Zwift’s recent race calendar cleanup announcement has racers discussing the merits of various Zwift races. Particularly, some riders are saying they prefer the smaller races which Zwift’s cleanup may remove from the calendar.

(I should mention that James Bailey, the former community race organizer and now Zwift employee tasked with cleaning up the calendar, has stated that this won’t spell the end of small races, women-only races, or events in less popular timezones. So I don’t think it’s time to pull out the pitchforks just yet.)

Some riders are concerned this will result in overly large race fields with 100 or even 200 riders per category. One reader commented:

When we speak of quality Eric, most of your Zwift Insider Tiny races are boring. Sorry to tell you that. You have large attendance of 100+ racers in each category but 95% of races end up with mass sprint no matter what route is chosen. I participate in them just to improve my race score but from the entertainment point of view, experience and overall satisfaction these races highly predictable…

A fair comment, and I don’t disagree that the Tiny Races have large turnouts, are fairly predictable, and often end in mass sprints. But this got me thinking: is this a bad thing? And if it’s so bad, why do so many people keep showing up?

Put another way: is a race’s predictability inversely proportional to its quality? Are mass sprint finishes indicative of “lower-quality” events?

What are the factors which make a race higher or lower quality, and who determines what quality is?

I think it’s worth discussing.

Viewing vs Riding

On the bricks of Champs-Elysees

Let’s get this out of the way early: the races I like to watch aren’t the races I like race. And I’m quite sure I’m not alone in this.

I love watching action-packed, drama-filled IRL bike races. Paris-Roubaix is a favorite, while flat TdF days are low on my list unless we’ve got crosswinds. I want to see riders attacking, making mistakes, getting shelled off the back then battling to stay in contention. All of this makes for good viewing, as any cycling fan will tell you.

But usually, when I take on a Zwift race, I’m looking for something predictable on a flat or rolling course. That’s because generally when I race, I want to have a shot at winning. And I’m not winning any climbing races as a B rider with an FTP of ~3.75 W/kg.

For the purposes of this post, I’m looking at race quality through the eyes of the racer – not the viewer.

Quality Factors

Is race quality something we can measure objectively before a race begins? ZwiftPower tried to do this with their race quality rankings years ago (read more about it here). But ZwiftPower’s ranking is just a measure of the rankings of riders who are signed up for the race. It certainly isn’t a standalone number that says “this is the right (or wrong) race for me.”

A more holistic view of race quality, from a rider’s point of view, probably requires a combination of several discrete factors including:

Participants: “the racers make the race”, as they say. How many riders are in the race, and what are their abilities?
Animation: how excitable are the riders? How steady is the pace? Course choice is a factor here, but so are the riders themselves.
Predictability: does everyone know how the race is going to unfold before it even happens? Or is there a level of mystery heading into the event?
Course: how long is the race, what climbs are involved, and how does the finish lay out?
Race Type: is it an iTT? A scratch race? A team points race?
Schedule: when does the race happen?
Powerups: a Zwift-specific factor, for sure, but some racers prefer races with powerups, while others prefer no powerups.

It’s quite possible that I missed some factors above (feel free to comment below). But the bigger point I want to make is that even if you could measure every race based on these factors and display some sort of “race rating” graph for every event, it wouldn’t give us an objective score that says, “This is a high-quality race” or “This is a low-quality race” for everyone.

I would submit that racers look at the factors above, even subconsciously, to determine which race they will take on. And it’s that subjective determination which largely defines race quality for the individual.

If I’m looking for a hard effort where I can work on my VO2 power and be pushed until I blow up, I may race in Yorkshire, even though I never come close to winning in Harrogate. But if it meets my goals, I’ll pick it and be happy with the experience.

Or maybe I’ve grown tired of getting my butt kicked around the ZRL B1 Division, so I head to Crit City for a 12-lapper that’s right in my wheelhouse. I know I’ll probably finish in the top 10, and will almost always have a shot at the win.

What one rider chooses on one day may not match what they’d choose on another day. So of course, the preferences of two different riders won’t always match either.

Race quality is subjective, not objective.

So how can Zwift measure it, in hopes of improving the Zwift race experience for all?

Descending Libby Hill

Quantifying Quality

If Zwift can’t measure quality based on the same factors racers use to subjectively decide which races to join, how can they measure quality in any objective way? Here are a few ideas:

Post-Race Surveys: ask racers to rate their experience after the race. (Zwift already does this, but rarely and randomly.)
Returning Riders: while a high number of signups or finishers doesn’t mean everyone enjoyed the race, when those riders return to the same event again, that’s a good sign.
Starter and Finisher Counts: how many racers started the race, and how many finished? Interestingly, Zwift says “finishing %s are one of our clearest measures of race quality.”

In an increasingly AI-driven world, it’s worth pondering how these three measurements could be combined to create a “Satisfaction Score” for repeating race events on Zwift. It would work for group rides as well. This score could in turn be used to automatically increase or decrease event exposure on the public calendar.

Wrapping It Up

In conclusion, I believe race quality is subjectively determined by each racer, but it can be objectively measured post-race by Zwift. I’d love to see Zwift start to use those objective measurements to help racers find the best races for them on any given day.

Your Thoughts

This article was meant to be a brain dump, not a manifesto. I’d love to read your thoughts on this topic, so please share below.