Zwift Chat

[bridgy]

This is where I'm still confused - I thought the new ranking was based on results, but it's still based on power and it seems power numbers are still the dominant statistic. I can't think of any real world sport where power is used to rank, it's always on results (and I assume online gaming is the same).

It uses 30s and 10m power over the past 90 days as a starting point, then adjusts your score based on results, and the relative strength of the riders you beat (currently only in races using ZRS ranking). But those 90 day power numbers also do affect how low your score can go.

This makes 100% sense to me for two reasons.

Firstly, you need some kind of starting point for categorising people or you'll have to start everyone on teh exact same ZRS score irrespective of ability and have pro's racing D Cats until the system catches up with results which would be ridiculous.

Secondly, without the power numbers baseline you could just game the system by purposefully losing a couple of ZRS races to make your score drop really low, then turn up and smash the competition with the power you're clearly capable of. Your score could end up jumping all over the place and that would also be silly.

As I've pointed out to you before, you did 30s power of 7.3w/kg in a race 2 months ago (so within the 90 day period) which will have affected your baseline score. I would assume those who are beating you in C races who have much lower ZRS scores: a) don't have such high 30s (or 10m equivalent) power in the last 90 days as you do; and b) haven't done many (or any?) ZRS races to get their ZRS number adjusted.

I don't understand the 'floor' concept, that seems flawed to me (pun intended) as it's using short term power to fix someone in a category. In a race of 40 mins surely the 30-40 minute average power is a better estimate of where someone will finish? If in C someone can hold 3.4 w/kg for 20 mins, and 3.2 w/kg for 45 mins, they are more likely to get to the finish line ahead of someone who can only hold 2.6 w/kg for the race even if that person does have a big sprint?

Not necessarily, no. It's a 40 minute race, not a TT. In my experience, most races split at some point and it's your ability to be on the right side of that split that has the biggest affect on where you finish - along with your sprint at the end. You don't do the whole race at threshold, so if you have great 2 minute (or 10 minute for example) power, you can avoid getting dropped and be on the right side of any splits, then settle back down in the draft and recover. Then at the end your sprint is obviously also really important.

All this will do again is make encourage people to watch their numbers and not push as hard as they can or race for every position, which is a shame.

Well. some people might do that - it's bit sad though IMO. Just race as hard as you can - if you keep working on that power it might not help you in the race you're currently in, but it could help improve your fitness anyway, and you might then make the split next race. That's how you improve

One day I'm sure someone will come up with a good way to look at the 'quality' of the riders in a race and rank the results accordingly. Ignore power numbers, rank on results against those in the race.

Well that's what they're trying to do with ZRS and the metrics they're using make sense to me, and they seem to be monitoring it and listening to feedback and tweaking things so hopefully it will only get better. But it's also still very early days and as I said, some people you see in C Cat races with lower ZRS scores than you probably have even done a ZRS ranking race at all yet, so focussing on that isn't really telling you much

 

[bridgy]

That's right. It's only the Zwift Labs ZRS races that affect your ZRS score while ZRS is in the test phase.

The only other factor is the seed score, where any race or ride can push up your seed score based on 30s and 10m efforts. There's been a number of people complaining about seed score, similar to the comments from @Whorty above.

I disagree with this. There's been a long history of sandbagging with the CE system in Zwift, where people have controlled their average power and unleashed a big sprint to win or podium. To be fair, there's a weight issue at play here as well, which pulls their w/kg down below a cat boundary. The same issue works against people like me, whose lighter weight gets them propelled up a category, even when they weren't winning at the lower category.

The winners under CE will claim this is race craft. That's fair enough the way CE works, but it doesn't make it a fair contest for all.

ZRS tries to place riders initially where their power figures suggest they should be. Recent replies from Zwift indicate they are going to take a wider set of figures from rider's power curves to place riders more fairly. There may also be more weight given to race power figures, rather than power figures from all rides, where it may be possible to sprint harder at the end of a group ride than you could in a race. I've got some sympathy with this complaint.

But the key point is that the seed is where you START. Once you're racing it's your finishing position relative to others that will determine how you move within the system and the effect of seed should be less relevant. In fact it should only be relevant to stop riders who try to game ZRS. The overall purpose of ZRS is trying to match riders of similar ability.

Of course there will still be those who try. No system in a game like Zwift can be foolproof. And the test phase has shown that there are some issues that need to be addressed. But that's the point of a test phase.

All I can say is that I've found the racing infinitely better in ZRS, than getting my arse kicked regularly in CE. Take a look at the results from Cyclopathy Round 1. The results across all 5 pens suggest the system is actually working pretty well at matching riders.

Good points 👍

 

[bridgy]

I think I've missed a lot of the discourse around the new scoring system. Is there anyway to stop people from deliberately trying to lower their score? Just looking at the guy who came last in my race, max heart rate was lower in this race than his average HR for previous race and power very low, clearly zone 2'd the ride (could have been injured to be fair). So his score goes down? Could someone not just do that fairly consistently, do a zone 2 ride or workout and basically ignore the race?

Well as discusssed above, your ZRS score won't drop endlessly if you just cruise around races purposefully coming last because the base level your ZRS can get to is affected by your 90 day power numbers.

It's one of the reasons why using some power metrics for a baseline, as well as your race results, is a good idea

 

[Del C]

It uses 30s and 10m power over the past 90 days as a starting point, then adjusts your score based on results, and the relative strength of the riders you beat (currently only in races using ZRS ranking). But those 90 day power numbers also do affect how low your score can go.

This makes 100% sense to me for two reasons.

Firstly, you need some kind of starting point for categorising people or you'll have to start everyone on teh exact same ZRS score irrespective of ability and have pro's racing D Cats until the system catches up with results which would be ridiculous.

Secondly, without the power numbers baseline you could just game the system by purposefully losing a couple of ZRS races to make your score drop really low, then turn up and smash the competition with the power you're clearly capable of. Your score could end up jumping all over the place and that would also be silly.

As I've pointed out to you before, you did 30s power of 7.3w/kg in a race 2 months ago (so within the 90 day period) which will have affected your baseline score. I would assume those who are beating you in C races who have much lower ZRS scores: a) don't have such high 30s (or 10m equivalent) power in the last 90 days as you do; and b) haven't done many (or any?) ZRS races to get their ZRS number adjusted.

Not necessarily, no. It's a 40 minute race, not a TT. In my experience, most races split at some point and it's your ability to be on the right side of that split that has the biggest affect on where you finish - along with your sprint at the end. You don't do the whole race at threshold, so if you have great 2 minute (or 10 minute for example) power, you can avoid getting dropped and be on the right side of any splits, then settle back down in the draft and recover. Then at the end your sprint is obviously also really important.

Well. some people might do that - it's bit sad though IMO. Just race as hard as you can - if you keep working on that power it might not help you in the race you're currently in, but it could help improve your fitness anyway, and you might then make the split next race. That's how you improve

Well that's what they're trying to do with ZRS and the metrics they're using make sense to me, and they seem to be monitoring it and listening to feedback and tweaking things so hopefully it will only get better. But it's also still very early days and as I said, some people you see in C Cat races with lower ZRS scores than you probably have even done a ZRS ranking race at all yet, so focussing on that isn't really telling you much

Agree with all of this!

 

[Whorty]

That race you did last night wasn't a ZRS race so unless I'm missing something it wouldn't have had any affect on yours (or anyone else's) ZRS score would it?

Not true. All rides, whether races or free rides, can affect your ZRS score. So if you bimble with a D robo, but blast a sprint, if you hit a new PB in that sprint it can increase your ZRS

 

[Whorty]

What a bizarre race. Only two of us in my pen . The other rider didn't really try.
I had a 2 min lead on lap 1

The lead kept growing, until when I think he realised I would lap him with about 1.5 laps to go. That's when he actually started to make some effort on the climbs. Nearly caught him , missed by around 100 metres before I finished the event

Stupid thing, my race score has jumped 24 points because of the win. But it wasn't a race by any measures. I'm now on 448

So RS doesn't take into consideration the quality of opposition


View: https://youtu.be/wodSHm66DrI?si=2dpS9prWcLSC84Rl

That’s one of my peeves too. Quality of race should be a factor.

 

[bridgy]

Not true. All rides, whether races or free rides, can affect your ZRS score. So if you bimble with a D robo, but blast a sprint, if you hit a new PB in that sprint it can increase your ZRS

OK yeah I know that - any race an potentially increase your ZRS if you increase your baseline 90 day power - agreed.

But you said your ZRS went down during that non-ZRS race and that's what I was referring to

 

[bridgy]

That’s one of my peeves too. Quality of race should be a factor.

It is. From https://support.zwift.com/en_us/zwift-labs-racing-score-faq-BkG9_Rqrh :


Q: How is my Racing Score calculated?

A: We calculate a new racer’s score using a formula based on CP30s and CP600s power outputs–the max power you have for 30 seconds and 600 seconds, or 10 minutes.

We then normalize the power by weight and make further adjustments based on our analysis to ensure the score is fair across various weights and performance levels.

After normalization, we apply a specific formula to transform these values into a seed score between 0-1000.

As you race, your score changes.

A higher finish against stronger competitors in races will increase your score more significantly.

Conversely, lower placements or less strategic engagement in races will result in smaller gains or even reductions in your score.

This method ensures that each racer's score truly reflects their performance and growth in capability over time.

 

[Whorty]

OK yeah I know that - any race an potentially increase your ZRS if you increase your baseline 90 day power - agreed.

But you said your ZRS went down during that non-ZRS race and that's what I was referring to

It did go down, by 6 points. It was 366 pre race and it's 360 now - I checked as based on wkg I was one of the weakest riders yet many have much lower ZRS than me so I wanted to see how the ZRS would change when I'm again beaten by lower ranked riders.
Given the riders in the pen tonight I again expect to be last (I may beat two guy who's wkg is lower than mine but that's only if I can hold 2.7 tonight which I couldn't yesterday).

I would expect my ZRS to drop again after tonight's race but if there is a floor, and this may be 350 (the pen boundary) then I'll be locked into a pen that is too high for me. Makes racing pretty pointless.

 

[Whorty]

It is. From https://support.zwift.com/en_us/zwift-labs-racing-score-faq-BkG9_Rqrh :


Q: How is my Racing Score calculated?

A: We calculate a new racer’s score using a formula based on CP30s and CP600s power outputs–the max power you have for 30 seconds and 600 seconds, or 10 minutes.

We then normalize the power by weight and make further adjustments based on our analysis to ensure the score is fair across various weights and performance levels.

After normalization, we apply a specific formula to transform these values into a seed score between 0-1000.

As you race, your score changes.

A higher finish against stronger competitors in races will increase your score more significantly.

Conversely, lower placements or less strategic engagement in races will result in smaller gains or even reductions in your score.

This method ensures that each racer's score truly reflects their performance and growth in capability over time.

As I said previously, zero incentive to fight for every place if you're going to get penalised for sprinting for 43rd place (which is what I did to get my sprint number, after a 5 minute rest on the KoM descent and run in to the finish). I'd have been better off rolling in and finishing 44th.

It feels like ZRS is trying to appease those who can't sprint by penalising those who can - forgetting that for most of us we don't make it to the finish to sprint because the roleurs and tempo riders have either dropped us already or ridden the sprint out of our legs. Definitely an advantage in the ZRS seed if you have a descent 20-30 minute w/kg but weak sprint.

 

[bridgy]

It did go down, by 6 points. It was 366 pre race and it's 360 now - I checked as based on wkg I was one of the weakest riders yet many have much lower ZRS than me so I wanted to see how the ZRS would change when I'm again beaten by lower ranked riders.
Given the riders in the pen tonight I again expect to be last (I may beat two guy who's wkg is lower than mine but that's only if I can hold 2.7 tonight which I couldn't yesterday).

I would expect my ZRS to drop again after tonight's race but if there is a floor, and this may be 350 (the pen boundary) then I'll be locked into a pen that is too high for me. Makes racing pretty pointless.

The last ZRS race you did was 5th September and it went down to 361 after that (i.e the 360.5 showing on ZP), so that's what it was before yesterday's race. So it hasn't changed or gone down in that race that wasn't using ZRS. I don't know where you're getting 366 from.

Your baseline based on power will change when that 30 seconds 7.3w/kg performance drops off the 90 days of data

 

[Whorty]

The last ZRS race you did was 5th September and it went down to 361 after that (i.e the 360.5 showing on ZP), so that's what it was before yesterday's race. So it hasn't changed or gone down in that race that wasn't using ZRS. I don't know where you're getting 366 from.

Your baseline based on power will change when that 30 seconds 7.3w/kg performance drops off the 90 days of data

Interesting - where do you see that? I got my numbers from ZP. On ZP yesterday pre race it definitely said 366 and 360 when I checked afterwards. Maybe ZP isn't updating 😕

 

[bridgy]

Interesting - where do you see that? I got my numbers from ZP. On ZP yesterday pre race it definitely said 366 and 360 when I checked afterwards. Maybe ZP isn't updating 😕

It's on Companion App in the race results ("Activities" tab). You've never been 366 - in the ZRS race prior you went down to 371, and then 361 on 5th Sept. I think you must have just mis-read 360 for 366 on ZP

 

[Whorty]

It's on Companion App in the race results ("Activities" tab). You've never been 366 - in the ZRS race prior you went down to 371, and then 361 on 5th Sept. I think you must have just mis-read 360 for 366 on ZP

That could possibly be the case, I should wear my glasses more Could have swore it said 366 My bad, sorry

 

[bridgy]

As I said previously, zero incentive to fight for every place if you're going to get penalised for sprinting for 43rd place (which is what I did to get my sprint number, after a 5 minute rest on the KoM descent and run in to the finish). I'd have been better off rolling in and finishing 44th.

It feels like ZRS is trying to appease those who can't sprint by penalising those who can - forgetting that for most of us we don't make it to the finish to sprint because the roleurs and tempo riders have either dropped us already or ridden the sprint out of our legs. Definitely an advantage in the ZRS seed if you have a descent 20-30 minute w/kg but weak sprint.

The way you word this sounds like you're taking what is basivally a mathematical model very personally! I don't think ZRS is "trying to penalise" anyone, it's trying to put riders with similar abilities together in races that's all.

I know it's not working for you right now - because of that anomoly of a performance in an earlier race, plus the fact it's early days - but I think that the basic principles of how its set up are sound.

I probably fit into that description you put of having decent 20-30 minute w/kg but a weak sprint and I'm seeded by ZRS in exactly the way I'd expect - basically in ZRS Cat B based on the ZRS score boundaries. Not seeing any "advantage" there.

What it will be doing is potentially under-scoring people who haven't done particularly high 30 second or 10 minute power in the past 90 days (eg people who haven't raced much over the summer) but are capable of better. Those people will see their ZRS score gradually increase if they do ZRS races and do well