With another year of Tiny Races under our belt, it’s time to crunch some numbers and share fun Tiny Race stats. How would recent numbers compare to last year’s stats? What do the week-to-week participation trends look like? Which country is home to the most well-behaved riders (and the naughtiest)? And who has completed more Tiny Races than any other rider on Earth?
Let’s dive in and find out!
Overall Data
With Tiny Race results saved since October 15, 2022, we’ve got over 21 months of weekly data to work with. Here are the total number of races finished each week since I began saving the data. This is broken down by time zone into the three event timeslots we run (Zone 1 at 2am, Zone 2 at 8am, and Zone 3 at 2pm Pacific time):
Zones 1 and 2 track quite closely to each other, with zone 1 being popular in the UK/Europe and zone being popular in the US/Canada. Zone 3 is always much smaller. While Zone 3 mostly exists to cater to Australia and New Zealand, it’s interesting to note that race completions have dropped off during the southern hemisphere’s winter, indicating that 1) Aussies and Kiwis don’t Zwift more in the winter and/or 2) a good portion of Zone 3’s racers aren’t in the southern hemisphere.
Next, we look at how many unique riders participated each week, combined across all three zones. We see a similar trend to the chart above, with numbers decreasing in summer months.
We peaked at 1636 riders on February 18, 2023, which was the second week of our special Worlds Experience races.
This Year’s Data
If we narrow our scope to the past 12 months (August 2023 through July 2024) we can simplify the charts and concentrate on more recent trends.
One thing we’ve been doing for the past several months is experimenting with alternative ways to categorize riders. First we started moving the power windows, then Zwift Racing Score was released and we’ve been testing it, first with Zwift’s standard category bands, and now with our own custom bands.
Each of these tweaks affected the sizes of the categories. You’ll see the E category is finally getting some use now that we’re testing our Zwift Racing Score’s 5-category approach, and category sizes are closer than they’ve ever been before.
Next, we wanted to break down participation by location. Which countries participate the most in Tiny Races?
It may look like the USA is first and Great Britain second, but if we look closer, ZwiftPower breaks out England, Scotland, and Wales as separate countries. If you add their numbers to Great Britain’s, GB edges the US as the most Tiny Racing country!
The Naughty List
The intention of the Tiny Races is to do 4 back-to-back hard races in one hour, but we found some riders were jumping into later races in the set so they could get a better result against fatigued riders. Therefore, we developed a unique “no sniping” rule that basically says “You can’t skip a race then return.”
How many disqualifications do we issue to snipers? A lot, sadly:
We’ve noticed a consistent trend:: Zone 1 gets less DQs than Zone 2, who gets less than Zone 3. Why? We’re not sure.
Disqualification rates by zone for the past 12 months:
- Zone 1: 6.83%
- Zone 2: 9.09%
- Zone 3: 9.58%
Going further down the rulebreaking rabbit hole, here’s a list of countries whose riders participate in the Tiny Races, sorted by who is naughtiest (or nicest, depending on how you look at it) in the past 12 months. This shows the percentage of race finishes which end up being disqualified for sniping. (Keep in mind you have to be registered on ZwiftPower to even show up in the results, so this chart – and all of the charts on this page – leaves out data from riders who aren’t on ZP.)
South Korea wins the award for being the naughtiest, followed by Russia. Both of these countries had relatively low participation rates, though, with around 250 race finishes for the year. And just like last year, Italy is the first major participant (over 2500 race finishes) to have a pretty bad score, averaging 12.5% disqualifications.
Kudos to Costa Rica for having the lowest overall DQ rate by a good margin. Also, a shout out to the good people of Sweden, who just like last year have the lowest DQ rate of any major participant, coming in at 5.6% with over 7000 race finishes.
Most Prolific Riders
Lastly, we wanted to see which riders had raced the most Tiny Races over the past year. The results are impressive, with last year’s most prolific rider maintaining their place at the top, averaging 7.2 Tiny Races per week. Nice work, Naofumi!
- Naofumi Nozaki (AHDR): 375
- Christoffer Wikman [ZWDL]: 222
- Tak Itagaki: 204
- E Payne: 190
- Mads Buster: 180
- Kaan Kayin: 180
- Berk Kayin: 180
- Robin Andersson [eSRT] CRP: 178
- Michel LAURENT: 176
- Henry Li-Zwift Fighter: 171
Looking a bit deeper, 67 riders have successfully completed 100 or more Tiny Races in the past year. Awesome!
Three Requests for Zwift
The Tiny Races have always pushed the boundaries of Zwift’s race events. For example: did you know we had to customize the events to only remind you of the next race 1 minute before it starts? This avoids riders getting reminders popping up as they near the finish of an earlier Tiny Race!
There are two features I’d love to see from Zwift because they would simplify the signup process and greatly reduce cheating in Tiny Races:
- Sign up for all 4 events at once: I get it, the idea of signing up for 4 race events isn’t exactly intuitive. I’d love it if Zwift could design it so you do one signup, and you’re signed up for all 4 events. This would reduce confusion, and reduce the number of snipers.
- Verify rider category when entering pen: if a rider signs up for an event in, say, the C category, then gets upgraded to a B before that event begins, Zwift still lets them join the C race. Zwift should verify riders’ current category before letting them enter the start pens. Signed up for to low a category? Just pop up a message and let them pick their new category.
- Don’t automatically move riders to an event when they’re already in one: if you do get within that 1-minute time window, Zwift might pop up the reminder of your upcoming Tiny Race then take you out of your current race and into the next one before you’re finished. That’s not ideal. It seems like an easy fix: if the rider is currently in an event, don’t automatically take them out to the next one. Force them to choose that option. This is how it should work throughout Zwift.
Looking For More Numbers
Perhaps I have a fourth request.
As I crunched the stats above, I was reminded of how painfully difficult it is to extract useful event data from Zwift’s systems. (There’s so much other data I’d love to see, and I’m only able to generate the data above because I have it stored in a custom database which I use for crunching numbers and disqualifying snipers.)
As an event organizer, I would love to have access to a portal that shows me participation numbers, rider satisfaction scores/comments, and other stats from my events. This sort of actionable data would help me create a better experience for racers, which would be a win all around.
Wrapping It Up
The Tiny Races have always been about trying something different. I’m not particularly interested in holding “just another Zwift race,” which is why these races have such a unique setup. It’s also why I jump at the chance to test beta features like pack dynamics and custom category boundaries.
Thank you for trusting me enough to race the races and share your feedback. It means a lot.
As we head into peak Zwifting months, we’ll continue pushing the boundaries of what’s possible in Zwift racing. See you on course!
Questions or Comments?
Did you find any of the data above particularly interesting? Got a Tiny Race question or comment? Share below!
Zone 1 and 2 are generally Saturday morning races for UK/EU and Americas. Would you consider putting Zone 3 before 1/2 and making it Saturday mornings for Oceania/Asia as well instead of Sunday mornings?
Thanks for including the naughty/good list…It would be interesting to see it weighted. Right now it looks like a bad stereotype for the Russians who race.
What do you mean by “weighted”, KJ?
I have joined a few tiny races (sometimes only one or two of the four and often whichever one lines up with timing for me) not to snipe but just to get a quick high intensity workout in trying to keep pace with the leader group. Wondering if there are a lot of folks like me that may have factored into the data for DQd or “snipe” people. ?
If you’re joining any race but the first one scheduled, you’re DQ’d, regardless of your intentions.
If Zwift auto-upgraded categories and checked for each race then people would get caught out in your cull of people not doing all four races unless you changed your rules and process to allow for that.
When it comes to checking if people are sniping, I don’t look at race categories. So if people were upgraded partway through, they wouldn’t get accused of skipping those earlier races or anything like that.
Is that what you’re alluding to?
Rankingpoints shoud only be given out aggregated after 4 races. Now a lot of racers use 2 or sometimes 3 races to save power for the last race against fatigued competitors
The other minor annoyance for people who use an actual power meter. If you enter the pen and (for example) go to the bathroom, chances are you’ll find yourself expelled from the event. You need to get pairing going again and get back into the pen. Which given the tight times between events can be nerve-racking.
Yeah, unfortunately that’s how Zwift’s hardware verification works. I agree it’s not ideal.
You began with over 2200 riders in August 2023, but by August 2024, that number dropped to 1400. I cannot concur with your assertion that the figures have not declined. The issue at hand is whether people dislike something about the race, or if it simply mirrors a broader trend of people moving away from Zwift. Are there any intentions to make changes, aside from the extreme pen allocations of 530-650 or 580-1000?
I didn’t see where he said figures have not declined, maybe I missed that. But in any event, I would suspect the minor declines may just be due to an initial drop once the novelty of these events wore off (and people leaned how difficult these were), plus the anti sniping rules being put in place. After that initial drop the year over year comparison seems pretty steady.
Anyway, in my opinion, one of the best race series on Zwift, and thanks Eric for all your hard work in putting this together.
Yeah, I don’t think I said anything about numbers declining (or not) above… did I? (Maybe I wrote that and forgot about it… but I don’t see it.)
I’m not sure where you’re getting your numbers though, Steve. If you look at the number of unique riders, the last two weeks (July 27+Aug 3 2024) average to 581 riders each week. July 29+Aug 5 2023 average to 653. So yes, numbers are a bit lower, but not as dramatic as the numbers in your comment.
(If you look at race completions instead of unique riders it’s similar: 1962 average for those dates in 2024, 2095 average for the 2023 dates.) I don’t know why the numbers are slightly lower. Sure, it could be because people don’t like the way Tiny Races are set up. But it could be lots of other reasons.
You ask about pen allocations and mention 580-1000… but I’ve never used that pen allocation. I’ve used precisely two pen allocation setups with ZRS so far. First, Zwift’s original one:
And currently, a custom one:
I don’t think the current setup is extreme… if it IS, it’s only extreme in the A cat, which is pretty wide. Certainly not for the B cat though.
Something like this?
That was in reference to signups a couple days before the race.
The participation numbers for the Tiny race have really taken a hit. I used to see 80 to 120, and occasionally even 150 people in my category. Now, it’s just around 30 to 50. What happened to everyone?
They went outside. It’s the middle of August… we’ve never had close to 150 people in a category at this time of year (or 80-120, even)!
If we had 150 in a category, that was a super rare anomaly.
Our highest rider count ever (for the entire day, so all three zones combined) was 1636 back in Feb 2023. Our typical week is closer to about half that… (see https://zwiftinsider.com/tiny-race-stats-2024/)
Also worth noting: we’re running 5 categories now, instead of 4 like in years past. So even if overall rider count was the same, you would expect to see on average a 20% decrease in pack size.
I don’t know if ZRS is ready for the “Verify rider category before entering pen” suggestion. At least, not in the sense of moving people between races (if they get bumped between the four races). I had a racing score of 424 before the last weekend’s races. The first race dropped me to 410 so I raced the lower category where I finished midpack. It then jumped me back over 420 but I had signed up for the next two in the lower category and my score varied wildly despite finishing midpack, eventually settling at 428 at the end of the four races.
The next day, the point swing was a LOT less wild with the exception of the first race where I was dropped from 424 to 399 and then the next three races were slight points bumps only getting me to 404. Had I followed the points swings, my racing would have been all over the place but once the score settled, it made a LOT more sense where I belonged.
Yeah, good observation. I get that would be the downside of it working that way!
My big concern is for the riders who don’t have a very accurate score coming into it (haven’t been racing consistently) so they join as a C but have clear B or A power.
But yeah, it’s tough when your proper score is on the edge of the bands…
Part of the issue is also that the immediate score updates are often wildly wrong, particularly for lower finishers when high-scored people DNF. Races 2-4 also seem to use the same starting score as race 1 for their immediate updates. They all update correctly 12 hours later but that doesn’t help here.
On a series like a couple of weeks ago with alternating sprints and climbs, you wouldn’t want people being forced to change categories between races when the series is already designed to level that out.
If it was 4 races in one event, or if the check was against the overnight score, it would be a great addition to this and all other zwift races. But it’s really hard to get an algorithm working between such close races that wouldn’t also have false positives.
Hi @Eric Schlange thanks for sharing the stats. Fascinating. Very appreciative that Zone 3 exists. Wouldn’t be able to take part without it. It’s a good mix of Oceania and North America, mostly.
I would like to see a greater variety of goals.
For example.
1 The Clyde Kicker finish at the Glasgow Crit Circuit.
2 A Village Sprint finish and Downhill finish at Castle Crit.
3 Downhill finish at Grand Central Circuit.
4 lead-in finish at Lutece Express.
5 Libby Hill finish at Richmond 2015 Worlds Reverse
6 Second Esse finish for Loop de Loop
7 Downhill finish at Queen’s Highway
8 KoM finish at Tour of Tewit Well
More non-flat finishes anyway!
If possible, can you get the top3 avg wkg numbers per zone 1,2,3? Or something like that. I get the feeling that zone2 requires a little higher wkg to get podium position but I might be wrong 👍
I love the Tiny Races and appreciate the work you do to put these on! On a personal note….it seems that the distances are longer than when the series was first started and now tend to really max out the available time. I really like the shortest routes that become an all out sprint from start to finish….but that’s just me. I also like the variety provided each week. Thanks again for making them available…
I did the Sunday morning (AEST), block of 4 tiny races.Signed up to all 4 in C grade, it seemed to me looking at the racing scores I was upgraded to B after the second race, I was allowed to race C in the 3rd, but the 4th I had no join event and had to race B. I think it worked ok but would definitely be better if all were in C grade. My legs were blown on the 4th and I was out the back, hence the cat’s working really well I think.