understand the reason behind decay, but it’s really not very sensible when you have so many people racing ladder and ZRL but none of their data goes into ZRS. I raced in the 390-510 with a previous ZRS of 416. I finished 16/43. My ZRS fell to the lowest of all people who finished the race….-39 down to 376 (this is 100% illogical), lower than the score range I rode and finish high up in…FTP went up in the app and velo went up to Amethyst. I’m now pointed to racing in a lower ZRS category than the one I finished top 35% in, this which in turn create even less need to seek PBs, hence an endless loop of complete illogical practice. This is not a complaint about specifically my score, it is to highlight that this new method is illogical as proven above - if the ZRS score isn’t influenced by the race result why is it a racing score - my understanding is my score will only really move up when I can add some PBs since most other data now not included as too old and other racing not included either. Doesn’t seem to reflect ‘racing’ very well where tactics are in play not only hammering yourself. Sent in good faith that you consider the implications of decay and its inappropriateness for vast numbers of riders, therefore impacting other riders and who they race against.
The reason they would do this is their fundamental misunderstanding of the purpose of the racing ‘score’. It isn’t needed as a prestige system but as a sorting mechanism.
But all zwift comms make clear they view ZRS as a prestige system. If it is, then it makes sense to decay peoples score if they don’t ride ZRS races.
If they want ZRS to work as a sorting system, then decay is either unnecessary or should only be in play where someone does no activity at all, including outdoor rides logged on zwift. There will still be returning riders with 150 points off their score dominating races….
And the extra frustrating thing is that OpenSkill can easily be used for both, and doing so would work very well for zwift where both skill and physical ability are key components to performance.
Mu for matchmaking, Mu - 3*sigma for ranking.
As someone competes more often and more consistently, their sigma will reduce so their ranking will increase. Obviously Mu moving up and down with results too. If they take time away the system can increase their sigma and their ranking will drop.
If someone goes away Mu would still be used for matchmaking, so they come back where they left. And because sigma would be higher their score would drop quickly if they couldn’t meet their previous performance level.
I do wonder if this bonkers decay was re-introduced, this time with activation only if a racer doesn’t race another ZRS for 30+ days (while perhaps doing great in ladder races; WRTL; TTs; workouts; freerides), because ZwiftHQ are seeing a number of riders get a ZRS boost for performing as expected in a ZRS race and then getting a pen promotion… So they then stop racing ZRS events?
I do hope that ZwiftHQ haven’t seen the failings of ZRS, and there are many failings and decided this decay is a suitable bandaid.
How long do we have to wait to simply stop decay if a racer uses Zwift at all?
Or better yet, ditch the whole ridiculously implemented ZRS system and adopt a vELO system similar to https://www.zwiftracing.app ?
This is what I don’t understand. From my understanding of OpenSkill there are a lot of things that should be working better than what we’re seeing (i.e. you should actually see some relation between who you beat and what your score does, smaller field sizes shouldn’t always provide bigger swings, tanking by riding up a level shouldn’t really work, etc.). But if they are always using Mu-3sigma for the number we’re seeing on screen then it does make it hard to actually understand what ZRS thinks your skill (or mu) is.
Well, this topic closes in an hour, so I’m sure all will be sorted by then 
This topic was automatically closed after 13 days. New replies are no longer allowed.