Great news zwift. Really looking forward to seeing how this plays out
1 Like
Is it directly taking the zrCS number (which uses 5 minute power) or is it still 10-minute based? Either way, really happy with the direction this is going and the quick responses to feedback and iterations.
4 Likes
It isn’t silly because of the way OpenSkill works. Having a decent number of races will give much greater confidence (smaller sigma) in where people are at. It’s like having a flying start instead of having to wait for everyone to do a few races.
There’s no question that some of the heavier sprinters who were celebrating yesterday with their sub-400 scores will now be put in a more realistic place. But then so will the lightweight and female riders who were severely disadvantaged by V2.
zrCS is proven to be a good seed, so this should be a very good change.
6 Likes
Well this is interesting … looking forward to hearing more on this.
Thank you Ryan (and all of the team behind the scenes making this happen!) So happy to hear that a compound score is going to be used, particularly the version used by ZR.app. No course on Zwift is perfectly flat or a perfect climb, so having a balanced blend of both watts and w/kg to score riders really is the way to go.
This should cover pretty much every extreme outlier out there while also doing a reasonably good job of initial matchmaking. Great work to iterate so quickly!
4 Likes
This sounds like a great change to me!
1 Like
Progressive.
1 Like
Wow great news … talk about 360° from v1 to v2 and not to v3
1 Like
It’s a factor in calculation - the calculation is neither 5min watts nor 5min w/kg.
just 5 minute
Not identical but very close. This is an improvement on zr.app which I assume will likely roll out to zr.app as well soon
2 Likes
So the changes we witnessed last few days were what?
They were a mistake? Because us lightweights were over seeded.
Today I raced against flocycle who was seeded with 999 zrs because he is 56kg… absurd
1 Like
Just think for a moment, where all this will end.
If ZHQ come up with a racing score system and it is allowed to run.
I think most of not all riders will be clustered around the category cut offs.
You will win a few races, move up to the next cat, lose a few and drop back down.
The past few weeks, I was in group D and my score was:
V1 194 and increased every race to 210.
V2 seed score 235 and reprocessed to 283.
All these numbers made little difference.
I stayed in the same group and raced the same people and probably beat and lost to the same people.
I would continue to gain 3-5 points per race until I crossed into group C.
There was one guy who beat me in a sprint and he went above 350 and into the next group.
He probably won’t be strong enough to win group C so he will be come back to D sometime.
This is going to be repeated.
Except for seed score floors , there should be few people in the middle.
A good anti tanking program should eliminate the need for seed score floors (they should be eliminated anyway).
I do like seeing how responsive ZHQ seems to be.
After we get the seed score straight, how about Burrito power ups for robopacer rides😃!
Wouldn’t that be a “hoot”.
1 Like
The seed score change does not correct the problem surrounding RS 11, and least not for all cases. Any update on that?
I do have contemporaneous and post-change Events/View/ZID data for one race series back to August 31 if you’re interested.
Thanks
Closing down this thread as we start our Seed V3 release. Please move any further discussion to the new thread. Thanks for all of the discussion and feedback! Zwift Racing Score - Seed V3 Release [December 2024]
3 Likes