With the introduction of ZRS, Zwift has hit the self-destruct button – I’m out!

Racing
1

I enjoy racing. I have done so for many years on Zwift.
Racing works best when riders with comparable physical capabilities compete against each other.
Here, FTP/KG is a really good parameter. Not perfect – but quite good.

However, Zwift has chosen to abandon the FTP/KG model and has switched to an opaque points-based system (ZRS), where a rider can gain or lose large amounts of points from race to race.
Regardless of one’s opinion on the new model, it must be acknowledged that a rider’s actual physical ability does not change at the same pace as ZRS does. That is an objective fact.

By cutting the last remaining connection to reality, Zwift has ensured that every race now has far too great a disparity in participants’ physical capabilities, making CE meaningless as a concept. The super-strong riders can now participate ‘legally.’

There are many ways to ensure fair competition, and many of them are not particularly complicated. But I have a clear sense that Zwift is not at all interested in having balanced race fields.

I do not wish to be part of that, which is why I am now canceling my Zwift subscription.

5 Likes
2

You are not the first that took off and many more will follow. ZRS can be better, it should be better. But for some stupid reason, it taken more than 6 months for nearly no progress. They need to remove the current ones running ZRS change the leadership leading racing.

They need to have accountability and goals. Neither of these are documented and months are allowed to pass with no progress.

You know it’s stupid when they spend weeks doing V1, V2, and have it change to V3 5 min CS, on the 11th hour.

1 Like
3 
By cutting the last remaining connection to reality, Zwift has ensured that every race now has far too great a disparity in participants’ physical capabilities, making CE meaningless as a concept. The super-strong riders can now participate ‘legally.’

That’s not true. The seed score is calculated based on physical capabilities.
You can argue that it doesn’t represent physical capabilities as well as the old model since it relies only on 5 minute power, whereas the CE system used 5 minute power and FTP. Personally I think they should add another “seed score” based on FTP to ensure that the FTP w/kg difference isn’t extreme.

There was nothing in the old CE system that would stop a sandbagger or excellent sprinter from dominating a low category. Now at least there is something. I agree it could use improvement but the way you’re portraying it is false.

I found the old CE system to be much worse since I was at the top of my category, and I’m heavy. It was not that hard to just ride away from the group in most races unless the pack was really large.
Now with the staggered racing scores for the main events, I can get much fairer competition and it’s much harder for me to win.

4 Likes
4

it taken more than 6 months for nearly no progress.

By now it is at least 18 months not 6. They DID fire the first Team working on what we now know as ZRS in Feb 2024. Zwift just does not care, they are a dying company focused only on financial survival. This always causes the customer being lost behind the balance sheets.

5

ZRS was a necessary move since WKG is such a counter intuitive and unfair way of categorizing racers. That doesn’t mean that I am happy either though.

I have had more fun racing lately than in a long time. These days people tend to work hard in races, almost everyone. And the categorization works OK’ish. So if I get dropped, then I deserve it. I did my best (or didn’t) and that was all I could do. I don’t get toyed around with like before by cruisers. The heavier guys have an advantage for most of the race, I have a hard time keeping up with them, but nowadays I can compensate for it somewhat where there is elevation (although I have to punch everytime a descent flattens out or I will get dropped, but this isn’t ZRS, it’s the physics). And in any case, a heavier rider who keeps outclassing me should get promoted sooner or later. Before, there was no promotion. The heavies could keep winning indefinitely. So, good riddance.

BUT THERE IS STILL ONE HUGE PROBLEM WITH ZRS!

It comes from ZRS not being much of a relative measure. What happened pre-ZRS was that lighter rider weren’t allowed to compete with heavier riders, because to keep up with the heavies they had to push higher WKG’s than the heavies, which would eventually bump them to the next category without ever having a shot at a podium. Very unfair to be disqualified by the mechanics of the categorization.

What happens these days is that ZRS does not, as you say, swing wildly with big score changes. You tend to get small adjustments after participation. And if you end up in the upper half of the field, you get a small upward adjustment. This puts every “mediocre” racer into the same damned-if-you-do-damned-if-you-don’t position as light riders were in pre-ZRS. Because what will happen is you keep racing and manage to squeeze into the upper half of the field every race, and then you (mostly) get a small ZRS increase. I don’t have data on this, but it seems to me this is going to push such a rider into the next category sooner or later, still without ever even seeing the backs of the front group. And it’s THIS that ruins the system.

The score must take other participants scores into account, like in ELO type ratings, or it won’t be fair. If I beat people with higher scores than me, then it should be reflected in a bigger score increase than otherwise. If I get beaten by people with higher scores than me, then that should only result in very small score changes, sometimes, depending on the relative score differences, no change at all. A rank score must work this way, and it seems ZRS doesn’t. Score changes seems mostly to come out of position in results and with little to no differentiation between positions in the upper half.

5 Likes
6

Exactly @Andreas_Traff ! This has become more self-evident during the Tiny Races. I’m not going to win one of these because I have no sprint. Or short game. But we’ve had a coupla guys in the C pen win all four races and still have a ZRS that makes no sense.

Here’s what I posted in the Tiny Race thread:

Alrighty, another week of Tinys and another week of Sandbaggers. The guy that won all four of the Z2 (8am PDT) C pen (350-510 ZRS) Tiny Races had a 466 ZRS AFTER the races. AND he won a total of nine races since the 22nd of March. His 5 minute floor is about the same as mine but he’s got a 11.5 - fifteen second and a nearly 7 - one minute. His one minute as about the same as my fifteen second. He took fourth in a ZG25 stage with 71 finishers in Zpower.

How does this compute in ZRS? And how does it keep happening? We had the same thing a coupla months ago in the C pen with a guy who had a 394 ZRS winning 3 of 4 Tinys. I finished something like 33, 13, 12, 9 and my score went up 15 points. You would think winning races with over 40 racers would result in some serious ZRS changes.

Screenshot 2025-04-05 at 17-24-35 ZwiftPower - Benny Axt🪓 Dragons 🐲
Screenshot 2025-04-05 at 17-24-35 ZwiftPower - Benny Axt🪓 Dragons 🐲970×1294 127 KB

7

If you scroll down on those results a bit you can see why this has happened.
Before the 9 wins he’s what I’ll describe as a run of bad form.
It seems like these runs of bad form have become more common since ZRS was introduced.
It doesn’t help that winning big races gives you a smaller boost than winning small races.

1 Like
8

Zackly. And that’s why ZRS is a non-starter. We’ve seen this movie and the ending sucks.

9

I was never expecting a call from Dave Brailsford, but in the first half of '22, I was quite happy aged 48 to hit ~320W for 20mins and ~360W for 5mins while 80-83Kg, considering I had proper flu for most of October '21 that left me incredibly weak for several months.

That would have put me around 511 according to ZWIFT ZRS v3 by VirtuSlo

Yet under Category Enforcement, I was close to the A/B pen border and went over a few times that year.

I hate that Racing Score isn’t a vELO system, but even without that big positive change, it needs some serious band aids such as…
Seed Score floor becoming 100%
Bigger fields giving bigger score changes
Signups and non-finishers shouldn’t affect score changes for those that finish
A higher % of top finishers getting bonus score points
Bonus points need to raise the 100% seed score of racers for at least a few weeks as success ballast
Racing Score updates need to happen in-between Tiny Races stages

2 Likes
10

The system is flawed (more later*) but it’s only as good as the people it applies to.

Too many Zwift racers want to win or place highly in races where they ‘ride their best’.

With only 4-5 categories, most riders are still not going to be able to do this, unless the course really suits their physiology and better riders are not there. And if they bring in more categories, there aren’t enough people racing to make it worthwhile…

So people adjust the metrics to ensure they get what they want.

More people need to understand that cycling is a sport where most participants can’t win and won’t win. Often, just not getting dropped is a feat in itself, or staying in the bunch until a certain point. My coach rode pro in Europe and used to set goals like ‘top 50’ or even ‘classified finish’ in a race like L-B-L because he knew he didn’t have the physiology to contest the finale.

It can also suck to be a non-explosive rider stuck between categories, for whom one category isn’t hard enough to give you training ‘bang for your buck’, or ‘going deep’ satisfaction, but you can’t win the sprint or drop everyone, and the next category is so fast that you get dropped. So some will adjust the metrics.

*I have gone from 652 to 711 by doing a few sparsely attended races instead of doing one race a week, but making sure that the one I choose is well attended and stacked with 900+ guys, and then placing poorly. I finish with a higher w/kg, but lose points. If I do midweek crits, I can rack up the points with less effort.

The problem with ZRS is that a rider who gets top 10 in sparsely attended slower races regularly will accumulate more points than a rider who finishes in the second half of much faster races with more (and more elite) participants.

Also, it just creates a new system for the sandbaggers to find workarounds. Until these people change their mentality, no system will work, unless they find a way of forcing everyone to do a VO2 max test and making sure that everyone gives 100% in that test.

ps. Zwift itself is almost certainly manipulating systems for reasons of inclusivity. As a 78kg A cat rider, I know I am not gonna be able to drop 50kg riders on the new Arenberg forest cobbles, whereas IRL if a 50kg rider hit the cobbles at the same speed as an 80kg rider going full gas they’d probably crash…

3 Likes
11

The big zRacing series events effectively have 11 categories, and this has made a huge difference in allowing riders find races where they can be competitive, especially in the lowest categories. Not everyone will be able to compete for a win (non-explosive riders will never win anything that isn’t a time-trial, Zwift or real life; that’s just cycling), but most people can at least find a race that they can hang in.

That’s the positive. The negative is that the ZRS ratings calculation needs a complete over-haul, as it deals poorly with both large and small field sizes. It should also react more aggressively to riders who consistently finish at or near the top without putting out huge 5 minute power (raising your power-based floor seems to be the only way to quickly increase your ZRS currently). It would be great if they just ditched for vELO, but that seems unlikely, and it’s not like vELO is perfect anyway.

You’re right on the money that eliminating sand-baggers is basically impossible, but fixing the issue with large field sizes producing small ZRS bumps for top finishers should at least minimize the number of riders inadvertently competing below their true level.

1 Like
12

I think also it needs to look at how hard the race was. Races with more participants, and more participants with a higher starting score tend to be faster, and more difficult.

This should be easy, right? Average watts, watts per kilo, speed, should also factor.

If your score is 675 and you do 4.6 w/kg with 5.3w/kg for 5 minutes to finish 25/47 in a really hard race in which orange tick riders dominate the top ten with World Tour watts, you shouldn’t lose more points that you would gain coming 4/6 in a race where you and the winner average under 4w/kg.

1 Like
13

You want participants of slow tactical races to have their ZRS pushed down? I don’t think that makes sense. The major reason to use a results based rating system is to prevent sit-and-sprint riders from staying in an artificially low category that they dominate by remaining fresh and crushing everyone in the last 30 seconds.

1 Like
14

If they only ride slow, sparsely attended races, why should they have an elite score?

1 Like
15

So that they don’t wreck races for slower riders.

ZRS is primarily a system for creating fair and competitive races. If anything, ZRS should err on putting a rider in too high a category and then correcting down if they can’t compete. When a rider is in too low a category, they can destroy the race for everyone else.

4 Likes
16

The reason Zwift’s lower categories are a haven for sprinters (self-proclaimed) is that they are categorized by mid- to long-distance ability so that they can’t run away with it.
That’s just it.
If you think I’m lying, check out the early ZRS threads where they were categorized by 30-second power.
A self-proclaimed sprinter (indistinguishable from a sandbagger to others) will tell you the answer.

Also, it’s not surprising that you poor legs don’t know this, but in non-discriminatory races such as the Zwift A category and real world amateur races, it’s common for a runaway victory to be achieved by running nonchalantly.

1 Like
17

Yes, 100%.

A major issue for the lower categories is that there are a lot of new and returning riders who have ZRS scores way below their actual abilities. And, if those riders like to sit in the pack and go for the win in a sprint finish, they will continue to have a far too low ZRS for a long time, because ZRS does react aggressively enough to consistently winning (or finishing on the podium) with that style of racing.

The riders who like to break away with 5-min or longer efforts are quickly moved up, because their efforts raise their ZRS floor, but that doesn’t happen for the sit and sprint riders as you note.

And this is why it would be a doubly bad idea to give riders who place well in slow races smaller ZRS improvements. That would just feed into existing issue of under-rating riders who don’t push the pace until the final sprint.

18

Not really disagreeing with you, but I think this behavior was also an issue with both 20min power, and CE. If folks didn’t push any real effort over 5mins they would be low end cat in any of these except the one short period where Zwift highly focused on 30s power in the calculation which is not good at classifying who can actually make it to a sprint at the end of a race - which is why Zwift had to move away from it.

That said Zwift should handle returning riders better in any of these cases. If someone was “A” category power last year it’s pretty unlikely they will be D cat, so it would probably be best to include a longer date range of power than they do today (maybe also include a small degradation factor to factor in possible fitness loss). If they’ve lost significant fitness they can let their results move them down.

Also, yeah, ZRS should give more points than it does for wins, especially in big races with lots of attendance, today you can win in a huge pack so many times before being moved up, but if you win in a low attendance race you can get more points for some strange reason. No idea why this hasn’t been corrected yet.

1 Like
19

I suspect that a major hinderance to ZRS working properly as a truly results based rating system is that Zwift really doesn’t want to put a rider in an excessively high category. Although that is best for the races as a whole, it will force the over rated rider to get dropped in a few races before they settle down in an appropriate category.

And I get it. Riders who are forced to sign up for races knowing they’ll get dropped are probably the most likely riders to stop racing or quit the game entirely. I bet this is why returning riders are always assumed to have lost fitness. Surely Zwift knows that they are more likely to have mostly (or entirely) maintained fitness after a summer of riding outdoors, but I’m sure they’re more than happy to feed these riders a few wins to make them feel extra good about renewing that subscription.

That’s why I think they need to focus mostly on aggressively increasing ZRS for riders who are new or returned from an extended absence. If a rider with minimal recent history is finishing on the podium, assume that they are under-rated and boost them fast. Don’t force them to move up by increasing their power based ratings floor. If we’re relying on the power floor, then all the old power CE sandbagging (intentional or not) issues come back, which is what we see today in the lower ZRS categories.

20

Summer is dead time of riding here now for the last 5 year. it been to smoky, it like off season here now in Canada. @Matt_Skalecki