Note: for the lastest news on Zwift Racing Score, see our post What’s happening with Zwift Racing Score?
Zwift officially launched their Zwift Racing Score (ZRS) metric this week, and as one might expect, there’s been no shortage of feedback from the racing community. I’ve found myself in an interesting position, hearing feedback from the community but also conversing regularly with the team inside Zwift responsible for creating and continuing to improve ZRS.
Feedback from the community isn’t easy to summarize. On the one hand, we did an opinion poll where 57% of racers said ZRS events have been better than standard category enforced races they’ve recently joined. (21% said “about the same” and 21% said they were worse.) On the other hand, comments on ZRS-related posts here on Zwift Insider have largely been negative, with a few repeating themes:
- The range of abilities in each ZRS band is too wide (“I’m a C racing A and B riders!”)
- Heavier riders and sprinters are disadvantaged while lighter riders are advantaged with ZRS
- “My seed score is too high and racing isn’t any fun because I get dropped every time”
- Riders returning to Zwift are seeded too low due seed score being based on their 90-day power bests
Meanwhile, over at ZHQ I’ve been asking for details about how ZRS works, and why the team built it the way they did. As I’ve taken the time to hear and understand their reasons for doing things the way they have, I’ve generally agreed with the approach they’ve taken.
Both the community and ZHQ have valid points of view. But some comments from the community have simply been inaccurate – people “guessing” as to how things are working, with limited data. ZHQ, for their part, hasn’t answered or corrected many of those comments, because they are keeping some of the details of ZRS under wraps since it is a proprietary engine and they do have real competitors these days.
But I’m a big believer in how true, shared information keeps communities strong. So I asked Zwift if I could get some more detailed answers to questions about how and why ZRS works the way it does. That’s what you’ll find below. Here are my questions, and Zwift’s answers…
Did Zwift consider different approaches before deciding on ZRS? Why not just use ZwiftRacing.app’s algorithm?
Yes, we considered and even implemented several alternatives to Zwift Racing Score, but none made it past internal testing. These included using Machine Learning models to calculate seed scores, variations in score volatility, caps on score changes, and different score decay rates. We also explored automatically moving a user up or down after placing in the top or bottom y% of racers x times in their current category.
We partnered with Tim Hanson, the creator of ZR.app, to discuss our proposed scoring system, and we continue getting feedback from him and from the community about ZRS. While ZR.app has a lot of very interesting features, and is not hard to grasp by people with some experience in cycling races, we were looking for a solution that would be extremely easy to understand by total beginners, as in a rating scale that ranges from 0 to 1000 (versus a typical ELO rating whose range is not just unbounded but also subject to inflation over time).
ZRS has been a few years in the making. What took so long?
Initially, we implemented a purely results-based solution, using USAC’s points system, similar to what ZwiftPower had done. It soon became clear that our USAC-style points system had problems. Points are earned every time one participates in a race, so people who race more frequently tend to have more points than others who race more seldom, but the former are not necessarily stronger than the latter. In short, points may work great as an incentive for engagement, not so much as a criterion for pen categorization.
The project to switch from the USAC points system to ZRS started on 09/29/2023. Its central piece is a complex algorithm that has required plenty of analysis and testing. We also developed multiple tweaks during public testing (e.g. podium bonuses and UI improvements). In addition to developing and testing the scoring system, we overhauled our events logic and tooling to be flexible to the new categorization system and enhanced our race results user experience to display scores.
What are the trickier problems you are trying to solve with ZRS, and how have you gone about attacking each of them?
1: How to best categorize someone who has never raced?
We considered some Machine Learning techniques, but those models proved significantly challenging to maintain and evolve. We therefore evaluated an approach similar to that in ZwiftRacing.app, having worked with Tim Hanson himself, who has been immensely supportive in our quest.
Ultimately ZRS predictions are not too far apart from ZR.app’s “compound score”-based solution in essence. We still employ a user’s recent critical power for some interval lengths (30s and 10min), both absolute and relative to weight, to obtain our predictions. The basic difference is that we were able to fine-tune the model a little bit further. Namely, we experimented with different relative weightings of those parameters, as well as different critical power interval lengths, to make our predictions provide a somewhat better fit in our tests, which we conducted on a massive amount of historical data.
2: How do we avoid Tanking and Sandbagging?
- We use 85% of the predicted score as a racing score floor
- We will soon start considering the player score at join time (rather than at signup, which may be done a long time in advance, in certain cases)
- We are considering discarding low-quality race participation (the player score will not decrease if we detect that there were no significant power efforts)
How did you test your ZRS algorithm(s) for accuracy? Was real Zwift racing data used in the analysis?
We used anonymized racing and power data from every distance-based race that has ever happened on Zwift to train and validate our scoring solution and the different score prediction models that we tested.
How did you decide on the 10min and 30s intervals used to compute seed score?
Intuitively, among the main factors that affect one’s ability to perform well in a cycling race are their skill/strength in sustained efforts (say, for breakaways and climbing) and in sprinting (for mass finishes). While skill (namely, being well-positioned in the peloton, having an accurate sense of timing as to when to launch an attack, being able to decide whether to respond to someone else’s move, pacing yourself well when time-trialing or escaping solo, etc.) cannot be inferred from power figures, raw sprinting power and high sustainable power are certainly crucial for success and can be measured by short- and long-interval critical power, respectively.
To arrive at our final interval lengths for the seed formula, we considered different power intervals and their correlation with the actual scores obtained from race results. The 30s interval was the one yielding the smallest RMSE (root mean squared error) overall. Among the long(ish) intervals, 10 minutes yielded good results and is more likely to be available and accurate in one’s recent activities (versus, say, a 30- or 45-minute interval), owing to the “sliding windows” nature of the critical power computation (smaller intervals are more abundant) and to the fact that, while training, most people tend not to keep their maximum sustainable power for a long period except in the rarer cases that their workout specifically tell them to.
How did you decide on the ratio between 10min and 30s in the final model?
When developing the seed score formula, we tested thousands of different combinations of power duration and rider weight with varying weights to each factor to find what best predicted the outcome of historical race data. What this revealed is 30s power, 10min power, and rider weight were the best predictors, with more emphasis on 10min power and less emphasis on 30s power. This even outperformed the 5min power interval used by traditional compound score formulas.
We are using the results of this analysis as our starting point for the seed score formula fully realizing it may need to be refined in the future as we shift racing categorization away from power-based categories.
What factors might cause an inaccurate seed score, and what have you done to mitigate these?
Sure enough, the seed score is but a best-effort attempt to estimate ZRS based on some input parameters (power and weight). It will never be 100% accurate. The error is inherent to the statistical regression method.
Let’s look into the most obvious problems that may arise and how we mitigate them:
- The outcome of a statistical regression is a curve (i.e., a mathematical function) that adequately maps the input (in our case, the critical power for the chosen interval lengths and the rider weight) onto the intended ZRS seed. For Zwifters who have already completed a scored race, we want to use the seed score as a floor in order to preclude sandbagging. However, since the curve was produced out of a “cloud” of actual scores from existing users (during the regression), some users will naturally fall above, and some below the curve that was found to be the best possible fit. If we were to use the curve itself (i.e., the very ZRS seed) as a score floor, then those users who sit a little below the curve for those input parameters would end up being overseeded (a floor would be set too high). To mitigate that, we use just 85% of the seed as our ZRS floor.
Note that, for someone who has never raced before on Zwift, we do not want to risk underseeding them too badly, because an underseeded race participant will likely spoil the party of a number of other players by winning a race that was too easy (whereas an overseeded participant may have a bad personal experience but will not ruin the fun of anybody else). That is why after the first ZRS race, when we already have at least one instance of power effort and scored race result, we let the player’s score go 15% below the seed. When the score of a user touches the floor (e.g., after performing poorly in a scored race), of course there are reasons to believe that the score of that user must indeed go down; we just don’t want it to go down too much to avoid abuse. A 15% deduction looked reasonable after gauging the number of Zwifters that would touch the floor. - Another possible problem may arise, so to speak, from the very user attitude to training. If, in the last 90 days, the user has failed to perform close to their peak in the relevant critical power intervals (namely, 30 seconds and 10 minutes), their seed score will be lower than it should have been. We are planning to mitigate this by establishing a threshold based on their last year performance. If their performance in the last 90 days falls below the threshold, we will use the full one-year critical power figures to generate a seed score.
Once you have your seed score, what are the main variables in the model that impact progression up and down?
- Race results (win/loss)
- Field quality in the races you participate (how strong are your opponents compared to you, in terms of ZRS)
- Variations in Critical Power (adjusts your ZRS floor)
How have you been tuning these, and do they currently feel accurate?
- We run analyses on anonymized data comparing power and ZRS, for all race participations (and also on sampled production data).
- We check ZRS distributions, and also how many Zwifters are sitting on their floor. Depending on the value, it could indicate that we should tune thresholds and other scoring model parameters.
- Based on results from ranked races we feel confident that ZRS is making racing at Zwift a fairer and more enjoyable experience.
What are the main things you learned during the public testing period that began in June 2024?
- Mechanisms that accelerate score progression are important to increase racing fairness. We implemented a podium bonus system after reviewing test outcomes and the feedback we received from the community during the testing period.
- The daily score decay applied after the user’s last race shouldn’t be too aggressive, or it might impact the score progression of Zwifters who don’t race too often. We tuned our decay rate during the testing period.
- The straightforward multiplication of absolute and relative power does not necessarily produce the best-fitting curve to map critical power intervals and weight onto a racing score that is meant to reflect a user’s likelihood to perform well on a Zwift race. Refined combinations of “weights” (exponents) of the input parameters do indeed provide better results.
How do the current ZRS boundaries differ from the previous category boundaries?
In the past, category boundaries were based on fixed power (to weight) thresholds, regardless of the number of users that would belong in each category. The way it is now, we rank the users by ZRS and impose moving thresholds meant to establish cutoff ZRS levels that will distribute the number of users into each category according to a predefined Gaussian curve (fewer people in the extremes of the spectrum, more people towards the “central” categories).
Another advantage of using ZRS for category boundaries is that it is much more flexible for customization by our event organizers.
Do our current boundaries please everyone? Zwifters who happen to be near the boundaries might feel like they are at a disadvantage. The difference is that, now, if you are in a category that feels too strong for you, you will soon have the opportunity to move down, since your scores will likely reflect the fact that you are not obtaining good relative performances in that former category.
Will Zwift Racing Score ever be ‘finished’? What’s next?
Rating systems need ongoing monitoring and occasional maintenance. We also anticipate continued constructive feedback and feature requests from the community, so we plan to keep evolving ZRS for the foreseeable future.
Future developments include implementing anti-tanking and anti-sandbagging measures, as mentioned above, as well as incorporating course profiles into our scoring system. We’re also exploring ways to improve our category ranges, including increasing the number of pens and also dynamic ranges.
Are you planning to reset scores regularly in the future?
No, unless the scores start to appear inaccurate for some reason. In that case, we would not only reset the scores but also investigate the issue and adjust the model as needed.
We might explore the idea of “seasons,” where scores reset at the beginning of each season, similar to other MMOs. However, this is not currently part of our roadmap.
What are some of the common misconceptions you’re seeing about ZRS, and how would you answer them?
A major misconception about ZRS is the overemphasis on the seed score’s impact on individual scores. While the seed score does set a baseline (which isn’t intended to be very restrictive), most Zwifters are not expected to remain at this score. Their scores will naturally fluctuate up and down as they participate in race events.
Another misconception is that the 30-second critical power used to calculate seeds may inflate certain scores, suggesting other intervals should be used instead. However, our correlation analysis between multiple power intervals and scores found that using both the 30-second and 10-minute intervals provides the best fit, with the 30-second interval having less influence than the 10-minute interval in the formula.
My Takeaways
I think there are a few key takeaways from Zwift’s answers above which speak to some of the recurring concerns I’m hearing from the community.
- First, their explanation of why they didn’t just use the ZR.app algorithm. Basically: Zwift wanted to make ZRS Zwifty. That means keeping it simple.
While ZR.app has a lot of very interesting features, and is not hard to grasp by people with some experience in cycling races, we were looking for a solution that would be extremely easy to understand by total beginners, as in a rating scale that ranges from 0 to 1000 (versus a typical ELO rating whose range is not just unbounded but also subject to inflation over time). - I’ve seen comments like, “Did Zwift even test this? Why did they choose 30s and 10-minute intervals, because those seem like the wrong ones to use!” But Zwift chose those intervals after lots of analysis.
We used anonymized racing and power data from every distance-based race that has ever happened on Zwift to train and validate our scoring solution and the different score prediction models that we tested. - It sounds like Zwift is working to include more historic data in the seed score calculation, which will be welcomed by just about everyone.
We are planning to mitigate this by establishing a threshold based on their last year performance. If their performance in the last 90 days falls below the threshold, we will use the full one-year critical power figures to generate a seed score.
While Zwift’s current implementation of ZRS isn’t perfect, I’ll say what I’ve said in other posts: it’s much better than the old system. Things may feel a bit messy right now as riders haven’t done enough scored races for their ZRS to accurately reflect their abilities, but over time I think we’ll see scores settle in and races become more competitive and fun for everyone.
Questions or Comments?
Hopefully, this deep dive has clarified some things for the Zwift community. Still got questions or comments? Share below!
They used all historic race data to see that 30s has a large impact on results, but then take in 30s data from all activities which overseeds anyone who does sprint work when fresh.
“But it’s just a seed, it won’t matter once you lose 20 races and get where you belong!” Unless seed and seed-15% are in the same category.
Lots of fancy words from people who think ZR.app is somehow complicated and difficult to grasp.
All my biggest sprints have come in races, I need the motivation to fight somebody to the line to squeeze out those extra watts
Sounds like you are not doing sprint work.
My biggest sprints come when I feel like testing a KOM segment while doing a fairly restful zone 1 or 2 recovery ride. A sprint after working a race at threshold or higher is never stronger.
Mike – your first assertion simply isn’t true.
What Zwift told me:
“When developing the seed score formula, we tested thousands of different combinations of power duration and rider weight with varying weights to each factor to find what best predicted the outcome of historical race data. What this revealed is 30s power, 10min power, and rider weight were the best predictors with the more emphasis on 10min power and less emphasis on 30s power.”
And yes, I clarified with them that the 30s power they’re talking about is the rider’s 30s power best (the same value used in ZRS seed score), not their 30s power in the race itself.
It would’ve helped if the press release you posted was more clear but as you noted they’re in love with overcomplicating their language (e.g. “gaussian distribution” is literally just a bell curve).
Did you ask them how they missed the fact that the data they analyzed was biased by being pre-sorted by 20min or zFTP numbers?
^^ This exactly Mike! It is ridiculous that Zwift are using the accumulated 30s power info from anything other than actual historic “Race” events to determine that element in the seed values they should use if they are also then capping the potential score reduction to a point where riders can get trapped long-term in a Banding/Category higher than they realistically should be in. If @James Bailey et al can’t see this, then they either aren’t paying attention or really have no understanding of what they’re doing!
I’m not surprised that 30s sprint power has been found to be a strong factor in predicting race outcomes when looking at who finishes on the podium. However, an enjoyable race is largely about being able to hang on and being there and getting challenged. The way the model is set up now tries to make it unlikely for the sprinters to be with the group. That way participants in one band have fairly equal changes to podium: you’re either a sprinter who rarely manages to stay with the front, but podiums if you manage to hang on, or you’re not a sprinter but you often stay with the front. While there is a certain fairness in that, it’s not as much fun. I wonder what the model would be if it didn’t look at finishing position, but finishing time. Then it would start to put people together in a band who are likely to finish within one minute from each other. Sure, the sprinters would podium more, but the whole event would be more fun to participate in.
Isn’t there an error in this – “Lastly – and this one’s a little hard to grasp, perhaps – the ZRS algorithm distributes riders’ scores so a band of, say, 100 points in the more densely-populated groups (C and B) represents a smaller range of abilities than a 100-point band in the A or E groups. So just because the E and A bands are wider doesn’t necessarily mean ability ranges in those bands are wider.”?
Shouldn’t it be – “Lastly – and this one’s a little hard to grasp, perhaps – the ZRS algorithm distributes riders’ scores so a band of, say, 100 points in the more densely-populated groups (C and B) represents a BIGGER range of abilities than a 100-point band in the A or E groups. So just because the E and A bands are wider doesn’t necessarily mean ability ranges in those bands are wider.”?
Otherwise a wider range of points and A and E with a wider range for eacg 100 would mean a huge ability range in A and E?
You passed the test! 🙂
I just removed that section, since it was confusing even to me, obviously…
To me, the new system is much better because it is finally fair. Before, you could end up finishing a well attended race in 10th position and being ‘promoted’ to an upper category while folks that had finished ahead of you would stay in the same category. That was not right, especially if you had finished well behind them. Most often this would penalize lighter riders. From what I have observed, this is no longer the case. If you are promoted because of a top finish, people around you are going to be promoted too. I am sure some small details can be refined overtime but it is a great improvement.
I have seen numerous examples of that happening. People finishing in the top half of the pack but not close to winning getting a points increase that has pushed them over the pen boundary, where the people ranked higher got points but did not get pushed up.
ZRS, in its current implementation, has most of the same problems as CE. They’re just impacting different people.
“Before, you could end up finishing a well attended race in 10th position and being ‘promoted’ to an upper category while folks that had finished ahead of you would stay in the same category.”
Exactly the same thing happens here. If you’re in the top half of the results you get a points increase, and it has nothing to do with the ranks of the people around you. So 10th place out of 25 can get you promoted a category even if you’re minutes off the lead pack.
You don’t necessarily get a points increase just because you’re in the top half of results. I rider who is ranked highly against the field, who finishes near the bottom of the top half, could very easily get a score decrease.
I’ve yet to see an example of this, but plenty of the opposite – lowest ranked riders losing points for finishing not quite at the bottom, highestranked riders gaining points finishing mid-pack. This isn’t vElo.
But are you telling me that the winner of the race wouldn’t get promoted as well? Especially with a podium bonus? Or maybe after winning a second race? In the old system, winning a race had no impact on promotion. Now I think it does. But I’m willing to be proven wrong.
Sorry, but being promoted from a poor result, usually just meant that you raced badly 🤷🏻
Not always to do with racing badly, it’s mainly because a flat w/kg limit with a watt floor is a terrible way to set race categories. It meant that heavier people, or those with power under the watt floor, were allowed to be fitter and faster than those in the middle of the weight range.
Thank you. That was exactly my point.
sounds like your race score is below 200 – an in that case, you were struggling at the end of C or in D.
this entire thread simply proves that no one knows what is going on! nor how we as riders are supposed to stay inside the rules (when we don’t know what they are). The old system sucked, but at least I knew, as a rider, when I was going to get a gain/loss or even a DQ on zwiftpower. Now the algorithm is so complicated we don’t know what is happening. My zwift score just went up 10 points explain to me mathematically, and not with fluffy words, as to why?
You don’t have to know the rules. You just ride the races. The only reason to know the rules (algorhythm) is to game the system.
I think there are a few very large holes here.
The first and biggest elephant is also ZR.app’s problem – ZRL doesn’t affect scores. ZRL is the race that’s the “goal race” max effort for more people than any other race. It has big quality fields. But the points config isn’t available to either ZR.app through ZwiftPower or apparently to Zwift.
The second is a drawback of ZRS vs. ZR.app -> ZRS doesn’t appear to adjust scores based on points races. Points races are fun. Probably 90% of the racing I do is not scored by ZRS in its current form.
I’m sure they’ll get to supporting points races. It’s great to see them making progress. I think it’s a complex task overall. If we waited for everything at once, I’m sure we’d be waiting for a long time!
Until Zwift as a game supports points races natively, you won’t see points races being included in the Racing Score algo. That’s my (educated) guess.
That said, I don’t understand why ZRL scratch races aren’t affecting our scores. Seems like that would work just fine.
Does Zwift even know if an individual ZRL race is a scratch or a points race? I don’t think they can tell a difference hence they exclude all ZRL races.
Glad to see this one finally get out the door. Good write up Eric.
Hey buddy – thanks for all the work you did on this in days of yore. Hope all is well!
So Racing Score currently works incredibly well for Tiny Races that last ~10 minutes?
No, it’s nonsense!
While playing with pen score ranges helps to some degree, the formula is so bad it still leaves huge ranges of abilityin pens.
0-300 is far too big for the weakest pen in Tiny Races as things are today, but you (Eric) said you don’t want to reduce its range, because it will mean the middle pens ability range will become too big.
The theoretical range of Racing Score is 0-1000, but in reality, it’s much closer to 100-875 for those with some power data.
The seed formula should be properly evaluating racers so that ~150 point ranges (eg. 100-250) per pen keeps ability ranges reasonable.
The range in the Tiny Races is 0-260. In reality nobody is a 0. If you look at the results almost everyone finishes fairly close to what you would expect based on their ZRS. My ZRS is 230 and I finish somewhere between the top 40% to 10% depending on the route. Exactly where I should.
In that “I have no idea what this means” section, they don’t seem to have been very kind in trying to explain things. They seem to be saying they tried all possibilities of multiplying different data values together, along with applying different weightings to produce even more combinations. So using a brute -force approach to see which values (e.g. which power durations) with which weightings provided the “best” seedings. However, what they don’t mention could be a critical flaw in their analysis in that all existing race result data would have been already pre-sorted based on zFTP categories, unless they only looked at mass-start races without the usual category segregation. So this pre-sorting influences the results (e.g. someone who might have won the C race if the category thresholds were tweaked is in the bottom end of B so had no chance of winning). If they didn’t take account of this then it calls into questions any conclusions they drew, because the data was essentially “biased”.
Agreed. Of course 30 second power is going to be a big determining factor in winning races when you’re up against riders who have similar w/kg numbers. Big sprint watts means nothing if you got dropped 25 minutes before the finish against someone with 2w/kg more FTP than you.
I’m dropped 2 mins into 35+ minute races by riders pushing over 2 w/kg more than me. I could have the best sprint in the world but when I’m over 5 minutes behind the leaders it’s pretty pointless. The new ZRS has landed nicely for those endurance riders with high w/kg over the race as they can just ride others off their wheels.
What is your ZRS? If it’s 370 and you’re racing in the 350-520 band then this is exactly what should happen. If you’re at the bottom of the category you should be finishing in the bottom half. Dropping the 520 down to 420 wouldn’t help. You would still be at the bottom of the Cat. The only thing that would help is to increase the 350 to 380 and then you might even win. The problem isn’t with the system, it’s that you don’t like where you are on the score.
Spot on. Exactly what happened to me. My 20min power is 3.2 the next closest to me was 3.8 (and went up to 4.4) result? Dropped like a stone on lap 1, so solo 2 laps of Innsbruck for me. Suppose I could have used my good 30s power when I was 5mins behind the pack and see if I caught them…
and if you did that, you would get “rewarded” with a massive score increase – LIKE I DID (60 pt increase to be exact), for just smashing out a sprint section by yourself to break up the boredom of getting dropped in the first 2 minutes.
Zwift just sent me an update to that section. 😀
It’s notable that Zwift is trying to improve race scoring. There are so many ways that a rider could be “cheating” on an individual level, Zwift could have called it good, but they’re trying to improve what they have control over.
> But some comments from the community have simply been inaccurate – people “guessing” as to how things are working, with limited data.
This is so true and so frustrating.
If Zwift was really worried about this they would release the formula. I am sure someone will reverse engineer it eventually, but until then they are at the mercy of people guessing.
They most certainly are not worried about it. Like Eric said in the write up, they can’t let users know how to sand bag the score while also using the score to test actual racers and events. I’m fine with not know how the sausage is made. I just want to race in races where I’m with similar abilities which I believe the ZRS will do better than the previous category system.
I’m one of those riders whose seed is inexplicably far too high. I’m a low C/D rider in old money but with a ZRS seed of 400 and in a pen with As, Bs and very strong Cs. I’m dropped on the first 2 minutes in the 350-520 pens yet my ZRS barely reduces. Racing is no fun at all in that situation.
Same
If you’re seeing A’s in a 350-520 pen, those A’s are the outliers – not you.
I’m a mid-cat B, and my score after getting killed in recent races is around 565. No way a legit A is below that… let them put in an effort or two and it’ll jump up.
If the system was functioning properly, there should not be any As or Bs in his pen. On top of that, the pen should not be in the 350-520 range since it obviously mixes a lot of racers with different skills and endurance. Most of the feedback about ZRS comes from racers in the 300-600 range, not just because it is the most populated, but mainly because the ranges are way too broad. It is definitely wider than the CE.
I’m a 2.5 w/kg racer, in a pen with riders 3.5+ and you don’t think I’m an outlier? In crit races I’m dropped within a minute then lapped. This never happened when I was in C races. My ZRS is currently 365 and I get beat by riders with ZRS below 300 is old cat C races. ZRS has ruined racing for people like me yet Zwift don’t care or try to fix it. Awful
i have a friend, we both have raced zwift since early 2017. we both are close in C (almost same weight), although id give him the edge overall in zwift races for outlasting me, but i have a better sprint. ive always been atleast 60-100 race score above him, and never been down in his ZRC cats – i ALSO see the same old people we race against also down there with him.
im convinced a certain zwift employee is trolling and altering my score. initials NM. is all i will say considering my history with this person.
Totally agree.Same experience for me.I was an old Cat C and had the occasional top 10 finish.Am now finishing consistently in the bottom 3rd of all races(350-520)
Usually find a third of the field are showing as B riders. So if my recent poor results are anything to go by I would say the older algorithms were more accurate than the new ones.
I’m a 3.1 getting creamed by 4.0 racers. I can only pick pan flat races and sit in. Then if I do well, I go up and it get’s even harder. Every year I end up quitting and starting back up after 90 days when I revert back to a blank slate. I’m a non-sandbagging C that gets moved up to B and then I get crushed and do Rouvy for a few months.
eric, im a C, and my race score has been in the mid 500’s, im currently 477. and I see A’s who are in this cat, and B’s who are less than 400, in the 300’s ( and when you look at their history results, they werent terrible )
Eric, the C pen as I will call it because that’s what it appears as under ZP, is absolutely full of riders over 4.0wkg 20 mins and D pen has more than it’s fair share. That’s why I say this system can’t work. You need to apply some form of common sense to this so called formula. If these riders have a base score that initially starts so low, all it does is create the new breed of sandbagger. The reverse of that is the riders with 2.8 wkg in pen C and not at the bottom of it, so they can never go as low as D pen, where they still wouldn’t be competitive. The system need throwing away and soon!!
The million-dollar question is: are those A/B riders in the C/D pens actually in the wrong category, and thus, they’ll be upgraded once they race a bit? OR is the system actually broken and putting too strong a rider in too low of a category?
My hunch is that it’s at least mostly the first one. I don’t know how long it’ll take the system to settle in for us to know.
If it’s the first one, and things settle nicely, this should also at least somewhat fix the complaints of bands being too wide. Upgrade the riders who are too strong for their category, and suddenly the categories won’t feel so wide.
Separate from these issues, but also important, is the discussion of whether Zwift could improve their seed scores so riders who take a break can’t return and race a lower category just because they haven’t raced in 90 days. (I think everyone agrees that looking further back in their power history would be a good thing.)
Eric, there is a basic question here that either you or Zwift are not answering directly. Why didn’t they put limits on the current ZRS pens based on the legacy CE pens for the seeding i.e. the 350-520 pen should have just started with C riders, 520-690 pen with B riders etc. We, the consumers, understand that there is this “growing pains” with ZRS and Zwift can simply say that it will take patience as they try to collect more data from actual ZRS-based races (for how long or soon, we don’t know). But this initial phase with this under/over-seeding could’ve been solved easily, in my view, by merging some aspects of CE to ZRS, as mentioned above.
My feeling is that there are far too many riders with an initial seed score too low and the reverse also evident, with many seeded too high. The lower seeded riders become your new breed of sandbagger, with the ability to drop into a category that fits their needs. Some will race and be true to their ability and find their level, but many will sit in a position they can control and win from. The new breed will come to light soon enough. The bottom line of this is that the initial seed score process is inaccurately calculating a riders true potential. Until the initial seed score is more accurately calculated the system will be even worse than the previous system. Even then ranking systems will struggle on Zwift as the field is just too large and varied.
I haven’t raced since last winter, and as this one approaches I’m back in Zwift. Did 3 or 4 rides. 3 were workouts and 1 Robopacer. 1 of those workouts was a Vo2Max. I’m a 4.15w/kg rider and Zwift “gave” me 510 points. I think it’s pretty accurate and should go up soon as I barely did hard efforts. The one reason I can see why people have lower points than they should is because they’re back from summer and did low-intensity workouts only. Pretty soon this should be automatically fixed. I’m planning to race quickly to see how it goes.
Wow, 4.15W/kg is definitely in A-cat territory! Having 510 seeds seems a bit wild since that’s where the old C-cat riders usually are. You’re going to be like that dude from tiny races last week in zone 2 who snagged a win in the C pen, even though he is a diamond rider with a crazy 3345 points and power radar values stretched to the max, just like my grandma’s old rubber bands! This is the racing we have now.
I was “B” last year, and think I won 1 race only ever. Got a few top 3, and that’s all. I’m not “A” for sure. I was 510 two days ago. Did a group ride yesterday and now I’m 515. I will be in the 520 range pretty soon, probably even before I race. That’d be the old “B”. So I think it’s not so bad.
same
You’re complaining because the system is working as it should. If you’re a 400 you shouldn’t be winning in the 350-520 cat. You should be somewhere in the middle or even lower. Those guys beating you aren’t A and B’s. Those are guys with a 500 ZRS, racing in the correct category. Focus on racing the guys near you and let those 500 ZRS guys go.
I think that the statement regarding why not just use ZR app is off. When I started to race I don’t needed more than to know which category I should join. After sone races I discovered ZR app and I did register myself to see more stats. Are they complicated or overwhelming for a person, which have ridden a few races? No. I think not being enough „easy“ should not be a problem. Zwift could tail what users see to its liking or make it layers, so that at all places users see just simple score and details only on a dedicated subpage. Well, but then we have just like in Ja aScript world, another day and yet another library, which solves already solved issue but a bit different.
5mins pure Watts x W/Kg x {weight adjustment factor for non-pro Zwift users}
No maths degree required for zrCS at zwiftracingapp!
The issue with the current and previous systems is that they does not take into account how much elevation there is on a route. Racing scores should be scaled based on the elevation gain in a race and the weight of the rider entering the race. A one size fits all system means that larger riders will stick to flatter races or will be incentivized to lower their weights to take advantage of the lower drag and faster climbing it provides.
I think it will be an improvement in the long term but certainly still needs some work done
“Note that, for someone who has never raced before on Zwift, we do not want to risk underseeding them too badly, because an underseeded race participant will likely spoil the party of a number of other players by winning a race that was too easy (whereas an overseeded participant may have a bad personal experience but will not ruin the fun of anybody else). That is why we use a modest 15% deduction in this case
What does that mean new riders are coming in with a 15% discount by default?
Also i’d like them to run the same analysis on ZRS races now to see if 30 seconds still has a high correlation with results. The problem i see with the analysis being done on past races is that they are all 99% cat enforced where you don’t get riders with widely different 10mins power like we are seeing with ZRS so i do still think that 10s power is overweighted in the seed score. Similarly those with low 30s are underseeded.
ps and slight moan when are organisers going to get the tooling to take advantage of ZRS rather than having to send smoke signals to zwift like it’s the year 2000
In my mind the biggest problem is strong riders riding and ruining races after a hiatus because they have been downgraded. Let them race where they left off and be downgraded if they actually have gone out of shape.
100%. They seem to have invented ‘decay’ for racers who don’t race very much in order to compensate for the inflation built into the model for racers who race a lot. COmbined with the 90 power intervals and it’s a sandbaggers paradise.
The decay is very minimal now, so I don’t think it’s much of an issue. The bigger issue is that they’re only looking at the past 90 days of data. If they could just expand that to 1 year, they’d be able to better seed a lot of riders who don’t Zwift much in the summer…
It’s hard to track that, obviously, but is good to hear. It’s hard to see how with podium bonuses, lower floor and 2/3rds of finishers having a score increase how they will avoid the system inflating out of control, then though.
Not hard to track the decay rate… just look at your score each day, without racing! 🧐
Wouldn’t that mean NOT racing every day? Literally impossible.
Today I became the only person to have raced every ZRacing Stage, I’m not letting that go!
Haha yeah on sunday I got smoked when racing 350-520 (my score being 498 at the time), doing 265W average and a guy doing 330W average over 40 minutes won by a large margin. Turns out it was his second event after summer and he was scored way too low, now the guy is sitting at 717 racing score, while I am at 514. He just finished like 2 minutes ahead of everyone else. But since I am bad at sprinting I don’t mind. Its just weird that some A racer can ignore Zwift for thre months and then blast throught the first ew races (after the race on sunday the guy was just rated in the lower 500s, so his score improved only by like 30 points, which is ridiculous when doing 330w over nearly 40minutes, but there was no need for him to spring since he led by such a large margin).
I have been racing ZRS since the Beta beginning and trying to give as much info to them as possible. As time has gone on, it seems to have leveled off to where it seems correct and close to my old days IRL. Some races do well, others get dropped and do OK, and some especially on Climbing routes I get crushed. At first, seeding was so high, after a couple K I was big time dropped cruising alone for the rest of the race SOLO and the effort was not race pace more like a Bot ride but admittedly not racing just a cruise TT (bored). Can’t even begin to want to understand this stuff, I’ll just keep pushing the pedals and see what happens. “ride on”
I’m excited about the changes to the ZRS! It’s great to see Zwift focusing on fairness, which will make races more competitive and enjoyable.
I hope that they can continue to invest and evolve the system so that it’s a positive experience for all riders!
Dynamic Pens would be amazing. You signup to a race and Zwift determine how many pens and what ranges for each pen based on number of signed up riders and their abilities. Finally match making!
Ok guys, let’s be honest: zwift didn’t use the zp.app racing score so they can say “it’s mine” – legal issues as I see, or just to say “we created this”. It’s not hard to understand that you’re “copper” or “diamond” as is the names of the scale in zp.app. No, the scale 0-1000 it’s not as easy than “copper, silver, gold” (and not as cool!).
That’s your opinion. I don’t share it.
I don’t remember what my ZR.app precious stone rating is, and it’s definitely not intuitive to me which ones are better than others. 0-1000 makes a lot more sense to me.
There are these three racers— A, B, and C—and they all have the same ZRS score. Sounds like the setup for a joke, right? But nope, it is just how things are. Not really sure what sense it makes to you if each of them got totally different power profile and abilities.
Sorry, but if ZRS runs 1-1000 and you have (for example only) 20,000 total racers, you SHOULD have 20 with RS 1, 20 with RS 333 and so on. The system doesn’t seem able to display 333.05, 333.10 and so on.
Now, I’m well aware that there are problems with distribution, but if and when all this evens out, you should expected to see someone in your pen with the same RS every once in a while.
I am an A cat racer in the old system and have been for years. My 20 minute power is over 5wkg and over 6 wkg for 5 minutes and I have done over 550 races on Zwift. I joined Zwift labs right at the start of this ZRS stuff and I was given a race score of around 580, because I am a lightweight climber basically. I thought that racing score was a joke, but wanted to prove it. I did loads of Zwift labs races and in 4 – 6 weeks I had got it up to 760. I think it went up every race I did. Then Zwift tweaked something and put it down to 750. Then they reset it a couple of weeks ago to 670, and they just put it down to 660. I always raced in the top category in that period and never had any bad results. So in my opinion its a total load of rubbish and I just hate Zwift now I have pretty much given up on Zwift racing, unless it’s stuff like WTRL or ladder league.
This is a rare complaint – a racer who says his score is too low…?! #fortbragg 😂
I can kind of see what you mean, but the point is the score is just wrong! And if it’s wrong for me then it must be for loads of people and then the scoring system just doesn’t work.
Instead of making fun you could also ask yourself if I see low cat riders, mid cat riders and now even the best riders complain about the seed score, that there might be a point and there is something wrong with it.
Perhaps pulling data from a flawed system wasnt the best way to start building a new system after all.
Kudos for calling it out and telling it like it is. I’m sure there are other lightweight A and B cats that are quietly pretty pleased with the pen they’ve landed in with ZRS. B’s are crushing every C cat (350-520) race at the moment.
The emperor is naked. I am glad that someone stopped to be politically correct and speaks up.
Man, this just proves how broken this system is. I am an upper C rider, and my 20-minute power is 3.2wkg while my 5-minute power is a bit over 4wkg. Guess what my original ZRS score was? A whopping 575! Whoever designed this must be completely nuts to think we are equal racers. You’d probably take me down with one leg!
are you pretending to be me, coz thats my story. nearly identical.
Same power, same formula, same score?
Ditto! I haven’t even done any “proper” races with ZRS enforcement, just group rides that I enjoy which often become spirited at the end. My 20 minute power is 3.4 wkg and my 5 minute power is 3.61. My score jumped overnight from just under 500 to 595 apparently based on just one 60 minute C category group ride in which I was sixth at the end of the hour. I would have never sprinted with the As and Bs in the front group had I known I’d get bumped way up for a result that wasn’t even in an actual race.
welcome, Im a cat C with about 1000 races.. i was there with you in the 500s – i came last.
The main arguement against this terrible system is very simple.
The biggest issue with the old system was that the categories were far too wide.
This has made that far, far worse.
It is absolutley awful.
Agree w the wide categories being the fundamental problem. No problem with the score per se with me, no better or worse that FTP tbh. Would love to see more races with smaller categories, or ideally matchmaking.
Is this true? I had assumed the opposite as there are effectively 5 pens now, where before I didn’t see E being used, so there were 4.
I’m personally a huge supporter of the new racing score. Overtime Our scores were all move away from our seed scores at the start. if you’re beating people ranked higher than you, your score goes up and you might earn a promotion. I do have a bit of sympathy for people who move up a category as the categories do appear quite wide. I did enjoy the low and high-end races for that reason.
I find it quite funny that heavy riders are complaining that it is unfair. The old system was so unimaginably biased in favour of heavy riders it was untrue.
Many of the heavy riders were fully aware that the previously system was biased, and had been advocating for a fairer system for years.
Sure, some thought it was all about their superior race craft, but as the quote goes “this man does not represent us”
Thanks for the deep dive Eric! After reading the article and responses to; it may well be brain surgery…at least for me. Am I too far behind the curve to simply appreciate a Top-Ten result, then age-group result listings? How an individual finishes against their age-group peers seems like a good race result benchmark.
Having done statistical modeling and data analysis for a living, it sounds like Zwift are on the right track in their thinking and processes. Those that think otherwise, well, maybe we should give them several months to see how it all shakes out rather than pass judgement after just a few days of anecdotes.
With the new system in place, it would seem that this may open up a plethora of new opportunities for new racing formats. While the Tiny Race Series was run under the old format, this is the kind of thing I am talking about. That particular series requires unique strategy and tactical plans to do well regardless of cycling power. As an example, if you burn yourself up in the first race, you are toast for the fourth one. Should you try and catch the person in front of you or be comfortable just sitting in front of the ones behind you? It’s a lot of fun making these decisions on the fly knowing you still have two or three races yet to go. It’s not just about banging out watts.
What kinds of other formats? One example might be “handicap” type races like they do in horse racing. They could internally alter the weights of riders, the power numbers, the perceived gradients, or back compute power numbers after normalizing the Racing Scores within the event to level the racing field making it more of a tactical race. Or they could set up races where you are competing against yourself, but in a group setting. Make it so that how you finish is dependent upon how much you bettered your 90-day moving average of your power numbers, for example, relative to how everyone else did against theirs. I am sure creative people could come up with even more interesting (and fun) formats.
The key is having enough options to let people self-select to match their personal styles of racing to optimize their own abilities and desires.
Kudos to Zwift and you, Eric, for keeping on top of it all. If it weren’t for Zwift, I wouldn’t be able to ride half as much as I do.
I have a question about the lowest ranked ZRS riders. Since the last recomputation, an average 10% of the Pen E racers have had racing scores of 11 for races I check. That was true of the “|| Using Racing Score” events and more recent ZRS live events. This includes me, but I try to be an advocate for new, beginner and older riders.
So, if there won’t be any more recomputations, will we have 11 racing scores forever? Earlier iterations of ZRS lead me to guess that they won’t change.
I know this raises questions, but I want to keep this focused and I stand ready to answer outsid of this original post.
Each rider’s seed score can change as they set new power PBs or old ones age out. Plus their score will change when they finish a race.
Would that were the case, Eric.
If you’re seeing something different, I’d report it to Zwift. Or share details here…?
Done.
I don’t understand some of the complaints tbh. I am a strong believer since a long time that the best cat allocation system should consider both racing skills (“race points”) AND power data. You will catch the regular racers who may not have the biggest watts, but who are really good at racing (and for some of them, also very good at staying just under the next cat limit…) and you will also catch those who are wattbombs but who do not race frequently or who are just poor at race reading. To that extend, for me, this goes absolutely in the right direction.
Whether it is perfect, maybe not. It will surely take some fine tuning as Zwift mentions it. But I find it nice that Zwift is also giving attention and invests into its racing community. I have heard a lot about experienced racers complaining about Zwift not focusing on its racing community. This is a good sign. I think that there were too many “competitors” (ZP, ZR, …) on which some race organizers are relying for cat allocation (if not their own class system), which I find is really “too much” for someone new to zwift. It will be much clearer and easier if all use the same base.
And regarding the fact that some complain that their sprint is useless on a climb, I agree. But Zwift has said the 30s and 10mn are the efforts that give the best curve for seeding. What I understand from that, is that taking like 2mns, 5mns, 20mns etc would have not significantly changed the seeding, but would have made it less reliable. In other words, I don’t think that having a very good 30s power will drive someone into a cat in which he does not belong if he has a poor 10mn power. On top, Zwift says that there is more weight on the 10mn power. And this is only one parameter, as the racing skills are also considered.
And finally, if you have a great sprint power, then, like IRL, you will be aiming for flat races in which you can draft and crush everyone on the finish line. Oppositely, if you have a miserable raw power but are super light, you will let your ftp speak on a climbing race. There is a great article on this site that explains why on zwift raw watts are more important than w/kg. I still find odd when I’m being passed by a 95kg rider on a climb on zwift whereas this never happened to me IRL in 40 years of biking. This will surely change the racing experience of those people, sorry for them. But if the goal is to have Zwift closer to what you can experience IRL, then in my view this is a step in the right direction.
Many will disagree, I know.
30s sprint inflated my score to 580. Once it expired my score dropped by 100 points. No racing at all as I am not the racer. Certainly, there were some score resets and other adjustments in the meantime, but still…
Personally I think Zwift would have far fewer issues and confused riders if they stopped attempting to provide riders with an updated ZRS value the moment they cross the line in an event and instead simply moved to a system whereby a Zwift Companion App Notification is created for you to tell you that your ZRS has been updated AFTER IT HAS BEEN FULLY CALCULATED! This current situation of showing people a score which then changes hours later, often more than once for people who do multiple events in any particular day, is a complete s**t show. People do not NEED to see some temporary guess at their ZRS when they cross the line in an event, they NEED to see where they finished, the ZRS stuff should only be notified to riders once it has stopped bouncing around.
Does it changing really matter? The only time the score has any real importance is if it causes you to either be promoted or relegated to a new race category. Also, if it does get re-calculated after you’ve initially checked it post a race, I doubt the change would be very large.
You can log into the companion app at any time and check your current score.
Something isn’t right here – i am a middling C Grader – FTP 2.8 w/kg – my inital ZRS score was 590. First race got dropped as soon as it went uphill – expected everyone else in my cat was B Grade and averaged 3.8 w/kg for the race, ZRS score dropped as expected.
Second race, mixed Cat start, held on to the front group for 1/3 of the race, got dropped, got picked up by the 2nd group and finished 3 mins down on the others in my cat – look at the results, ZRS score drops as expected – the system is working.
Fast forward 45 minutes – a rider in my car has sailed round at 95W average and finishes 40 mins behind me – now my ZRS score has gone up – despite getting dropped and finishing 3 minutes down – there were only 5 entrants in my cat, i came 4th and the guy that was 5th was clearly a non trier.
I think it is just better. Racing is more fair and less risk of getting dropped in the first two minutes. Better category fit. thanks for the insights on how it is supposed to work
Would like to see the addition of each rider’s ZRS listed on the event signup screen. Would make it a lot easier to see what your competition is like without having to click on each riders profiles. And then you have the private profiles so you can’t even see theirs. Knowing where you stack up against the competition before entering is nice to see but can also see the argument that this would lead to smaller races knowing that you may be the lowest ranked rider and not register at all.
It’s weird that Zwift have chosen 30 seconds and 10 minute power as the only important power numbers, but in the results field of the events they only show 15 sec, 5 minute and 20 minute power. Why the disconnect?
What is the function of Zwiftpower and why does Zwift give out a letter ranking still? Should it all move to ZRS?
they are generalizing, and EVERY person is different, and EVERY person changes (whether by accident, age, fitness, or life change, power source)
they try to tell you, your score will alter. in my case its the most ridiculous POS ive come ever across
30 sec for me is easy, 1 minute is getting close to my “max power to weight”. after that the fall off is dramatic.
Has the topic of single pen/mass start event been discussed? The simple change of making races mass start seems like it would solve a lot of the current issues faced today with sandbagging and seeding. It would also make the events during off hours more interesting where rider turnout isn’t very high.
Mass start events, in my experience, are always more fun, engaging, and challenging… pushing me to new PRs that I normally wouldn’t achieve racing in my normal category. The most common complaint I see with mass starts is that if a rider is “lucky” enough to catch a break with stronger riders, they can finish much faster than their peers in the same category. But with the ZRS, this will self-correct and the “lucky” riders will cat up.
Eric, I beg you! I know the Tiny Races don’t suffer from low rider turnout, but it would be an interesting experiment to try having those races be mass start and see how it goes. My future FTP improvements would be very grateful! 🙂
to solve the problem of low A riders riding in the B group, they’ve created a system in which low A riders who don’t have a sprint get to compete against C riders who can sprint.
There are also B riders who have an A level sprint, but a definite B level FTP, who have race scores above 700, who now get to race against A+ riders and ‘pro’ e-sports guys.
I know a genuine A+ rider with a mammoth FTP (over 5) who has a race score of 570.
I also know a B rider with a 3.8 FTP who has a 1400w sprint, who has a race score of 750. He is only getting DNFs since the new system (although to be fair he was winning almost every B race he entered and might have been sandbagging by holding back on 20m watts).
The new system also encourages riders to enter races in which few riders attend. If you do an 690-1000 race that only has a handful of riders, you can get on the podium with a max 30s effort and you’ll get a bunch of points, with maybe a 3.7 average for the race.
Conversely, if you ride a race with 60 starters, and finish 32nd, 5 seconds down on the winner, with 4.4 average – you can LOSE points – even if you raced a category higher than your race score. This is mad. Finishing a race a category higher than your score with what IRL would be ‘ST’ (same time) as the winner, should be rewarded, not penalized.
Also, it seems that light riders are now really favoured now on flatter races.
I raced the Japan race (1300 CET Friday) today in the old A category. It’s usually a flat race, and as a 78kg rider with an FTP of 330, I like riding against lighter Japanese riders on flat courses and usually finish in the first group (at the back of it as I am not a sprinter)
Today I was absolutely blown away. Did 3 watts off my PB for 10 minutes and then couldn’t stay with the bunch. It felt way faster than normal. I was 30% of the time in Zone 6, going all out to get back on.
I noted around 20 DNFs, way more than usual, and lots of heavier riders. Getting my as* kicked by 56kg guys doing 240w, while I am doing 350 watts ON THE FLAT is a joke. It’s not remotely realistic, especially if you in the bunch. This will just encourage weight doping imo. There’s already a massive problem with unrealistically light riders in the highest category…
Ye Gods, please let THE FLAT not be changed into THE HILLS and all about w/kg just because light riders had to work a little harder…as they do in real life.
I’m still amazed Zwift doesn’t use any IRL performance data for seeding. I understand the complexities of that, but it would catch those who ride outside all summer and come back to D. It also appears that performance in ZRL doesn’t affect ZRS at all?
Do we know when ZwiftPower will show RP in the signup and results tables?
I find myself at the top end of the 350-520 range. I started on 440 and now up to 499. So I’m closer to bouncing into the next category.
I usually finish in the top 20%, but normally way off the leading pack, especially on courses with long climbs. So it feels like I’m being unfairly rated compared to those finishing ahead of me. Perhaps this is the relative scoring settling down?
ZRS as it is… Is silly.
30 seconds has no place here. Sprinting leads to results. Results lead to better positions in a field. As a result people with a good sprint will be more likely to move up than people without a good sprint, all else being equal. Time would automatically take care of this and not enable 30s to impact the initial seed so people are in pens they can’t keep up due to large different in longer term power.
ZR.app does it much better. Scores are most more in line with expectations. I havent seen large discrepancies in score there myself while I see numerous in any given ZRS race. Zwift likes a 0-1000 score instead? Well, this is just a normalization they are applying. Same can be done to scores of ZR app. It would have been one minute of work.
Also the talk of simplicity is ridiculous when they are adding more parameters to improve the fit. That’s data mining and it’s clear that’s what they are doing here. Not good data science, regardless of how many marketing words they throw in. Especially on a biased dataset as those are heavily impacted by the old categorization.
It’s a failure by design.
That’s pretty unfair for a rider who repeatedly gets dropped in the first couple of minutes of every race and rides alone for the next half hour. That rider is unlikely to want or need to do any serious sustained efforts in that race from that point onwards,once he/she realises there’s no chance of getting back to the bunch. It means that they’re locked in a cycle of ‘off the back’ racing every time they race. Hardly much fun at all.
I hear ya. BUT my guess is those first couple minutes would count as a serious power effort…
Than what good does it do ? That would not take long for those abusing it to find out that if they ride with the group for 5 minutes it counts as an effort and they can zone 2 the rest of the race to get a lower score.
And besides that, doesnt the floor already prevents lowering your score a lot. I am not Phd Scientist like Zwift has, but I think it is impossible to drop more than one cat with the 15% we have now.
So I dont see the point why they are considering this.
Your ‘guess’? So those of us pushing our nuts out of our backsides that first 2 mins, only to be dropped and then riding 30-40 mins solo, have to hope your ‘guess’ is right? Else we’re purpetualy stuck in a pen we should never have been placed in there is no way to get demoted, even though the whole point of results based racing is just that, the results normalise riders into their correct ranking?
I’m glad we pay our subs, that have increased by 30%, to have racing ruined.
Funny too how it only seems to be the Ds and Cs affected adversely by the new scoring.
Look forward to giving it a go. Any idea if the racing score will be visible in the zwift power sign up list?
Sharing information can indeed make a community stronger as most can buy into an idea when it’s fully explained. The only thing I’ve gained from reading above ud that since 2019 they’ve spent a lot of time and effort to develop something that is completely useless. That I think is very worrying!! I believe a reasonable percentage of racers also currently believe this to be the case and that percentage will increase day by day, which will make the community weaker. Let us have the actual formula so we can pull it to pieces, rather than the marketing pitch.
It needs time and data from individual riders so you really need to enter a few events to get you in the right group ,I have been doing a lot of the events leading up to go live and it doesn’t have any advantage to light or heavy riders I’m a 55 kg rider and have to work extremely hard to keep in the main group of similar riders , I think riders who are used to the old category enforcement system are probably set in their ways on how a race unfolds and which races to enter but I’m suer this new system is much fairer more fun and makes you think more about race tactics and positioning in the group, like every new it won’t appeal to everyone b
ZwiftPower should now let riders select 10min output in column choices
Quite inexplicably, Zwift’s main event pages which are relatively new, inconveniently exclude both 10m and 30s values. The question really is whether it’s intentional?
When will ZRS replace the Cat ABCD system?
That’s basically what happened this week.
It mostly already has for racing. There are very few remaining ABCD races on the public calendar. There’s Zwift Academy, which is only for a few more days. Two or three other races per day are using category enforced ABCD, and a similar number not using any category enforcement.
It looks like it will continue to exist as a pace guide for group rides.
Let’s admit the sole reason Zwift won’t release the ZRS formula is because it is equivalent to 2+2=5. And they surely know that we would notice it. 😀
On a positive note, I highly recommend everyone to try racing in IndieVelo. You can create any adhoc event (with many race types that are not even supported by Zwift) for just yourself and fill in the pens with bots. And it’s is really fun to race with those bots! Once you get comfortable with iV racing dynamics (you need to unlearn Zwift peculiarities), you can join bigger crowd races with fellow humans. iV categorizes you in 4 different metrics and in the races it will put you in a dynamically allocated pen with racers (or bots if your pen has not enough human participants) of the same abilities. It makes it really fun to participate.
With all this ZRS ridiculousness I have been trying branching out into Mywhoosh and Indie Velo. I have only done one long IV race but I thought the pack dynamics were brilliant and I really enjoyed the race. There are human only races now, so I think they just need a slightly bigger user base and then it could be really good.
I’m pretty much in the same boat. After more than a year of ZRS news and now 4th month in of actual ZRS racing I have no hope of any meaningful change will be happening any time soon.
One thing about IndieVelo I want to emphasize is that racing with bots is fun in IndieVelo! This is counterintuitive but give it a try.
Hi @Eric Schlange it doesn’t look like the ZRS system takes account of the achievements of other riders in your race in calculating the score increases/decreases. If it’s right that it doesn’t do that, the results based element can’t work. It will simply mean your
score goes up if you’re on the top half of results/down if your in the bottom half. I really hope that’s not right, it would be crazy.
You’re right, that would be crazy. 😜 But that’s not how it works!
Your result is based on the number of finishers, their Racing Scores, and your finishing position.
Phew thanks Eric
The way it works now is that you can finish 5 or more minutes behind the leader and still get your racing score increased. So next race you’ll struggle even more. Makes perfect sense! not really. 😀
An example mass start race 0-1000 earlier today https://www.zwift.com/uk/events/view/4536513
With a few oddities around mid position finish, top half got score increase and bottom half got decrease.
Similar story in pens in Tiny Races one yesterday zone 2 https://www.zwift.com/uk/events/view/4539680
Racing Score is not an ELO system like https://www.zwiftracing.app/ , where in theory your vELO score.predicts your finish position and you get more points for finishing higher than expected.
not diving deep enough for me! please just simply tell us how the score is calculated? there is an algorithm behind this so why not simply tell us the what the limits are? we are racing in categories and we don’t know how they are defined/calcualted and as such we don’t know the rules we are racing under? so we suddenly get a DQ in zwift racing or we go up/down a category and we have no clue as to why? not saying the old sytem was good, but at least we knew what was going on
Who got a DQ (link please)? Those basically shouldn’t happen with ZRS.
Hi Eric, was speaking hypothetically, but thanks for clarifying. However you can still go up categories during a league season and be disqualified from them? This got me thinking, and this may sound conterintuative, but a lot riders are happy in the category they are already racing and don’t really want to move up: Who wants to go from a being a big fish to a little fish. Further, if you are in a league you want to stay in the same category for the duration. As such, it is quite easy to “cheat/manipulate” the new system by simply entering non-target races to lose and thus reduce racing score prior to important target events, as such mainting current category. Once people realize this we are going to get a lot of races with people simply riding at the back, the opposite of sandbagging. Also, we may see a broader distribution of rider strengths within category defined events (riders with “manipulated” vs”unmanipulated” racing score), the opposite of what racing score intended to acheive, and ironically making racing less competative. Correct me if I’m wrong but this is a) quite achievable and b) Zwift have been naive to this possibility nor do they have anything presently in place in the algorithm to combat it?
Zwift certainly isn’t naive to the possibility. In this article (and others I’ve published) Zwift references that they’re looking at countermeasures to purposely tanking your score (and indeed, they’ve taken some countermeasures already).
I think it’s too early to see how easy score tanking actually is. Yes, it’s fairly easy right now. But the more you race, the “firmer” your score will get. It may be that it’s a big hassle to tank your score in any meaningful way (like, it may take too many races).
If it turns out this is happening to much, I’m sure Zwift can figure out some countermeasures. They’re watching it, for sure.
Also, worth mentioning that, with the previous category system, you didn’t even NEED to tank your score in order to sandbag. You just sandbagged. So at least ZRS adds an extra (and not insignificant) step!
cheers again. I now know what “tanking” is :). Sounds good!
There are many considerations that can be made but the ranking of the tiny of 12/10 in cat. E and race 4. Seems to me to say it all
not exactly from ex D with CE where in 5′ max was 3.2 wkg😟