Racing Score Updates [Sep 19 2024]

Racing
1

We’re excited to share another batch of improvements to Zwift Racing Score as we prepare to graduate the feature from Zwift Labs. We are so grateful for all of the participation, discussion, and feedback as we continue to tune and polish.

The top 3 finishers in a race will receive a podium bonus. This is meant to help move racers up a category faster if they are regularly winning. Podium bonuses will be awarded to 1st, 2nd, and 3rd place finishers. The bonus will only take effect if there are at least 5 racers in the field and will not kick in for the unlikely case the score decreased. This change is in the final phases of testing and will anticipate this to be enabled early next week.

We have defined the subgroup score ranges we will leverage for Zwift-owned race series and recommend for any community organizers. These ranges were decided based on factors like potential field sizes, probability of winning between max and min scores, and even how previously used power-based A-D pace groups distributed into the new categorization. For more information, see Changes to Score Ranges on Zwift Owned ZRS events

Scores are now calculated closer to real-time. Before this, scores would finalize after 12 hours, and would only then take DNF’s out of the field. Now scores are calculated and adjusted as each racer crosses the finish line. This will eliminate issues where individual scores would drop when checking the day after a race.

Zwifters are now required to do three 10 minute activities within 90 days to be eligible to race. This is to ensure the activities are long enough to give us complete data for the seed formula. We realize that there is still work to be done to ensure that those activities are going to produce an accurate seed, and we plan to continue making improvements here.

ZwiftPower league results will support score range filters. This will allow users to continue tracking General Classification (GC) standings after we transition event series to score-based subgroups. This change is in the final phases of testing now and will be released soon.

We’ve fixed an issue where scores that drop too low get stuck with low values even after winning several races. We realized that due to the way our scores are calculated, it was possible to have very minimal score movement from winning a race if the score is too low. We’ve made changes that prevent scores from getting stuck at a low value if consistently winning and reprocessed scores today to reflect these changes.

After graduation from Zwift Labs, we will continue to make improvements. Some of the next racing score features we’re investigating are:

  • Enhancing the seed algorithm to incorporate a broader range of historical data, considering the performance of racers who are transitioning back indoors after an outdoor riding season
  • Measures to prevent intentional score manipulation
  • Category enforcement applied when joining an event instead of at sign up
  • An historical view of how your score has changed over time

Thanks again for participating in this test period and providing feedback! The data we’ve been able to gather from this group has been extremely valuable.

22 Likes
2

Appreciate the update ryan, glad to see support for alternate race formats is coming

3

One of the best things about this system should be that the pen boundaries can easily change, so an individual might be towards the top in one race and the bottom in another. Static boundaries might have some value for a series with a GC element, but other than that you should be encouraging community organisers to use different boundaries.

Other than that, good updates that should improve the system. In particular not having to wait 12 hours to see the actual results will prevent a lot of confusion.

4 Likes
4

Why fixed ranges for pens?

This is perpetuating the mistake made with the old w/kg-based A-D system, where the pens were fixed instead of being mixed up from race to race.

I don’t buy the argument that people should always know which one they’ll go into each time.

It would be so much better if racers didn’t always slot into the same pen, and that should have also happened with the previous system.

Why limit things in this way and encourage score manipulation to stay towards the top ot a pen instead of being in the bottom of the next one up?

6 Likes
5

dynamic pens are the way but look at the moaning and whinging on here already, imagine if people didnt know who they were racing then looked it up later and saw a huge disparity, be carnage

1 Like
6

3x10min rides is rather an odd and very specific requirement?

With no minimum effort level, what do the other 2 efforts give you data wise?

Edit to add -

Another vote for dynamic pens - the insistence of fixed boundaries is frankly laughable having seen the issues from the previous systems.

7

for zwift owned events. community organisers can go buck wild with whatever pen distributions they want. there’s quite a few different ones out there already

8

but they are still fixed boundaries allowing manipulation of races to stay winning trinkets

1 Like
9

of course they can but i don’t think ZHQ giving recommendation on how organisers run their events is a good thing.

What we need is more flexibility on what organisers can do which will grow racing.

7 Likes
10

yeah, i do agree.

11

Instead of this why not just go back in time past 90 days to get the last 3 10-minute+ activities? This way someone who primarily races won’t have to do a random 3 non-race activities each new season, and they will still be seeded using the last 3 efforts on Zwift. If someone has done 200 races in Zwift and starts a new season it seems non-ideal to require them to do 3 non-race activities in a new season rather than using the last 3 races they did as input to the seed.

3 Likes
12

Ryan said above:

We have defined the subgroup score ranges we will leverage for Zwift-owned race series and recommend for any community organizers.

So we all know what’s going to happen there. And why fix them for Zwift-owned series?

13

i won’t speak for anyone else but myself but i am of the firm opinion that community organisers should be able to set the boundaries however they wanna set them

1 Like
14

I can see where they’re coming from with the last 30 day requirement but as someone who mainly uses Zwift to cross train during cross country season, it’s great to just hop into a race whenever I feel like it.

15

We were already requiring 3 activities in the past 90 days, this just ensures that those have 10 minute power curve data, which is required by the seed formula.

1 Like
16

We’re doing some data analysis on this right now, and this is what I was referring to with the first post-graduation feature I mentioned. The challenge is in considering all of the different use cases (ex: new Zwifters without any data, injuries or low training load causing a significant drop in fitness over the summer, etc).

17

If it’s not driving a more accurate seed score then it just seems like a barrier to entry.

It seems a barrier to the customer with no actual gain over doing 1 10min ride?

Just seems an odd change.

Does a 30min ride count as 3 10min efforts? Or does it need to be 30 mins in separate activities.

18

’ * Category enforcement applied when joining at event instead of at sign up’

Finally :+1:t2:

8 Likes
19

Hopefully Zwift will continue to offer the split “Low End” / “High End” options to effectively provide more than 5 (ie. 10) pens?

Is this same ability going to be offered to community organizers if they so want? Or alternatively, perfectly ok for a community organizer to focus on a narrower band of ZRS for their event (eg only from 300-700 ZRS)? I don’t know, is there a requirement that a community race be all-level encompassing?

20

the Best line in there is this * “Category enforcement applied when joining at event instead of at sign up” about time. @Gordon_Rhino-Racing @S_A_ccc

5 Likes