Note: for the latest on Zwift Racing Score, see What’s happening with Zwift Racing Score?
In October 2024, Zwift Racing Score (ZRS) officially graduated from testing in Zwift Labs to become the de facto race categorization method on the platform. Today, the vast majority (over 85%) of public races on Zwift are using ZRS.
It was a big change in the world of Zwift racing, and it came after no small amount of development work by Zwift’s team plus months of testing and feedback from the Zwift racing community. We’ve dived down the ZRS rabbit hole in other posts, but below you’ll find straightforward info covering key Racing Score topics.
Zwift Racing Score: How It Works
Every rider is given a ZRS between 0-1000 (higher is better). But where does that score come from, how does it change over time, and how is it used in Zwift races?
Your Starting Score
Your score begins with what Zwift calls a seed score, and it is based on your weight and historic power numbers. Zwift looks at all your activities in the past 90 days and pulls out your best 5-minute power performance, plugging that value into their super-secret algorithm to produce a seed score between 0 and 1000. It’s this initial score that determines which group(s) you can join for your first scored race.
Your Race Category
While not all races on Zwift use ZRS for categorization, starting October 7, 2024, almost all races do. Race organizers can define custom score bands to break races into competitive groups, and Zwift continues to experiment with different band ranges in their own events.
As an example, here are the ranges used when ZRS was first released in October:
- 690-1000
- 520-690
- 350-520
- 180-350
- 1-180
See below for more on joining your first scored race.
Updating Your Score
As you race, your score changes based on where you finished (podium? last place? somewhere in between?), as well as who you raced against (strength of field).
This makes intuitive sense, of course. A high finish against strong competitors should increase your score more than the same finish against weaker competitors.
Your score can also change (increase) if you set a new 5-minute power best. And it may decrease if your current power best ages out of that 90-day window and gets replaced by a lower number.
If your score decreases, take heart: your personal score is just a moving probability of predicted race results. As it adjusts based on race results and 90-day power numbers, you should find yourself placed in groups that deliver a challenging yet enjoyable race experience.
Finding Your Racing Score
Your ZRS can be found in the Zwift Companion app and on your profile at zwift.com:
It is also visible on your ZwiftPower.com profile page.
Joining a Scored Race
You don’t need to know your racing score to race on Zwift, though! When browsing races in Companion or at zwift.com/events, Zwift highlights the “Recommended for you” group based on your current score:
Click the (+) sign to join your desired group. You can join a higher-scored group, but you cannot join a group below your current Racing Score.
Note: you must have at least one 10+ minute activity recorded in the last 90 days in order to race a scored event. This helps ensure an accurate seed score, thus improving Zwift’s choice of category for your first race.
Understanding Race Results
When you finish a scored race, your ZRS is recalculated so it remains an accurate reflection of your current abilities.
You can see your new ZRS under your profile as shown above, and it is also saved in the event results in Zwift Companion and on zwift.com under your activity details or the event’s homepage:
Race results show your new Racing Score, and an arrow indicating whether it increased or decreased as a result of this race:
A down arrow with a line below means you’ve hit your floor score – the lowest possible score you can reach based on your 90-day power numbers.
An up arrow with a line below means you set a new power best, which increased your seed (and thus your floor) score. This arrow only shows when your new floor exceeds what you would have otherwise had as a result.
What’s Next?
While ZRS is officially live, Zwift staff continues to monitor race participation and rider feedback in order to continue improving the racing score algorithm.
If you’re curious about what Zwift is working on regarding ZRS, or want to see the details of its evolution, visit our What’s happening with Zwift Racing Score? page.
Concluding Thoughts
The implementation of Zwift Racing Score is the most game-changing update to Zwift racing that we’ve seen in recent years. Now, instead of categories being based purely on riders’ power numbers, they are results-based, taking racing smarts and other factors into account.
ZRS is the results-based categorization system Zwift racers have been requesting for a long time. While it may not be perfect in the eyes of some, it’s vastly improved in the eyes of most, including myself. Here are some of the ways the new ZRS system beats the old category system, in my opinion:
- Harder to sandbag: in the old system, riders in every category below A could simply sit at the top of their category, winning race after race, without ever being upgraded. That happens much less with ZRS, thanks in part to podium bonus points being added to scores.
- Custom categories: race organizers can define custom category bands much more easily with ZRS than with the old system’s zMAP and zFTP metrics. This encourages events that group riders in fresh ways, mixing up who is at the top and bottom of each category.
- Immediate feedback: your updated Racing Score provides immediate feedback as to whether you placed above or below your expected position in each race. This is a new, immediate feedback metric that wasn’t available with the old system.
- A global KPI: While ZwiftPower’s Rider Rankings work decently well, the metric is flawed in various ways and not available on Zwift Companion or zwift.com. Zwift Racing Score is the closest we’ve ever come to a simple Key Performance Indicator that can be used to assess a rider’s overall racing ability.
There are improvements still to be made to the new ZRS system, of course. I’m particularly looking forward to Zwift incorporating more historical data into riders’ seed scores to help prevent them from blowing up lower-category races when they return to Zwift racing after taking a few months off.
But for now, I’ll take the much-improved system over the old one any day. Ride on!
Questions or Comments?
If you still have questions about Zwift Racing Score, you may also want to visit Zwift’s support page on Racing Score.
If that doesn’t answer your questions, or you have feedback on your race experience in scored events, please share below!
I love the racing score, I dont love the new categories.
I think the categories are too wide and that 5 is too few. Probably the bottom part is ok in size and the top end needs to be that large to have enough racers. My problem is with the 690-350 span just having two categories. I think thats way too few. This is where the majority of riders are, so population is not gonna be any problems, but the problem is the huge spread of riders in the two categories. The step from 180-350 into 350-520 is much more brutal then the step from D to C. Same from 350-520 up to 350-690.
There for sure is enough riders in 690-350 to support 3-4 groups in that span.
…and how will this work in ZRL. Are they still using Zwiftpower categories?
They are still using CE at least until next year but i think plan to move to ZRS when zwift ask them. probably for the 25/26 season seems to be the suggestion unless it sneaks in for round 3
yes for now ZRL have said they will use Zwiftpower categories. What I’m not clear on is wether ZRL results will affect the ZRS?
My 5 min power is a B category and everything else is C. I was a B in Zwift and a C in Zwiftpower. Today I got moved up to B without even riding. Are they pulling that data differently too. Zwiftpower didn’t used to count Tiny Races and now they are?
zwift power category has always been the same as your zwift. Tiny races have always counted too
I imagine it will? Unless non-Zwift Race Score races don’t count, but I don’t think that’s the case. Any race or power effort should affect your score; that is how I read between the lines at least.
But with ZRL, the rider first across the finish line is not necessarily the winner of the race. Does Racing Score support points races, and will ZRL results get updated with this in mind?
Not holding my breath on either, to be honest.
Placement shouldn’t affect your ZRS for this season. When ZRL starts using ZRS,then placement MIGHT count. That’s up to the powers that make those decisions.
ZRL results will definitely affect someone’s ZRS if they do a new 30-secs or 10-mins 90-day PB, but I’m not sure if/how someone’s overall performance in a ZRL race would affect it (my guess is that currently it wouldn’t, but that’s just a guess)
They will. At least any change in power. My ZRS was 99 and went to 113 after completing a ZRL event because my 10 second AND 5 Minute maximum power went up. It’s also increased after another ZRL brutal effort. I don’t expect placement in a ZRL event to affect the score.
I agree that 170 score-width pens are too broad and I can also tell you that, as rough at those pens are, the 310 score-width pen that is 690-1000 is even rougher. From a code perspective, Zwift is tied to 5 pen categories per race but a solution that was done during ZRS beta was HIGH and LOW versions of races (so effectively 10 pens per race). I’m in favor of fewer races available at each time of day, more pens (i.e. a high and low version of the races we do have each hour). That combination could solve the population issue per race while also keeping each race more competitive from top to bottom of the pen.
I totally agree I’ve been racing 2 or 3 times a week and have been bouncing between categories (what used to be C and D). Cat 180-350 where I am somewhat competitive and I’ve even won a race to category 350-520 where I am dropped by the bunch usually around half race distance (depending on the route profile). After one or two races in each category my score has changed up or down to switch me again. I think the 350-520 category is way too tough this is the equivalent of the old 3.2w/kg and I frequently see many riders finishing with 3.8, 3.9 or even 4w/kg at the end so I’m pretty sure a lot of these could go up to the next cat. I get that Zwift are trying to evenly split the riders but from what I am seeing someone with a racing score around 400 or less is not going to stand a chance in the 350-520 cat
I was hoping that the RS numbers did not fall during many riders ride outside. Means that every autumn, we will see all the stronger riders in lower cats again, and that it will take time to adjust, probably just before spring, when they are going out again – Total groundhog day 😀
I do like the change and think its a step in the right direction 😀
The seed will prevent this. I’d argue it’s too restrictive at the moment. We had some riders come inside but not be able to sign up for the general community race due to the 3x10min.
I thought that part of the sign-up restriction will only be an issue in this “first time ever getting a seed score” situation though, and in future just coming in from a summer of outdoor riding having previously had a ZRS value would mean they’d just start with their most recent ‘Floor Score’ wouldn’t it? I wouldn’t imagine they’d have to do 3 new 10-mins+ activities, if so that’d be a big flaw. A slightly ‘decayed’ ZRS which stops decaying after X amount of inactivity on Zwift would seem to be a far more sensible/workable approach to ensuring there isn’t the constant category YoYo-ing that we get with the A-E cat system each autumn
There is a floor to how far your RS can drop based on your seed (initial) value.
Not if you dont cycle at all during 90 days.
I would love to see the points we get after a race explained in more detail. So far it looks like only the place in your race group matters and not how well you did compaired the ZRS score of the riders you raced with. And how much the podium points are.
No, your placing AND the ZRS scores of you and your competition matter. It’s just like scoring with other online games and even IRL events. If you’re a lower scored player and you beat a player with a much higher score, you get more points and they drop further. If you beat someone closer-matched, you don’t gain as many points.
Tjhat is why we need to see the points we get after a race and not everyones new total score.
But if you look at all the races. 80% of them there is a clear cut around the 50% of the riders who signed up where people above go up and people below go down.
If it would count for the ZRS scores also that should not be possible.
I agree, seeing the number of points would be nice.
I don’t think I understand your second paragraph. Isn’t that how it should work?
If you look at race result you will see a cut off at around 50% of the entries.
So if 22 people would have joined a race the cut off is around the 11th place. All higher get a rise in the ZRS score, and all below go down in their ZRS score.
You see that in a lot of race results.
But what if someone in that race ended 14th and all others who he beat, so 15th to 22nd were a higher rank than him before the race. Should he than not get a rise also since he beat a lot of riders who are better than him. But you dont see that in the socres at all.
Thats why I asked if someone could explain the scores a bit more in details since it looks like the ZRS score of your opponents has no influence at all.
Oh, I see! Thanks for clarifying. Well I think the reasoning there is that out of the whole race, you could have done better if you raced in a different set of racers. Meaning something like, “you weren’t in the top 10, so a less-bad loss could be had in a different group” so they take away points.
I think this then lends to the notion that the range of scores per category should be narrower.
Tim, the guy behind the zracing.app site, who is working with Zwift on the ZRS score has confirmed that ZRS doesnt work with an ELO score. So what your score is and the score from the riders you lose too and beat doenst impact your score. Only your placement in the race does.
I’m not sure that’s a complete description though, Peter.
Yes, it’s not ELO.
Yes, your placement matters.
But the scores of the other finishers also matter.
It’s just not comparing the score of who you beat or lost to specifically. It’s computing an overall strength of field based on field size and rider scores.
So if you enter a race with 50 highly-ranked riders and win, your result will be higher than if you enter a race with 50 low-ranked riders and win.
Raced my first Racing Score event yesterday – Community Races at Innsbruck. Speaking as someone who as a result of the paradigm shift went from a strong C to the bottom of the 520-670 category (racing score of 527), here’s my impressions and observations so far:
–I’m hoping that the interest in racing as a result of these changes hasn’t been killed off. For a series that would usually draw 70ish riders at the time I race, there were only 20 this morning.
–Either I was grossly overseeded with my initial score, or the pen allocations need to be narrowed a fair bit in scores. Granted, I’m at the bottom of my current category as mentioned above, but routinely finishing 2-4 minutes after the leaders do isn’t going to be much fun for most. Especially in a system purported to level the playing field some so more people have a chance.
–Directly after this race, my score was immediately dropped to around 515, but then a few hours later, it was bumped back up to 525. I took 13/17, and the bottom three weren’t even trying to get a good result at all, so I only placed ahead of one person actually competing. So my questions are why did it drop and then raise up again? And for placing so poorly, why only a net drop of 2 points?
These aren’t meant as complaints by any means – getting my ass handed to me is only going to make me a stronger cyclist in the end. But not everyone is likely going to have that attitude about it. And I also realize that the only way to refine the scoring and get placed in the correct pen is to keep racing and let the algorithm do its thing, and for everyone else to do the same. I’m just hoping that the new system intended to improve on the old isn’t actually worse than what we had… Time will tell.
I totally agree that the categories are too wide. Doesnt matter what rating system is used if categories are too wide, leveling is always gonna be a brutal non desirable experience.
This would be easily fixed by narrowing down the mid categories to say 100 point spread instead of 150.
We dont have enough pens for that I am afraid. Or you should stretch the top and lowest pen a lot and I guess the people in there wont be happy about that.
With just 5 pens available making them narrower is hard.
Questioning from a standpoint of admittedly complete ignorance, couldn’t they just build and implement more pens?
That would be a great idea indeed. And I really have no idea why Zwift didnt think of that before going live with this.
There were testevents with 10 categories. Worked fine for me.
There were and there we are also a lot of races with just two people in them. Pretty lame “winning” a race against one other person with .5 w/kg less ftp than you. Small field sizes make FTP the dominant factor and not your ZRS. I would much rather compete against 100 people and not place than race 10 and get a podium.
In terms of your “Why did my score go down then up again” question: This is something that a lot of people have been annoyed by. Basically the ZRS value that you see immediately as you finish a race is NOT the ‘set in stone final value’ and some further checking and computing goes on in the background before that ‘final’ score then turns up.
If you’re familiar with the ZwiftPower site results processing then it’s kind of the equivalent of when you show up in the results with power numbers with a “green bolt” next to them which means it’s a snapshot rather than final values and it’s only really your true power values once it goes to a “blue bolt”.
With the ZRS score though I personally would far rather it not show a score at the end of the race and only show it on your profile/results etc. once it is “genuine/final”
Yeah indeed. The waiting till Zwift finalizes your score is terrible. After a race you see a score only to see that changed again 12 hours later.
It shouldn’t take 12 hours anymore. The lastest round of ZRS changes included a change so scores are updated in near real-time. So once the last rider finishes, the score should basically be finalized…
Not on your Tiny Races last Saturday. I did the 11PM Races and had to wait till next morning till my score was adjusted.
If you’re doing more than one race within that 12 hour period then they all seem to start from the same initial value. The score movement from each race seems to be the same, but the running total only adds up once the scores have been recalculated 12 hours later.
More pens with narrower score ranges will far from solve everything, the current seed score formula is garbage and you will still have racers not placed in anything close to the correct ability pen.
Under-estimated racers, often lightweight high W/Kg, might eventually get a pen promotion to racers of similar ability.
Over-estimated racers, often heavyweight high pure Watts, can only drop 15% under their rolling 90-day seed score, which may well not be enough of a drop for a single pen drop.
Agree with this as well. Two biggest issues it seems right now are the wide ranges of ability in a pen to start, and making sure racers are getting the right initial score and pen placement to begin with. How the score reacts and changes according to race results will also likely need refining.
I just figure adding and implementing more pens is the easiest aspect to address and improve the problems currently faced. All the while also figuring out how to best tweak scoring, both seed score and subsequent scores as one participates in events, as more races become available to draw data from.
This is me too. I am a stronger C & sprinter/puncheur but neither am I close to upgrading to B standard. My initial racing score is 531 and it looks to me like I am going to go from competing for Top 10 on hilly races (lower on mountains) and podiums on flat/rolling courses to getting my ass kicked and like you, coming in a few minutes down on my own, as I mostly do if I do a B race. I’m not upset by the idea of probably not going to be as competitive & towards the top end in a new category but getting dropped and riding on my own doesn’t sound fun either… lets hope it will adjust itself and I possibly see my racing score drop but I am concerned the bands do looks too wide. I would be also interested to know if ZRS will adjust for the terrain like Zwiftracing.app does.
Same here. I was on Top in C, sometimes upgrades to B, then i got after some time in the middle field there, dropped back to C 🤷♂️. I had four weeks vacation: Still B (523), higher kg and not in my best shape. Last week dropped to 518. raced the same race (Innsbruck) were dropped hard in the last lap (i would have attacked usually at the leg snapper), came in in the middle, but also got higher 523 🤔. Inreally don‘t understand 🙈
This happend to me to, pretty much exactly. What was even more frustrating was to see the mix of ‘old’ B cat riders dropped down to C whereas I’m there being the only ‘old’ C cat rider in the B race. Guys were putting out my sprint power time and again on the leg snapper. I would’ve just stopped as it was boring and annoying when I was left behind and the fun was removed, but figured I would have to finish and hope my score was reduced
Good to see attendance across cats increased now that it’s the only option.
a decent chunk of folks seem to be keen to evaluate it as a math problem. The real question is whether races are attended and challenging but fun. Looks good so far.
The attendance is likely to decline significantly by November. There has been considerable criticism directed towards ZRS, particularly within the discussions in the race pens chats.
Let’s see how things develop. Usually people who criticise things are much louder and seem to be more prevelant than people who are fine with the changes, which then doesn’t reflect the widespread opinion. All I can say, is that my recent races were quite ok. The new system is without the slightest doubt objectively a step into the right direction. Does it need tweaking? Yes, definitely. One thing to consider is also that a lot of people got a random racing score asigned that might not reflect their actual racing ability which will need some time and races to settle. This then also makes it appear that the different skill brackets are currently too wide (although I also would wish that there would be more than there currently are). But it should get rid of many issues of the old W/kgs system especially in connection with weight doping in whatever direction.
It is not just a little adjustment that is needed; it requires a complete overhaul. Even those who worked on the ZRA have admitted that the new ZRS has plenty of issues, including problems with the seed formula, a lack of ELO ranking, and a misunderstanding of what “fair racing” really means. The chances that a critical mass of racers will be happy is close to zero.
Can anyone comment on why KISS races in the middle of the week uses ZRS while longer races (KISS 100) on Sunday still use the old A,B,C,D system?
The organiser needs to switch them, assuming they want to. Zwift is only in control of the races it schedules.
The KISS 100 is also using the ZRS 0-1000 system now, but bear in mind that the Sunday just gone was the 6th of October and the new ZRS system went live for most events from the 7th of October, so Sunday just gone may well have been their last “A-D Cat” event.
Any idea what happened to them updating the ZwiftPower site to incorporate ZRS properly?
Currently it still only shows events with the classic A-E coloured pens, no way to see if an event is using its own ZRS Bandings (such as the Cyclopathon events for example), it also doesn’t show ZRS of riders in the list of signed up riders for an event (again just the A-E icons) making it harder to gauge who is likely to be the toughest competition for you to keep an eye on etc. when you start racing, nor does it show ZRS changes in the results still just the old defunct 600-0 ranking thing
Have they fully given up on actively working on and updating ZP now? Seems like it has been treading water ever since Zwift officially took it over 😭
Hopefully they’ll start applying this and scoring to WTRL TTT. The same teams and riders dominate frappe and latte and do so just within the legacy category limits.
WTRL is in charge of their own categorization system for Thursday TTTs. So it would be up to them to make changes as needed…
I really want to be a fan of RS, but so far I have been avoiding those races, after trying a few times. Just look at riders stats on races on ZwiftPower in the 350-520 group; in one case I see ppl with 2.8wkg 20min power against ppl with 3.9wkg 20min in next case 3.0wkg vs 4.2wkg and so on. How that is more fair, I don’t know. Then I actually prefer a few sandbaggers, because at least they to not rip everything apart until the final sprint.
The algorithm doesn’t include 20 minutes because not many people have recorded a true 20 minute maximum effort, and people who haven’t would get an artificially low score.
Of course the exact same argument applies to 10 minutes and more so to 30 seconds.
So, currently races show up.with cat A B C D and E..What will appear for races using this score? A Max score# instead of the letter category?
Never mind. I see the screenshot showing that each level displays the range.
Love it, best upgrade Zwift has done in a long time!
I’m a C… Had a little Stint as a B (caught me by surprise), because I got Stupid during a Race, and I guess I had a good day 😝… I then had to Suffer through 90days of the B’s, until my numbers where readjusted, and I was moved down to the C’s again (Yipeee). Now with the new Racing Score (317), I’m what I would guess is back down in the D’s (even Happier!)… I guess I’ll just have to wait to see how things play out? 🤔🚴🚴🚴
Eric, I applaud Zwift’s efforts to come up with a more fair ranking system, but I think their algorithm needs more work. I’m 70 years old. You are a lot younger. Zwift thinks I’m a CAT B with a 565 rating. According to one of your graphics above, you are a CAT B with a 568 rating. You typically win or attain a Top10 in most of your races. I’m lucky to finish in the Top 50% of any CAT B race I enter. Your w/kg power numbers put you near the top of the B category. Mine place me at a mid-to-low level B. I’ve got to question the validity of a ranking algorithm that says I can compete fairly against anyone with a 565 ZRS and believe that it probably needs to add additional power metrics to improve its accuracy. I find it hard to believe that race performance can be calculated based only on 30 second and 10 minute power levels, despite what Zwift says. My current 30 second effort followed a relatively easy ride and my 10 minute effort was a max effort that again followed a relatively easy period of riding. Neither reflect what I could do in a CAT B or 550-680 ZRS race throughout the start, mid-race, and sprint finish segments of a competitive race. In theory, if I find a race with a ZRS rating of 565 near the top of the range, I should be very competitive. But I’ve “raced” in pacebot group ride sprint segments with lots of riders 100 ZRS points or lower than me and can’t compete. So for the time being, I’m stuck at what I believe is an unfairly high ZRS and will just have to wait 90 days to see if it resets to something more reasonable. I’d like to think that as more race results data is inputted, it will sort itself out over time. But with what may be a large number of both inaccurately high and low ZRS initial ratings, it could take awhile if it can happen at all, without an improved algorithm.
My power numbers don’t put me near the top of B… not by a long shot. See my ZP profile at https://zwiftpower.com/profile.php?z=58160
I can finish top 10 in flat races, but I get dropped on climbs.
Curious what your power numbers look like… link up your ZP profile maybe?
But really, none of that matters now. I understand your complaint, but it’s important to realize that you’re comparing an old category system based on completely different parameters to the new one, and saying it’s not working well because people are being categorized differently!
Here’s the thing: if people were being categorized the same as before, THAT would tell us the new system isn’t working well. The fact that it’s different… that’s a good sign. Or at least, it MIGHT be a good sign.
If your seed score has placed you too high, my guess is a few races will see your score drop. If your seed score is 565, that means it can drop all the way to 480 (15% below your seed score).
Unlike the old category system, which let people just sit at the top of the category and win repeatedly, your score should get more accurate the more you race. Remember, it’s just a race predictor. So it’ll go up or down based on how you do.
I’ve got to be honest with you—this is not going to work. Once he drops to 480 after a bunch of rough races he will bounce back super quickly. So, for like 90% of his racing time he will have an extremely bad race experience. Is this what Zwift want?
Based on this seed score, I will never drop below this category? 612 score.
I am a C category rider, 92kg good sprint. I am racing monsters and get dropped even on the flat. The system isn`t right and if I can only drop 15% this would only put me down to 521 so I will never get demoted?
Nice, clear explanation. Looking forward to results based racing at long last! For context I have been a top end, part time B racer for a long time but welcome the gradual input of results over arbitrary power only. One of the great things about Zwift – you can always find a race within a race…
My experience has been,since ZRS became a thing,that I am no longer enjoying the races. Why are the C category races (now 350 to 520) so heavily populated with riders still showing as B riders? A little research shows that a good number of them were doing pretty well in B races too So,clearly, the racing score system has not brought them down into a category where they deserve to be. Whatever time and money has been thrown at racing score, it’s really not a resounding success. Not just my thoughts, but those of many others on here.
Tried my 1st ZRS race on Monday in the 180-350 pen with a score of 270ish. I’m a heavyweight cat D so high pure watts but not fantastic WKg, generally place well in cat D races, the pen was mostly cat C with a couple of Bs and me. I got absolutely battered, could just about hold wheels on the flat and blown out of the back on the 1st hill and came in close to the back. Looking at the other scores probably about right but ZRS dropped about 3 points. I don’t fundamentally disagree with the principle but it doesn’t feel like the point reallocation is quite right and there is no clarity on how it’s worked out.
A quick reset of points is all when and good but if it is seemingly arbitrary then people won’t have confidence in it.
Although, it’s a different system it ends up with the same issues. What was effectively D C B are still quite tightly banded then hits the top band with a huge variable. 680-1000 is a big jump!! And it still doesn’t stop the blatant cheating of the higher banded riders tearing up the races. There are far too few riders in the races I can do, and when I do they are 9/10 full of riders who should be pro. (Cheats) so I’m slowly falling out of love with the platform. For me, the algorithm is just pointless, have a simple scoring system win, 10 points, 2nd 8pts and so on. If there are less than 10 riders, adjust the points awarded. It’s not needed to be complicated.
Is it your best 10 min and 30s power in a race or in any activity from the last 90 days that will determine your seed score?
Any activity. But now that it’s gone live it’s the score floor that is important, which is 15% lower than seed score.
So two identical riders can have scores 15% different if one did their power records a week ago and the other does them now.
I’m presently a Cat B. I’ve been watching my performance decline quite significantly over the last few months. I’m 68 years old now and I’m expecting to have a knee replacement operation in November this year. My return to Zwift will be gradual, but if it is within 90 days, I expect I’ll be racing in races that my previous performances suggest I am still at the same ZRS category. This is going to be quite a while before I drop down to a more realistic lower level during my recovery period. Can Zwift seed me to a lower level a bit quicker so that I can compete with those I can conveniently race with after my operation until I start to improve, or do I have wait 90 days to see any changes?
As of today, I’m not a fan. As a mid-bottom cat C rider I compete with cat B in third category. It steals the joy of racing.
I understand it’s early days and hope it will get better with time.
It’s simply shocking currently, 3 races completed and dropped within 5 mins. Even lapped on Duchy! rode home alone each time….
Cannot fathom how I can compete in any way with people who have 5 min power well beyond my 1 min power👎what do I do join the weight dopers or get a kickr snap😕
It’s been a bit humbling. Just done a D race, averaged 2.8W/Kg and a years best 20 minute power which would have got me upgraded under the old system. Got dropped from the front group after 5k. The top 7 riders all averaged over 3W/Kg
same happened to me twice on the community racing series and i’m sure when i try again today it will be the same, my racing score is falling though so i’ll hopefully drop into E pen, now how is that motivational? i used to enjoy racing with the D’s and occasionally when my 65 year old body was feeling good i would podium, now its just like doing a ITT not fun! 🙁
I like the concept but think they’ve pulled the trigger too early in this. God knows why they take into account riders weight, 60s and 600s power. The average race is probably 30 mins so why not make the assessment based on 20 and 30 mins power outputs? Or at the very least include longer ride outputs in the measurement.
Its like Park Run creating grouped 10km running events by measuring someone’s 100 metre and 400 metre sprints
Why does my Profile page in Zwift look so different from yours? I don’t see that at all? And my Companion app is smooshed and doesn’t show my score? How do you get yours larger like that?
Mine too… No ZRS anywhere to be seen. I can view it on Zwift Power, but not the change up or down from my last race.
The problem you may be having is that your devices (Android or Apple) default font size is set too big. Try changing your devices font size in your settings. I made my font size smaller on my Android phone…boom, I can now see my ZRS!
I’d like to fewer zones vs more. Alternatively, they could reduce the number of races during non peak season. What was the “A” category often has fewer than 5 folks riding.
Raced in a C group ZRS race yesterday — 18 riders mostly around similar power etc some B and C pace riders. But 1 guy was an A pacer with 270 watts = 5.0 w/kg so he rode away. That caused the race to split as half of us tried chasing him for a lap. So over time that A guy will score up into the next range. The thing I’m not liking is that half the people were not in ZP and hence no ZP results and no zwiftracing.app scores for those riders. I wish that ALL racers were just auto put into ZP so that we can see everyone’s power ranges and if they don’t wear a HR monitor – fine DQ them for final results in ZP, but at least count the rankings of them in ZP and zwiftracing.app score.
I think the move is in the right direction but compared to the vELO system on zwiftracing it’s still fairly unsophisticated. It takes some account for terrain, power profiles, effort levels etc and seems so much more nuanced. I’ve found it to be an excellent predictor of some of the races I’ve been a part of. It might be too complex for Zwift’s taste but it could always just display the numbers to the masses but let the more serious racers pour over the data.
Time will tell, but I honestly believe it should be better in the long run. Of note, my ZRS hardly moves regardless of the results of races so I am not sure how effective it is currently. However my ZRapp score does move around a bit so I believe currently ZRapp is still the gold standard for this system that zwift should attempt to emulate.
i wish that Zwift could select riders to put in the race based on their ZRS at the point the race started. Ie the algorithm selects optimum number of categories looking at numbers who enter the event, right up to start. Then groups riders into the best number of categories to balance evenly matched riders with categories. Eg 45 riders enter by time of race start, algorithm groups riders into 3 categories of 15 riders etc.
I joined the last World Series community race with a seed score of nearly 300 or something. And it was brutally unfair. My W/KG were around 2.5 and the winner had nearly 4.
After reading this I suspect the issue was that these may have been summer outdoor riders with a low score, but I do hope they get this under control.
It’s a good step forward, but now it needs to be dialed in well. It seems like they’re on it already.
The formula is heavily weighted towards pure Watts, so relatively high W/Kg riders get a lower score and depend on race results to put them in a higher pen, where they should have been put from the seeding.
High pure Watts combined with heavier weight are stuck between a rock and a hard place with the formula, they are given a higher seeding score and even with a consistant series of bad finishing positions, their Racing Score can currently only drop 15% under their power seed score. If you’re lucky in such a situation, you might get a single pen demotion, but it’s not guaranteed.
There are returning riders who have spent summer outdoors and limiting the seed score to power data in a rolling 90 window doesn’t help, but it isn’t the whole issue.
The formula is garbage.
what happens if you have signed up for an event and when the event starts you are no longer allowed for this category, because your score increased in the meantime?
For now, just like with Category Enforcement, pen restriction is at sign up and not trying to enter the race pen start.
Allegedly, this is supposed to change Zoon.
I am delighted that Zwift has introduced the racing score. In my feedback on Zwift surveys over the 6 years I have been a member I have been positive, with the caveat that the A to D categorisation seemed to be unrealistic, based on my experience. I have always ended up in the low end of the B category but in racing I ended up “TT-ing” the rides on my own a lot of the time, which is a lonely experience 🙂
When I have been bothered, I have reduced my riding for periods of up to 90 days so that I drop back into a C category rider. This rarely lasts long after I ride an event enthusiastically, when I find myself back at the low end of B again.
Having said that, I agree in principle with categorised events.
From my brief experience to date, the racing score seems more closely tuned to my ability compared to other riders. Thank you Zwift for trying a modified approach.
There’s absolutely no need for 30secs power to be part of the seed score.
If you are with the lead group within the final 1Km and contest the sprint, finishing place success should simply increase your Racing Score and repeated success should result in a pen promotion.
Unless there is a fundamental change to the garbage seed score formula, the current limit of a series of bad results only allowing Racing Score to drop 15% below seed score isn’t enough and perhaps should even go altogether.
Unless your power data of the last rolling 90 days (too short a window IMO, should go back to last winter) decreases, only when you finish races an your score drop.
There’s a limit to how much your score can drop per race and there’s no reason the drop amount formula cannot be tweaked (eg. less points drop for doing badly in a very short race like a Zwift Insider Tiny Race).
Limit finish position Racing Score reductions to final ~40% of finishers; increases to top ~40%; no change for middle ~20% finishers; bonus for at least the podium (and more in bigger field races).
The implementation of ZRS just seems totally half baked. Event labels still include letter categories. Riders’ colors in races still correspond to the letter categories. Zwift Power results still show letter categories and don’t correspond to the ZRS pens. It’s all a jumble.
Besides that mess, the category boundaries seem much wider than the letter categories, even with one more pen.
I hope this stuff gets ironed out long term but it seems like Zwift is shooting themselves in the foot by making such a mess with their most loyal users.
This should be interesting. When they went to Category Enforcement a few years ago, I thought that they would base it upon course type. They didn’t. Now they went to a scoring system and still haven’t. They need to. It shouldn’t be that hard. Why can’t they do A for flat races and B for climbing races for example for a bigger rider? I hate to be stuck doing only a certain type of race all the time. I like being pushed to get better and don’t really care about winning, but when you have to do 5+ W/kg for greater than 5 minutes or get dropped up the bigger climbs it kind of sucks when there are only a handful of people in the race to begin with. That’s the problem. You’re dropped and nobody is coming to ride with. I think mass category races are the best because there are always groups behind to push you to get better.
Is there a way to see your score history or once it’s updated after a race the previous scores are gone?
Currently no way, although Zwift says they’re working on it.
One thing you CAN do is look at your Racing Score result in past races, as that is your actual new score, not just a single race “result”. So basically each race’s score is a data point on your Racing Score history chart.
Hi Eric, I’ve looked at a few results and believe that Pen E is completely wrong unless someone was running a beginners race event. Pen E imo should be 0-299 not 0-180. 180 is too low. I organise the Kiwi Crew Rides and have asked ZHQ to adjust the categories to the following for my events.
Pen E 0-299
Pen D 300-450
Pen C 451-574
Pen B 575- 699
Pen A 700-1000
I actually think Zwift needs to add 2 more pens to split A into 700-849 and 850+ and to split B’s in half which would actually lower the C pen slightly to around 550 so organisers have the ability to merge or separate them depending on how large their events are.
Hi @Eric Schlange – I am around 660 zwift score, would I improve my score by racing up in the 690+ category and probably coming at best mid-pack near the bottom or racing in the 520-690 and possibly getting on the podium?
Not sure! Depends on the actual field you race against, and exactly where you place in that field. I don’t think I can give you an answer that will always be right…
Only way your score would likely increase is by producing new best 30sec and/or 10min numbers in the top pen, increasing your seed score.
Essentially, you need to place in top half of finishers to get a Racing Score increase from where you finished, with a bonus for podium, outside of a seed score increase. Those finishing in lower half get a score decrease, if they don’t have a Racing Score that is 15% below their seed score.
I would think they could change or update races based upon number of participants at or around certain scores. So if there are ten racers with 185 Zrl and twenty with 325 zrl then these could break into two separate groups.
Still need to sort weight doping
how does it actually work? at least with the old system we knew what was behind, but now? why can’t we simply see the calculation behind so we know what is going on? the article has big words but no substance for me!
Any word if ZwiftPower will start to post these power numbers as far as 10 min power and 30 second power? Would be nice to see them posted somewhere.
How do I find a ZRS enabled race? On Oct 16, if I go to the Companion app, every race I see is still using A-D. What a mess.
If I filter the Companion list to just show races, most of what I see say 0-1000, not A-B-C-D.
But you can further filter that list to just show racing score, by deselecting ABCDE in your filters.
Racing score is ok for lower scored riders but just as with categories higher scores equals cat b and a riders only racing a few events and killing higher cat c riders just like the weight dopers. Nothing changed, you do to game the score is not race as often. It’s never going to be fair unless you race in real life.
The new C category is too broad and the D category is too low. I rarely see people racing B. They really need to take a hard look at this. It’s ridiculous how bad/unfair it is.
I’m coming back to this article because I am still seeing lots of ‘B Cats’ in the 350 to 520 ZRS dominating these races and we have a clear two-tier race formed, usually at the first climb, where the lead group is overwhelmingly B riders and the second group C riders. The races I have done with my ZRS score of 510 to 515 have all been dominated by B riders with ZRS much lower than mine and 20 minute power numbers at 3.5w/kg to 3.7w/kg. When you then go to Zwiftracing.app and take a look at their vELO you can see why they are dominating because not only is their pace category B, their vELO score predicts and then shows them finishing above the C’s, in my case with a current ZRS much lower than my top of the category score but massively higher vELO than me. Now, i do get the problem might be that vELO has been collecting data points for much longer, so the question for @Eric Schlange is do we know how long Zwift thinks it will be take to smooth this out and we get the system we all hoped for? I’m not complaining about coming 17th in a race vELO predicted I would come 16th in (so almost bang on as predicted) but that with a ZRS of now 515 why the riders above me mostly had lower ZRS when clearly it appears they should be higher?
For me, the best thing is that the boundaries of the categories don’t matter as much. I can compare myself with the racers who have a similar ZRS. Podium results reflect the boundaries, which are aribtrary. Getting the gold isn’t as significant to me as coming close to people with a nearby ZRS.
I’m not sure what to think of this new scoring feature. I did a Tour of Watopia event the other day, rod hard, and increased 1 point. That type of event is not even a race. Then I did a race event yesterday to get the route badge, didn’t actually race it, which should be obvious from stats, and I lost 10 points. So if anyone wants to drop a group and sandbag they just need to sign up for an event and not go hard. Seems a bit odd. I’m also noticing in the rides I’m doing, ppl are now like 3.9watts/kg which means they really increased the ceiling for ‘c’. I’ll hopefully get use to this in time. I agree that the new categories seem a bit big. But having maybe 10 categories would make the events too small.
https://www.facebook.com/photo/?fbid=10160181258056644&set=gm.2311422105869948&idorvanity=205036953175151&__cft__%5B0%5D=AZVujZz4YnmwyFik3dNWvL8a_HVDaiWJYrrgsUeXoVkPP-Y40TtyjFmx6hpkodYX5Evvz78MA9FMwAPowdOfe_nCS4AnBUF5j8aMxgVgRvHavrqMk0Z24Dcsve4-INcaoLE6kWwyjn2gaKMghgGI4Ma1&__tn__=EH-R
where have all the cat D racers gone? is this a question being asked at ZHQ? of the 17 sign ups 1 is cat B, 13 are cat C and only 3 are Cat D. This is not an isolated incident. it is every ZRS ranked race. The lead group in these cat D races are averaging 3 or even 4 W/Kg. The leads to the old money D’s getting dropped from the get go. If we wanted to ride on our own, we wouldn’t have signed up for a race. If you can’t manage to cater for races covering the complete demographic of riders you will see the old D’s disappear from races (and shortly afterward the platform). The only saving grace at the moment is ladder races and ZRL (which uses the old catergories)
Completely spot on. The Zwift Racing Score has destroyed the race experience for Cat D riders. I argued this would be the case when it was in beta, but was lambasted with “but you should want to improve your performance, and this system will help you do that” comments. I was previously a solid Cat D racer, pre-ZRS. Sometimes win, sometimes not. But usually competitive. Now I’m blown out and a minute behind the lead group within the first 3 miles of a race. I’m in the bottom 5 every time, competing against 3.5 to 4.0+ racers in every race. Not enjoyable. I’ve loved Zwift for years, but the change to the ZRS system is making me consider alternative platforms to find an enjoyable experience again.
I’m is races with half B racers and some A racers from time to time. Guys pulling 3.8 and 3.9 where I can maybe eek out a 2.6.
I race group B now following the new Zwift Racing Score change, and not at the most sociable times of day either, but always used to have numbers in the teens and often more
All if the races I’ve done in the current series have had very low numbers – sub 10 and the 2 most recent had 4
Is it because of the ZRS change (or the multitude of race routes on offer at the moment?) or have B races always seen much fewer racers than C
Be glad of some perspective, because it often means I end up time trialing to the end with no help/competition and it’s definitely not so fun as before
Anyone else experiencing this?