Last week’s launch of the new VO2sday Micro Races was a big success! 371 riders started, making it the biggest race of the day by far. Read the story and watch the video of my race day >
But we’re not stopping there. This week, we head to Makuri Islands for a new set of routes, with a race structure modified to help make the experience the absolute best it can be. Read on for details…
Changes This Week
I’ve made two big changes to the races this week:
- Two more time slots! I’ve added more times based on rider feedback. We now have three time slots to choose from:
- Zone 1: 11am UTC/7am UT/4am PT (see results on ZwiftPower)
- Zone 2: 5pm UTC/1pm ET/10am PT (see results on ZwiftPower)
- Zone 3: 11pm UTC/7pm ET/4pm PT (see results on ZwiftPower)
- Compound Score-based categories: Riders are now grouped based on your compound score, which is a metric Zwift records but nobody knows anything about. 😄 It’s based on your 5-minute power and body weight, computed as (5-min power in W/kg) * (5-min power in W). Example: my 5-minute power best is 359W, and I weigh 84kg. My compound score is (359/84)*(359)=1534. In theory, this should work well for grouping riders into competitive categories. We’ll see how it works this week, and adjust accordingly!
- A: 1800+
- B: 1400-1800
- C: 1100-1400
- D: 1100 and below

May 12 Race Details and Signup
Here are this week’s courses:
- Neokyo Crit Course (2.75km): a nice flat start
- Castle Crit (2km): time to climb!
- Suki’s Playground (2.75km): another flat one, with a downhill start in fact…
- Valley to Mountaintop (2.4km): ends on a bit of a climb, in the Golden Forest
- Three Village Loop (3km): finishing our series with a slight downhill

Sign up at zwift.com/events/tag/vo2sday/ >
Time-Based GC Results
Final GC winners in each category are crowned based on overall completion time across the 5 races. The leaderboards are available on ZwiftPower:
- Zone 1: 11am UTC/7am UT/4am PT (see results on ZwiftPower)
- Zone 2: 5pm UTC/1pm ET/10am PT (see results on ZwiftPower)
- Zone 3: 11pm UTC/7pm ET/4pm PT (see results on ZwiftPower)
My dream with these races is to make them be time-based, so each race would be (for example) 4 minutes long. This would let us dial in precise VO2 max interval lengths for the races, would ensure that all riders go hard for the same length of time regardless of ability, and (maybe the most fun part of all) set up a unique GC that is based on overall distance covered in the allotted time.
The problem is, ZwiftPower doesn’t support this. So I’m talking to WTRL and a few other community organizers to see if we can pull it off some other way. Stay tuned.
Structured Workout Racing
The VO2sday Micro Races are nothing like your typical Zwift scratch race. Here’s how these races are distinctly structured to be engaging and fun while delivering a proper VO2 max session:
- 5 races in less than an hour – that’s 5 hard VO2 max intervals
- Very short efforts: Each race is very short, with a target completion time of 3-5 minutes, just like a good VO2 max interval
- Recovery time: Races are 10 minutes apart, so a 5-minute race leaves you with 5 minutes of recovery before the next interval
- Mix of courses: Each week’s race courses feature a mix of flat, climb, rolling, and even downhill parcours. Riders with lots of pure watts have the advantage in some races, while riders with strong w/kg have the advantage in others.
- Compound Score categories: Since results will be driven by your power numbers, riders will be categorized based on compound score, not Zwift Racing Score. This takes into account your zMAP and body weight.
- Mass start: While riders are broken into categories for results, these are mass-start events with all categories starting together, so everyone has riders ahead to chase. (Remember, the goal is to push as hard as you can for the duration of the race, not to sit in the pack and conserve so you can sprint to victory in the last 15 seconds!)
- Drafting is disabled, so these are effectively time trials. (Yes, you’ll want a fast TT setup.) Hopefully this forces you out of the “sit in then sprint” mindset, and into “hold steady high power for the duration” mode.
- GC-Based final results: overall rankings are based on your cumulative time across the 5 races (see links above).
Questions or comments?
I’d love to hear your feedback after you’ve completed the latest set of VO2sday races. Share it below, along with any questions or comments you’ve got beforehand!
Zwift’s zMAP is a 6min prediction, not actual power recorded.
Compound Score is 5min Watts actual data x 5mins actual data W/Kg.
Tim Hanson’s zrCS adjusts Compound Score to take account of the much wider weight ranges of Zwift riders, compared to the professional(?) ~21 year old riders used in the original study data.
For a four pen event, the top end early signups for pen D in the 1800 BST are diabolical, 5mins 3.97W/Kg with 265W for the strongest rider so far (1043 zrCS or 1056 CS)!
Those sound like positive changes @Eric Schlange. The only other thing I’d like to see are the assigned, colored A, B, C, D kits so it’s easy to tell who we’re chasing.
Based on the calculation given, at 67 kg and 267 W (5-6 min PB), my score is 1048 ie Cat D. Zwift recently kicked me up to Cat B! Does Zwift consider this metric in their normal calculations for determining categories?
Compound score is not included in the normal pace group calculations. Those are w/kg, compound score is w*w/kg. Both are biased systems, but the bias basically swaps. Pace groups advantage heavy riders (ignoring the watt floor), CS advantages light riders.
Sigh, thanks for the extra two time slots, but for AU riders:
Tue 9pm, Wed 3am (original time slot) and Wed 9am are again difficult to accommodate.
Possibly the Tue 9pm event which means a late night trying to unwind and get to sleep after the finish at 10pm.
Alternately, take the morning off work Wed and do the 9am event 🤷♂️
Yeah … just doesn’t work for me. Same times as Tiny Races would have been good. Oh well.
cat D riders don’t seem to fit into the compound score pens. Is it not possible to have a pen E to accommodate lower power riders 🙃
I am confused about the categories. Were riders automatically placed into categories based on compound score or could they select which category to compete in?
I just did the first race in category D. My racing score is 236 and I finished 19 out of 40.
The top 7 riders had the following racing scores which would normally put them as category B or C riders…
538
473
328
427
434
390
394
Of course a person with racing score 538 crushed category D. Shouldnt they have been in category B or C?
Zwift Racing Score has nothing to do with VO2sday categories. Riders are grouped strictly on Compound Score – formula for that is explained above.
That said, this week was our first week doing that, and the D category had much too large of a window. Next week the cap will be lower, so riders will feel more competitive…
Feedback for the bottom PEN… top racers in the Pen D are top (legacy) C riders.
Maybe consider lowering the 1100 compound score boundary?
We were the biggest group at 40ish riders while the other pens were just around 20
Would love a chance to race with more traditional Cat Ds
Yeah, next week we’ll definitely lower the D limit. A bit of experimentation is required when using Compound Score, but it should be quite dialed in next week!
I noticed that today in all three VO2sday leagues (time slots), women placed *significantly* higher than last week. Perhaps this is due to the new Compound Scoring categories and/or the vibe of VO2sday attracting strong female athletes, but either way, this is a development worth following, sharing, and celebrating particularly in the context of growing women’s cycling and #WatchTheFemmes!
Last week, women barely placed in the Top 10: 10th, 8th, and 13th in Cats B, C, and D respectively. Today: League 1: 5th in B, 2nd in C, and 1st in D as the best-ranked women. League 2: 4th in B; PODIUM SWEEP + 5th in C; and 1st, 5th, and 6th in D. League 3: 1st in C; PODIUM SWEEP + 5th and 10th. Congratulations and cheers to them all!
Eric, please cover this in your next or an upcoming VO2sday article, and let’s hear also from the women who are racing and winning. Thanks for creating and evolving VO2sday and, whether by intent or by coincidence, growing an inviting and competitive race format for all regardless of gender. Chapeau to everyone who has participated and/or invited friends to this awesome series!
—
Some longtime members of the cycling community might know that I personally have worked for years in the industry specifically in support of opportunities and visibility for women in cycling, so this discussion is very much within my experience. These stats today are important, and I am very happy to see and share them!
Interesting point Chris – looking at the results (as I DIDN’T accidentally get tricked into riding these yesterday), there were no women in the A field and only one woman in B. And the women in the C field are mostly categorized in the old Zwift system as A or B. Do you think categorizing mixed fields in this way for longer races might work to help women feel competitive, or do you feel that the longer races would benefit women too much in Eric’s system? Also – if you’d like to discuss these things, please join us on the FCC server or TWG server, we’d love to hear your thoughts as someone following this for some time.
Great questions, Beccah. I assume by “longer races” you mean the more traditional Zwift races like ZRacing that are typically 15-30 km long. It also important to note that these new categories use Compound Score, which as Eric explained takes zMAP and rider weight into account. If there is a similar type of compound score that appropriately takes zFTP into account (as that is largely irrelevant for the short VO2sday micro races), then my answer is that probably yes, such categories could help women feel more competitive and just as importantly allow all the rewards of increased visibility and race podiums/wins (including comparing to Top 20, 10, or even Top 5).
Let’s be nuanced and consider though that I’m making my educated guess based on everything I know over the years about women’s cycling, but I am still coming at this as a man within our sport; I can’t say I know 100% exactly how women would feel and would turn the question back to you for your take and your direct peers’ take on that.
Again, though, if I previously was barely getting into the Top 10 based on categorization that didn’t fully take into account my power numbers AND my weight (which is generally lower for women compared to their male peers) and a revised system like what we had this week is more accurate across genders, then I personally would feel not just more competitive but also that everyone’s results were more accurate and representative of ability.
Onto your second question about benefiting women “too much,” I would want to know what “too much” means as there is nuance there as well and is perhaps somewhat loaded. If that simply means women would be placing higher than they had previously and legitimately out-ranking men in the same races more often, I would have to ask whether anyone has an issue with that and why. As I explained in my previous paragraph, if the categories are more accurate for the factors that matter for a particular race, then all racers regardless of gender, power output, weight, and w/kg can feel that the results are valid and justified.
We also have to consider this in the context just of Zwift racing or even slightly broader with eSports, where instant access to these power numbers and categories is feasible. I don’t see this system of categories as realistic for outdoor racing, although I suppose it would be possible if absolutely necessary. It’s probably obvious that it’s an apples-to-oranges comparison, but I bring it up in case anyone’s reflex is to compare to IRL categories.
Please do elaborate on this idea of “benefiting women too much.” I’m coming at this from someone who has followed women’s racing for a couple decades and has worked to lessen the disparities in visibility, opportunities, etc. in IRL racing/cycling in general, especially in the context of Half the Road, which I have supported both morally and financially. Thus, if the results are fair and accurate, this concept of “benefiting women too much” in cycling feels foreign even now after all the progress of the past decade, but it would be quite a fascinating swing of the pendulum that arguably is overdue.
Hi Chris – Thanks for being a supporter of women in sport, the world needs more of you.
Eric’s calculation reminds me of the seed scoring behind vELO – and I think it’s a genuinely interesting categorization approach worth exploring. Having followed the evolution of Zwift categorization from the old A/B/C/D w/kg bands through to Cat Enforcement, vELO, and now Tim Hanson’s vELO2 system, I’m always curious how each new method handles the nuances of mixed-category racing.
My questions to your initial post were really just me thinking out loud about how a zMAP-based system might play out across different race durations. One thing worth considering: women’s power output relative to men’s tends to become more competitive over longer distances, so in a zMAP-based system – which is typically derived from shorter efforts – women might systematically outperform their predicted category in longer races. That’s not a problem in itself (I genuinely love seeing women race well in mixed events!), but it’s worth asking whether the categorization still feels equitable and satisfying to all participants across a range of race formats, which is something vELO2 seems to be specifically designed to address by accounting for race type.
I don’t have a clean answer – just a genuine interest in finding ways for women and men to race together in mixed events in a way that’s meaningful and equitable for everyone.
Thanks for your kind words, Beccah; I’m always happy to do my part.
You bring up some interesting points worth exploring. “One thing worth considering: women’s power output relative to men’s tends to become more competitive over longer distances.” I am aware of this in the context of endurance, such as ultra-endurance (e.g., Lael Wilcox) and the original women’s Tour de France (multi-week). However, I had assumed that the typical Zwift race distances even including “Epic Race” and Fondo distances wouldn’t show any significant advantage for women, but now I see I should have checked that assumption.
Are you specifically saying that based on data and observation you’ve seen that a zMAP-based system disproportionately gives women an advantage compared to their male category peers in longer races? If so, what is the threshold for “longer” in which this begins to show statistical significance? I see now where your “benefiting women too much” thought can come into play.
I’ll admit that I don’t know the formulaic details of the vELO/vELO2 system and will look deeper into that for my own understanding. I can say for now based on what you’re sharing that for any particular race, the most equitable solution seems to be the categorization system that takes into account race type and rider attributes (e.g., weight, height, power output that is relevant to race distance and topography).
In the context of VO2sday, I’m seeing now that this could serve the purpose as being an accessible case study in equitable race categorization with more visibility and participation from the community. As I understand it, and let me know if you see other elements worth considering, the new zMAP-based categorization for MAP-oriented (MAP-duration) races is a fair and accurate system for mixed racing. As these next few weeks of the series develop, this will be a valuable opportunity for us to see how we can apply the lessons learned from VO2sday’s categorization amd results to other races.
I guess what I’m saying is that a zMAP based categorization system used for longer races might benefit women, in that the delta between a woman’s power curve and a man’s over time decreases. If you looked at Coggan’s power chart for instance you’ll see that the best male power records are 23.8% higher than the best female power record at 5s and 1m, 15% at 5m, and 12.4% higher at 20m. Men are punchier than women, but the longer out you go (enduro events as you mention) the closer they are. I wrote an article last year for ZI on the different categorization systems, I think we should try them all and see – Zwift is a great place to do this as we compare amateur athletes worldwide. Fascinating!
As soon as you cited Coggan’s power chart, I thought, “Of course! That’s the obvious answer to my question.” That is interesting to consider how a zMAP-based categorization could benefit women at longer distances, that is, within the range of standard weekly Zwift races.
I did go back to find your article and a few others you wrote, and your thoughts and the comments from others are fascinating insights into what the community wants.
I am curious to see more zMAP categorization in action and its acceptance by the community. If this allows more fair recognition for women’s efforts in races through higher placings, let’s explore that! As you and others have said, there isn’t a one-size-fits-all approach, so a more tailored approach (perhaps using vELO as you said) should be applied more often than we see now.
Compound Score is similar to the original vELO seed, but it was quickly identified that it’s not balanced over the range of rider sizes found on zwift. vELO seed uses a modified version that is more balanced, and ZRS seed uses a practically identical formula mapped to 0-1000.
Zwift could build a very fair categorisation system based on their physics model, it’s kind of baffling that they haven’t.
Yes I love the vELO score! I think the adjustments made to the compound score based on rider size and the way the score adjusts based on performance makes total sense. Pro Tim here!
Personally I like the Compound score cats. Together with vELO cats, the best so far. WKG based cats does not work well for lighter riders/women IMO, the lighter you are, the worse it gets. Only 1 scenario where WKG cat is an advantage for lighter rider, when a race ends with a long sustained climb. Almost no races ends that way. What goes up, 99% of the cases goes down (Muckle Yin et al) All other scenarios/terrains are fight-for-life, threshold/vo2max efforts just to stay in the draft.
Imagine if power lifters would be grouped based on (max lift in kg)/bodyweight but the winner is based on how much weight you actually lift. or boxers categorized by (punch power)/bodyweight. I think the heavier guys have an advantage, no matter the number.
Another aspect that is never considered when discussing comparing different cats/riders etc is the training-volume (not sure how to catch that in a good way) but I get the feeling that when looking at pure numbers/results and not knowing if it is a weekend warrior loosing toa very serious athlete with +15 hours a week, one will make the wrong conclusions/assumptions. If Pogi would have joined SISU Pinkki race 2 in disguise, many would think he won because he was lighter and had an unfair advantage. Maybe intervals fitness value or Zwift training score at least gives an understanding if it is a rider barely training, vs a guy that trains a lot.
Looking foward to next Tuesday 👍