Pretty static yes but they did make a change, not for the better, and organizers had the ability to customize the category boundaries but that was painful for both staff and organizers.
In the early months, zftp was based on predicted 20min power. It was how I was briefly promoted to CE pen A in '22, around summer iircâŚ
Oh how things can change!
Yes. Organizers could request custom limits, but those limits couldnât cross the A/B/C/D boundaries, only be within those.
At the point we set up Herd Beginner Racing, anyway. If that changed afterwards I didnât know about it.
For zFTP, true, but there were other options. We copied from @DejanPresen in one of our events.
(subgroup.label == 1 && (powerCurves.powerCompoundScore >= 2100 || powerCurves.zFTPwkg >= 4.1 || powerCurves.zMAPwkg >= 5.0 )) ||
(subgroup.label <= 2 && powerCurves.powerCompoundScore < 2100 && powerCurves.zFTPwkg < 4.1 && powerCurves.zMAPwkg < 5.0 ) ||
(subgroup.label <= 3 && powerCurves.powerCompoundScore < 1750 && powerCurves.zFTPwkg < 3.6 && powerCurves.zMAPwkg < 4.4 ) ||
(subgroup.label <= 4 && powerCurves.powerCompoundScore < 1400 && powerCurves.zFTPwkg < 3.2 && powerCurves.zMAPwkg < 3.9 ) ||
(subgroup.label <= 5 && powerCurves.powerCompoundScore < 1100 && powerCurves.zFTPwkg < 2.8 && powerCurves.zMAPwkg < 3.4 ) ```
1 Like
Pretty sure that changed after that but you are right when first released that was a problem. Mute point now really with no public CE races
that was the date they cut off CE races for organisers, the reset was the date it went live in my opinion
Wasnt that October 7th ?
Cat 350 - 520
Signed up - 19
Started in pen - 13
Finished - 11
Place - 4th
ZRS Score +9
Finishers ZRS - 5 down, 6 up.
Using the web-based tools (zwiftâs events page, and zwiftpower).
52 signed up
28 finishers (only 25 show in ZP)
Using LIVE tab in ZP, there are 9 entrants who DNF
There donât appear to be any DQs
28+9 = 37. Thatâs 10 off from your 47 number
Could those 10 be non-ZP registered riders? Or riders that were at the starting pen, but never crossed the start line?
Old formula to determine # of up vs downs that Iâve used would be:
52 signups Minus 9 DNFs = 43 / 2 = 21.5 ups/downs
Iâm pretty convinced that the general formula is 50/50 up/down - but determining the number to use is a bit trickier. Aaron showed a graphic that had 3 numbers of racers - 52 signups, 47 finishers, and 28 included in the final results. But none of those numbers is what they use. You need the number that actually started the race, which isnât (easily) shown anywhere. Click on one of the images taken from the race and look at the rider listing. For the race he showed Iâll bet that there were about 36 or 37 people that started the race (because the up/down looks like it changes at 18/19).
What are some of the defining goals for ZRS to see is it a success or failure?
Todayâs Coastal Crown Loop shows the disfunction of ZRS. When riders are 1.5 w/kg in difference in power for their 10 min power, there is no way to survive the initial climb. That 30s power doesnât come into play when you are 2 mins behind the front group after 10 mins.
2 Likes
The screenshot I have saved to strava shows 46 people (Iâm 12/46 at that point), this is at the 10km point for that race. So there were at least 46 that started, but that one person might have quit after starting and before the 10km mark, so the starting number still might be the 47 maybe?
So the numbers now are the three in that screenshot, and the rest:
- 52: Which I think is number of sign-ups?
- 47: Either the number that were in the pen, or folks who started the race?
- 46: Number of riders in the screenshot at 10km into the race (did one drop out since the start maybe?)
- 28: Number in the results in Zwift Companion app.
- 25: Number that Wannie mentioned showed up in ZP.
- 9: Number that Wannie mentioned showed up as DNF in ZP.
So my assumption is the rest of the riders who DNFâd were not on ZP. That would mean 9 riders DNFâd on ZP, and 10 or 11 riders DNFâd outside of ZP (either once the race started or in the pen). That is rational since about half the folks donât show up in ZP (25 of 46 or 47 were in ZP) from this race, so about the same amount of non-ZP users DNFd.
That does seem like a lot DNFs. I might go back and average DNFs from my last ZRS races and my last CE races to see if thereâs any big pattern there.
For me it would be that ZRS solves the problems I had with CE and so far it has failed in every aspect or made it even worse.
We know that one question they asked (which was posted on Zwiftinsider) was whether or not people enjoyed ZRS better than CE. So we can infer that is a clear goal for Zwift (for that survey to come back as more positive than negative).
So, that is one qualitative metric.
I would also assume that whichever PM is in charge of racing has some quantitative metrics based goals such as increasing racing sign-ups as a proportion of active Zwift users, decreasing DNFs in racing on average (if itâs funner presumably there would be fewer DNFs - the portion of ârage quitâ DNFs), there might be a metric that is based on race organizers as well (both a qualitative one and quantitiative one. i.e. They should probably be asking race organizers if they prefer ZRS as a form of qualitative result, as well as have some form of metric in terms of race organizer participation ensuring there is no drop etc in some way shape or form.
Thereâs possibly a metric for average races per Zwift racer to see if on aggregate people who currently race opt to race more with ZRS than CE.
Anyhow, these are just conjecture, but they would be a few things I would look at if I were the PM on this project. They would want both qualitative, and quantitative values to see if racing has become more fun and popular with ZRS over CE.
Itâs the right question. What was the goal?
If you were a customer focused org that embraced the joy of user experience work then youâd probably go for âensure that joining a competitive race is seamlessâ and âracers who do well in competitive races are ranked higherâ
So metrics like âhow long does it take to find a race and join itâ considering time to read the description. Or number of active cat systems = 3 and percent of races using ZRS = 95%.
Iâd also add a Velo comparison and CE comparison metric to say there is agreement on cat and field strength by around 90%.
Maybe some repeat racer and racer retention metrics.
Maybe a 80% of riders try within 85% of their best ability per race.
Some potential snark metrics;
-
If this thread makes 2k posts
-
number of angry YouTube rants
-
Number of âcareer B winnersâ forced into A
I think one of the funniest aspects of this is that people who are getting upgraded are struggling to lose enough to get demoted because there are so many people trying to lose.
Nice analysis.
Something to consider. I checked out a race result today. Out of 20 race finishers 7 had either an up arrow or down arrow but their score actually didnât move.
Zwift doesnât currently have an arrow system to show no score movement.
They should just show how much everyone in the race goes up or down.
Something like +13 (534) with 534 is the new total score.
1 Like
I âthinkâ they do but itâs Up and itâs just 0 = remove your decay % for the day. Iâve had this happen 2x where I have the score from yesterday with an up.
5 point bump today for 2/3 with no DNFs. 11min ahead of 3rd and I had higher power than the winner but got dropped on the grade. Clearly, the 5 points is a skill-based reward. I had the highest ZRS in the race.
1 Like
Or, thereâs always score movement at the decimal level due to decay, and if your score went down 0.1 between races and you got 0 on the latest race it still shows a down arrow?
So if they store your score as a decimal then most folks wonât get exactly 0 since their last race due to decay etc.
As mentioned it would be much better if they just showed how much your score was affected by this exact race along with the new score.
3 Likes