We’re excited to share another round of Zwift Racing Score updates!
Testing and development of the new seed formula continues, and we are targeting a rollout in the next couple of weeks. We have two candidate formulas identified and are in the final stages of validation before implementing the change and rolling it out. Over the next week we will be finalizing what the release will look like. To reiterate, we will not do a score reset, but may do some reprocessing of previous races in order to fix instances of over/under seeding. We will have another update with more specifics on these logistics in the next week. [UPDATE]
We are simplifying the race requirement by reducing it to a single activity, down from the current three. We’ve received feedback that the barriers to race are too high. We still plan to look further back than 90 days when seeding new racers, but in the meantime we are removing barriers for getting started or back into racing.
We’re finishing up testing on the change to require a minimum effort in order for ZRS to decrease. This took slightly longer than anticipated to develop, but we wanted to take the time to ensure the mechanism catches instances of intentional score tanking as accurately as possible. We anticipate this mechanism to kick in before the end of November.
We will be revising our default event category ranges to better align to the score distributions after the seed formula updates. We’ll also be taking this as an opportunity to find ways to make the subgroup ranges smaller, and therefore more competitive.
Category enforcement when joining events is coming. This feature is mostly implemented and is currently under design review. However, since it involves a change to the game client, we need to wait for the next game version, 1.80, scheduled for release between December 17 and 19. This functionality is expected to start appearing with events in January.
Other features and improvements on deck include:
Improvements to volatility of score movement at lower ranges (est Dec 2024)
Looking further back than 90 days for the seed score (est Jan 2025)
Arrow indicators when min effort mechanism kicks in (est Feb 2025)
We are committed to continuing our cadence of monthly updates via forum posts on our racing roadmap and updates. As mentioned above, we are planning another out of band update in about a week with more information on the release logistics for the new seed score.
Since it’s impossible to require a certain effort, what is the upside of demanding any activity at all? Just let someone race right away. ZRS should quickly adjust based on efforts during races.
I am so curious what you have come up with for this. And how a computer sees the difference between someone doing it on purpose and someone who really cant do better, or people who are genuinly getting dropped and riding solo.
And does this also effect people who are wining but are masters in drafting and so have a very low w/kg for the race but won the sprint.
But, these are all nice en welcomed changes. Rather have them in place tomorrow of course, but now knowing a timeframe is nice.
What’s the estimate for when ZRS will be taking Points races’ scores into account? There is a LOT of points racing on Zwift, including what are likely the highest quality races in Zwift Racing League, which should have the highest quality data as most riders peak for those races. Many riders ONLY ride those races regularly, and ride scratch races only rarely.
How will one be able to see their new seed score? You really should consider a full reset if it’s a significant change, and if it isn’t a significant change, there’s no point.
@Nick_ZHQ any estimate on being able to view the history of our ZRS ?
And other things that popped up :
Will the score be changed at the result screen so we can see how many points we gain lose something like +7 (356).
When will Zwiftpower be adjusted to ZRS ?
Will our seed score always be visable somewhere. Even if it is on our Zwift profile website.
If a strong rider has no ZRS and races in a low pen their score will take a long time to adjust to their actual score because in a low pen they would be racing at a level that is nowhere near their max, so there will be no max seeding efforts, and at this point we’re only seeing like +20 points (give or take) for winning races, so getting there via podiums will also take a long time if you start in a low pen and manage your power to be just at the top end of that pen too.
Forcing any random activity is also not a great solution because if someone is told to do an activity before racing they will not do a max effort anyhow, but the system as is will not work well if all new riders can just enter the lowest pen (this would be mostly a problem for the lower pens).
I would guess that the best thing they can do is to largely expand the 90-day window to include activities from the previous season rather than start everyone fresh in the new season, which they say they are estimating for January. For racers with no data at all, those will still be an issue, but hopefully a smaller set.
Outside of ZRL, who is putting on points races? I only ask as I think Zwift should sort out the racing score situation first and then worry about the intricacies of points races.
I am wondering if given that power efforts at different intervals (especially short term sprint efforts) play into the racing score, whether mid race efforts for points would show up in racing score results even if finishing points are not captured.
Any chance for more daily/weekly variability in the category ranges? As long as the defaults are static, and broadly adopted people will be able to manage their efforts to stay below a set number. If there is some variability in ranges that would become more difficult.
Does this link with the overall scoring going in the opposite direction of where it should go? Or is this something different?
Seems seed is still the focus and points distribution also needs work?
Today you can sign up to a race on the calendar in the future with your current ZRS, your score might go up and at the point the race starts your score could potentially be too high for that category.
I think this line item is to check category at the time you enter the pen so if your ZRS is no longer in the correct range it will tell you then that you have to move to the next cat up.
So before you could sign-up in advance and it wouldn’t do any additional checks at the time you race, I think this says they will now do the category enforcement when you actually go into the race that day.
What is a subgroup range and how do you make them smaller when the point scale is 1-1000 in size?
As has been suggested many times, why have ‘default’ ranges in the first place? Random and myriad different pen size definitions seem to be highly desired feature request, and solves a lot of issues in itself.