Should be easy to figure out if it wasnt for the fact that Zwift hides almost everything for us to see.
Even the ZRS we see before the race isnt the actual one since it only counts from the last race everyone did and we dont get to see the decay.
Wondering if this system is so basic and easy that is why they are hiding it all and we wont be laughing at all those Phd professors studying data for years to come up with this …
too complicated. my theory is that the guy who writes the FAQs for racing documentation on zwift has a way with words
Well I had a race a week ago or so, and my racing score ended at ~215.
Yesterday when doing another race it started at ~240 and ended at 249 after the race.
It would be very good to see how the score is developing through time, because I’m not sure what influenced the score like that.
New 30 second and 10 minute PBs which you set in between races is the most likely cause of your ZRS change
Check out your Power Profile and see when you set the most recent 30s and 10min PBs
Zwift have said that they’re working on that - Zwift Racing Score - Next Phase! [September 2024]
In basically all the races I overlooked in the Companion App so far I saw that there seems to be a simple general rule, the upper 50% of the riders improve their score, the lower 50% receive a decrease in their score. Is this really the case? Because if you don’t improve this algorithm you will end up with most drivers having their score around a category threshold. I even plotted this with Phyton to confirm my hypothesis. So most drivers will eventually go up a category, perform in the lower half of this category and will go down again. There would be a huge bunch around the four category thresholds. Please tell me you improve this logic and I am wrong.
The solution to this is to implement a rule that changes your score relative to the drivers and scores around you. So you might get 5th in a 50 rider race but still receive a decrease in your score because everyboby in front and around you has way lower scores and vice versa.
So far today, times are CEST and where possible data is from CA.
01:10 - 6 entries, 3 finishers
02:10 - 5 entries, 3 finishers
03:10 - 0 entries, 0 finishers
04:10 - 0 entries, 0 finishers
05:10 - 0 entries, 0 finishers
06:10 - 1 entries, 1 finisher
07:10 - 1 entries, 1 finisher
08:10 - 6 entries, 6 finishers
09:10 - 0 entries, 0 finishers
10:10 - 3 entries, 1 finisher
11:10 - 1 entries, 1 finisher
12:10 - 27 entries, 18 finishers
13:10 - 16 entries, 14 finishers
14:10 - 14 entries, 11 finishers
15:10 - 7 entries, 3 finishers
I think it is more 60% goes up and 40% goes down. There seem to be a little more going up than going down.
And most of the times there is one or a few who differ. But most of them are easy explainable by last races and so decay, summer rider returning, or havent raced at all for a long time.
Is this observation right? When you finish the race 30s later, that is your highest score. Then as weaker riders finish their race the mean score of the race drops, therefore your net score gain goes lower. This goes on till the last riders finish, so your score continues to drop till that point.
Not from experience… It goes the other way.
Sunday +1 directly after the race, +5 (roughly) after the re-calc.
Its not as linear as that, riders are moving up & down outside of that pattern but there some oddities.
Things like the following (made up numbers):
550 (+3)
456 (+3)
600 (+2)
590 (-1)
560 (+4)
In this example I would look at the history of the rider that went down. Most of the time it is explainable why he/she differs from the rest.
ZRS doesnt get updated daily, so we dont see the daily decay changes. If someone hasnt raced for days or weeks, his actual ZRS is a lot lower than we see. That could explain rarities within the scores.
But overal that 60% / 40% is pretty acurate.
Is that actually correct?
Race movements may be trying to show movement from last race result score to current race score, which IMO does show odd score movements due to PBs and decay.
But if you check out their ZRS score immediately before the race start isn’t it showing you their current ZRS score not their last race score?
What’s the daily decay rate these days? Still 0.25?
My score hasn’t dropped since racing on Saturday.
As far as I know we dont see that. The next time you race it gets updated and starts at the number we cant see now.
As far as I have seen it is. The daily decay doesnt show on the website or the CA.
If it gets updated the moment you sign up for a race I dont know.
But all this speculation could have been prevented if Zwift was a little bit more open how the scoring actually works and shows us the points gained or loss after a race and not just a new total.
I think what Ian means is if you proactively look at the racer’s profile ahead of the race it ‘should’ show their current racing score including decay.
I also think he’s agreeing that the up/down arrows in the racing results are the change since last race rather than starting from what you had at the beginning of the race.
So if all you’re doing is looking at the races results and you see a down arrow (where the people around that person had up arrows) it can be the case that they had some decay between races that was more than they went up during the race.
The two riders I looked into who went down around people who went up were specifically that case (folks who hadn’t raced a ZRS race in weeks and were close to the halfway point in the race - so maybe decayed by 2, went up by 1 in the race, and Zwift showed a down arrow in that case sorta thing).
It should indeed. But with Zwift you never know 
Question though. You can only see the racers ZRS by clicking on their name on the CA right from the race they signed up to. Cant see it anywhere else or am I missing something.
Ok, checked with a friend who hasnt raced for weeks. So I asked him to join a pen.
ZRS before he joined the race was 216
His best 30s and 10m were from August and September. He didnt race since the reset so no new ZRS updates.
After he joined the pen his ZRS stayed at 216.
So no decay to be seen. So my guess is really that it is counted after the race and not before, so we cant see it.
He cant race right now, but I will also check if a DNS also triggers a recalculation.
So “for science” I did the 1710 Festival race, 10 laps of Downtown Dolphin, at a fatigued z1/2 pace and finished last (86th?) in the 180-350 pen.
Provisionally, my score dropped from 280 to 259 (should drop a bit more as there were ~5+ DNFs), which isn’t 85% of my seed score (which I think is ~227 as I reset to 266).
Funny guy who won the D pen, showing how bad initial seedings can be, provisionally increased to 616! 
