchmair:
You’ll notice his best numbers are all on the w/kg side, indicating he excels in races with substantial climbs. His 20-minute power of 5.79 w/kg places him in the 98.66% rank, indicating that very few riders can beat him on long climbs. This is the power profile of a rider who is nearly unbeatable when the road tilts upward!
Not all top racers are lightweight climbers, though. Holden Comeau is the 2019 Zwift USA National Champion, currently ranked 3rd in the world. Check out how different his profile is from Stefan’s:
Holden doesn’t win races by attacking on long climbs – he sits in the pack and wins in the final minute with a super-strong ramp up punctuated by a blistering sprint. With 1-minute power of 768 watts (99.07%) and 15-second power that is ranked over 98% both in pure watts and w/kg, Holden is a rider you have to shake off well before the finish if you want to win!
Holden vs Stefan
If Holden and Stefan raced against each other they could use these critical power charts to come up with a plan and give themselves the best chance at a win. Let’s say they were racing Watopia’s Mountain 8, which begins with a climb up the reverse Epic KOM and radio tower (~18 minutes), then descends to a flat finish of approximately 9km.
- Stefan would want to get away from Holden on the climb, putting as much time into him as possible while the road is steep. Then he would want some high-wattage teammates to tow him to the line, keeping him well ahead of Holden so it doesn’t come down to a head-to-head sprint.
- Holden would want to limit his losses on the climb, perhaps using teammates to lend a bit of draft and pacing. Then he would want teammates on hand to help him reel Stefan back in on the descent and final flat portions of the race, forcing the race to end in a head-to-head sprint.
Using the Data: Training
Just as the phenotype chart tells you your strengths, it also tells you where you are weakest. Coach Shayne Gaffney says “race your strengths, train your weaknesses.” Even if you aren’t working with a coach, you can use these charts as a simple guide to spot your weakest points.
I know I need to work on my 5-minute power while dropping a bit of weight. What about you?
If you’re looking for a resource to help you train, check out the latest edition of Training and Racing with a Power Meter. The other book I’d highly recommend is “The Cyclist’s Training Bible” by Joe Friel. This is a very well-respected resource from one of the most trusted cycling coaches in the world.
There’s much more we could discuss with these rider phenotype numbers, but I hope this gets you started. Questions or comments? Post below!
Hi my phenotypes are slightly confusing: For my 20 min power I have scored both 3rd (light blue) & 1st (orange). It was 3rd for Long Climber & 1st for Time Trialist. Perhaps this is to do with me being a medium weight (80kg) rider? Any thoughts? Thanks.
The left-hand column is for w/kg, while the right-hand is pure watts. My 1st, 2nd, and 3rd bests are all on the right, because I’m a bigger rider (82kg). Bigger riders don’t generally rank well against the population in power to weight!
I’m glad you mention data quality. Consider two riders, both doing the same time trial. One does a killer time trial and finishes in 19:30. The other has less power, and finishes in 20:03. Depending on how ZP handles the data, it may not give the first rider credit for 20 minute power (in reality, 19.5 minute power is also 20 minute power, reduced 2.5%, filling in the rest with zeros).. So when they compare numbers, the second rider may appear to have better 20 minute power. This is a problem with using fixed-duration power points instead of considering the… Read more »
TTs in Zwift (such as those on the Bologna course) are of 8km or 18km so they flank the 20min duration on either side. Someone doing an 8km TT will have their best 5min taken into account while someone doing a 18km TT have a chance to record their 20min power. I’m not sure if a scenario as what you describe exists, but you can share a link to it here.
I did Bologna in 19:56 and it didn’t give me a 20 min value and no FTP-raise (I raised it manually afterwards). The funny (or strange) thing: On the Zwiftinsider-verified Strava-Segment my time was 20:03.
It would be great to see how these phenotypes break down for different categories… whats a good w/k for a short-sprinter in cat a, b, c and d for example.
Rick Wenger’s series has actually analyzed this pretty well! Here are some posts:
Hey I havent done a race since july but i do workouts and rides all the time. Everything on ZP shows zero for me, not sure if that is becuase I have not done events in last few months, or is something is broken sync wise with my account.
Zwiftpower only takes into account your power from events, from the last 90 days.
So no broken sync, you just need to do events.
This is really brilliant and incredibly helpful! Thank you, Eric.
I’ve pushed myself in lots of group rides and races this year, and couldn’t really figure out what I was doing right or wrong. With this article I’ve just been able to look at my ZwiftPower data and understand how I got my best ever Zwift result, and how I might have avoided my last “blow-up” if I’d been ready for the part of the course which highlighted my weaknesses. I’ll definitely be using what I’ve learned in my next race!
The Phenotype are useful but they what happens if the race lasts a lot longer, they give no clue as to a rider endurance. I riders profile may show 4w/kg for 20 mins but what can he hold for 40, 60 or 90mins in a race situation. As you say the type of course makes a big difference, maybe the course could have something similar compiled from the phenotypes of the riders who do well on it. Could both of these be used to place a rider in a category or to adjust how many points you get in a… Read more »
Trouble is, where does the cyclocrosser and MTBer fit in thoes power profile groups?
Cyclocross and MTB requires the ability to repeat short, hard efforts… and the repeatability really isn’t quantified by rider phenotype.
Just a general comment on the ZwiftPower website – as a newcomer to Zwift and Zwift racing this website is horrible – it looks like something out of the 80’s; no explanation of what data you are looking at; or a FAQ page that responds to what might be real FAQ’s (like what does the data on this page mean? why is my NP got a red bar? why is my weight red in every race result?). I love Zwift and racing on Zwift but ZwiftPower is not helpful for on-boarding new racers. It would be great if Eric could… Read more »
Remind me how much you’re paying to use ZwiftPower?
It’s currently a free service so stop complaining (albeit Zwift have recently taken over stewardship)
Those that feel the need to respond in such a manner must lead an awful life. Everyone constantly disappointing them day in and day out. Let it go.
In John’s defence, I doubt he knows the history of Zwiftpower; with all the data on there it’s incredible to think that it wasn’t actually set up by Zwift themselves.
I think there are a lot of Zwifters who could benefit from being pointed towards ZwiftInsider help. I think Eric fills many of the gaps in knowledge that Zwift itself doesn’t seem to answer.
No way to know but ZwiftPower was a community established and run site until very recently. The layout and look of it harks back to this free, community run content. That said there is a ton of useful data on it. Perhaps start here for the inside scoop on some of the features https://zwiftinsider.com/top-videos-45/ (video 5 on that post) 🙂
Hi Eric, how does ZwiftPower calculate our phenotype numbers? Presumably a max figure looking back over a set period? Mine all seem to be a bit lower than when I last looked!
As mentioned above, “ZwiftPower uses your last 90 days of group ride data for its phenotype charts.”
great article! very helpful to understand how to use Zwiftpower data to see how best to improve. Little side-question: I wanted to check my last race (class B) on zwiftpower to see if it would explain why I was dropped where I was dropped, and indeed I saw that all 4 guys who ended in front of me had higher 1 and 5 min power and I was dropped on a short and steep climb (I hate Libby Hill) so that would make sense. However I saw that some of these guys had much higher stats overall than me, and… Read more »
Almost all Zwift races are set up so, if you exceed category limits, you are DQd on ZwiftPower.
If yours didn’t… well, I’d have to see it. Are you sure they didn’t race as A’s, but all categories started together?
No they raced as B’s and their output also was as a B so they weren’t DQd. But I guess
my point is that an A can take it easy for 2 laps and have something extra for the third lap, whereas in this case I was cooked in the last lap and couldn’t answer the last push. Don’t know but it would make sense to me of you would just race in the class that ZP says you are 🙂
This is really interesting, thanks.
I also came across an online calculator that gives some recommendations based on these figures.
It’s at — https://fft.tips/rankme
(It shows why I need to go early if I’m to stand any chance in a race.)
I used this calculator and it gave me the exact opposite result that Zwiftpower did.
ordered the book and gonna give it a quick skim. seen it referenced in a few places. interesting stuff :3 p.s. interesting that you labelled Kirch as a climber – tbh he’s more of an all rounder, his 15sec sprint is absolutely no joke and his 1min lead-in to it is also absolutely bonkers!!! not many ppl in the world who can push those kind of numbers!
Agreed – his numbers are incredible! They’re good all around, but his 20 minute w/kg power is his strongest relative to the population, hence my using him as an example of a climber. But yeah, I’d agree he’s a strong all arounder.
I agree, his numbers look pretty incredible all around. Definitely does everything very well.
For comparison, the only thing I do pretty well is long climbs. My ZP 20min is 5.60wkg but my 1min is only 7.60wkg. So his 1min power is almost 40% higher than mine, surreal.
What’s going on with Zwiftpower @eric Schkange?
No update since 28. December 2020…
Sorry. Eric Schlange
What do you mean “no update”? ZwiftPower is constantly updating.
I like the charts provided by ZwiftPower very much, but I would like to say that they can be misleading for beginners and riders who never had the privilege of being the with the best of one particular category. My stats are suggesting that I’m best suited for long climbs, but that’s because racing is for me is a matter of hanging on as long as possible. I’m almost going flat out from start to finish as a mediocre C-rider. My 15-seconds Wkg is 7.3, while I’m capable of hitting 11.5 Wkg for at least 5 seconds when I’m still… Read more »
Eric, the 4+4 bars on the screenshot at the top of this page are EXACTLY what I’d like to see, live, for each ride as I ride that ride (not 90 day bests). Can they be set up either in Zwift or with a 3rd party app overlay, or are they just Photoshopped in and I’m out of luck?
What about U shaped ”type”. Or is it simply I need to work on my 1 and 5 minute power!? Equal best attributes are sprinting and 20 minutes efforts. equal worst is 1 & 5 min efforts. Do well on a flat or rolling course but always struggle with climbs in races unless they are very short and punchy. Am continually dropped and then left to pace myself but end up with power curve similar to the leaders except for a drop off between 1 and 5 min power!. Any training suggestions. (other than intervals which I haven’t seen much… Read more »
Wow, this was a very informative and easy to comprehend article. I’ve looked at my zwift power chart before but didn’t understand it as much as I do now. And I couldn’t agree more that it is only as accurate as the data coming in is. It’s important to ride events often to get accurate numbers. Thanks for the great explanation and breakdown.
Is there another way to get this data accurately from Zwift or another program using a DD trainer?
I mean using data from races and ride on Zwiftpower seems a pretty haphazard way of testing and measuring progress.
I want to do a specific test to find these limits which is repeatable so I can properly gauge my progress (or lack of!)
Trainer seems an ideal place to do these tests but for 15sec power I don’t think I can produce the same as I would on the road since a fixed bike isn’t ideal for out of saddle sprints.