Racing score is 4 months old ( approximately ) so how's that going?

Racing
1

Roughly 4 months down the road from the replacement of Category Enforcement with ZRS.

back then Zwift said.

“Using our 1-1000 scale—where your score increases as you improve—we can recommend the most competitive band for you, ensuring a fun, rewarding, exciting, and fair experience.”

I don’t know about that…

I’m still seeing riders with very high power to weight figures with very low racing scores

riders very obviously managing their scores to remain in lower tiers

nothing apparently more scientific than
For the top half of finishers in a race the score goes up and for the bottom half, the score goes down.

no apparent weighting for the relative strength of a field

weird score outcomes for small field races

and an utterly incompetent version of a monthly race series where the result from a race at any level counts in the lower bands and riders disappear and reappear from a band (Cat ) during the month ! In January the top 2 riders achieved some of their best times racing in the equivalent of Cat A races to win the approximate equivalent of Cat C

some riders may be having a hoot, but is this new system an improvement over the old one for the majority.

many people didn’t like Category enforcement - fair enough

in my opinion ZRS is actually worse - and I was all for a results based system when it was introduced.

if I chose to, even I, could ‘fix’ my score in this system - that’s saying something - its so easy and blatant, It cant be called cheating - its almost as if it was intentional.

have Zwift given up trying to fix it?

or do they actually think its working ok?

13 Likes
2

Given that there are now regular six hour blocks with literally no races that I can enter with a score of 802, I’m going to stick my neck out and say it’s not going well.

8 Likes
3

Now that they’ve added the offset categories for the ZRacing series, I’m very happy with the changes. I’m no longer jumping back and forth between C & D (getting blown out or feeling like I’m sandbagging).

2 Likes
4

Getting my :popcorn:

3 Likes
5

I’d say we don’t actually even know how it’s supposed to work at this point. If you look at OpenSkill it seems it should behave reasonably in situations where ZRS does not.

But in Zwift:

  • You get bigger points swings in races with small fields
  • It seems the “skill” (or ZRS) of the people around you does not matter as much as whether you’re top or bottom half of the field.
  • If you join a mass-start race as a D cat rider (and many Cs) you are mostly guaranteed to drop score.
  • If you join a mass-start race as an A cat rider (and many Bs) you are mostly guaranteed to increase your score.
  • If you race up a category you are more likely to drop score.
  • They still haven’t really figured out how to handle Tiny Races.

It is unclear if any of the above is “by design” from Zwift’s perspective because they have not communicated what should be expected.

That said, the introduction of different ranges was an awesome change, which they could have done in CE or ZRS. Anything they can do to keep the ranges either smaller, or less static in general would be a good thing.

6 Likes
6

I don’t think ZRS works too badly for me. TBH, mine bounced around a bit before settling at 570 or so, and nothing much I do seems to change it. vELO shows more volatility but this is currently drifting down from an artificially high level driven by 5 minute power.

My ZRS seems roughly consistent with vELO, assuming the latter is on a scale of 0-3000 or so.

No comment on race bandings, calendar etc. beyond observing that the restrictions on race availability on high ZRS riders appear unhelpful.

Where riders are improving rapidly or where they are deliberately managing their ZRS down, it seems clear that ZRS is not good vs vELO as it produces for more counterintuitive outcomes.

1 Like
7

Rider 1 just blown the field by over 1 min 40 second and have a seed score lower than half the field. That says a lot.

IMG_7005
IMG_7005873×1801 191 KB

8

With just this screenshot it doesnt say anything.

2 Likes
9

If a rider is beating dozens of riders by over 1:37 min and still have the same seed score, that doesn’t say anything? You are not asking the right questions.

10

I can not see the history of that rider from just this screenshot.

But since he upped his seed score in this race my best guess would be that he hasnt raced in a while.

11

He has 5 min power of 4.5 W/KG and a score of 303 in Cat D :grimacing:

12

I am really not going to go through all the ZRaces to find this race, so I will take your word for it.
But still probably someone who hasnt raced much. And without the correct data its no wonder Zwift cant calculate his true ZRS, since as far as I know, they still dont look further back than 90 days.

13

You miss the point, he has a score of 303 underlined which means he set a new power record for 5 minutes at 4.5 W/KG so he can still compete in the same Cat D with 4.5 W/KG 5 minutes which is absolute crazy.

1 Like
14

Just use the ZRS Score site to see if there is anything wrong. They figured out the formula :
https://peca.vigred.com/blaz/zrs2.php

The formula is (cp300 * cp300/w) * (65 / (65 - (0.5 * (65 - w)))) * 0.364 by the way if you are interested. Where cp300 is the 5 min watts (not w/kg) and w the weight of the rider.

2 Likes
15

Thanks, will have a look.

Edit- I cant access the site.

16

Hence the system is broken.

Rider should have 471 score.

1 Like
17

That sounds reasonable, since with 4.5 he is in my range and I am at 456

1 Like
18

It depends on weight, I have seen 60kg riders with 4.2 w/kg below 350 points in Cat D.

19

But they have terrible pure watts and no sprint, so beat them there. It is not all about w/kg.

20

On hilly courses you cant beat them at the sprint if you are 2 minutes behind them at the finish.