Then it has to be a bug and they have to correct it.
As you say you have only done 5 activities so quite easy to find that you fully completed the FTP test lite. Completing this test might give you zMAP of around 243 possibly but certainly would give you a 5 minute power of 225+watts or around 4.8 w/kg. How did it feel, effort wise, when you were doing the FTP test?
This 5 minute power figure might be enough to give you a ZRS of 414, I donât know but possibly @DejanPresen could confirm as he has a very good idea about ZRS formula.
The puzzle for me is where you are seeing April information unless by chance your Herd Ride was on 4th December and you are seeing power profile activity with a date of 4/12/2024 and taking that to be 12th April rather than 4th December ?
2 Likes
That indeed appears to be the case. I can only see one ride for her in ZP, on 4 December.
Also, her Ramp Test Lite had a max of 261w.
1 Like
Youâre not. Zwiftpower uses US format dates, so your ride on 12/4/24 is 4 December, not 12 April.
Also your Ramp Test Lite went up to 261w, so I donât think thereâs anything wrong with your zFTP value.
1 Like
Along with your ramp test showing a 5mins of ~240W on 6th November, the first ride visible on Companion app from 5th November shows average 194W for just over 26mins, with a max of 340W.
All activities are used to decide your 5mins seed score, not just races, as was the case with the old Category Enforcement system that began around Feb '22.
At a listed weight of ~53Kg, itâs no surprise you have a v3 seed score of 414.
The average power on your 5 November ride and your 6 November ramp test lite look hugely more than your later rides. Are you sure that someone else didnât ride your trainer without you realising? These two rides/tests are the reason for your current zFTP and your ZRS.
1 Like
What happened to the idea of scoring based on where you finished vs other riders and their racing record? What youâre saying now just sounds like a pure power and weight based system which doesnât sound like a whole lot different from the old w/kg based categorisation.
1 Like
The seed score is based on the power profile. From there itâs all results based. Without the seed it would take exponentially longer to get anything resembling competitive fields.
Whether the results part will work effectively is yet to be seen. Itâs not ELO, it doesnât seem to compare you to the other riders in the race. Itâs looking at how you perform in race as a whole. Long term it might get to the same place but short term it is producing some counter-intuitive outcomes.
2 Likes
Iâve not got a particularly large dog in this fight, as Iâm mid-50s, have competed in one form or another in races (rowing, running, biking, triathlon) for nearly 4 decades and have long-since come to terms with the fact that racing is all about losing, and even the all-too-rare âgood resultsâ are extremely painfully secured. All I can do it prepare properly, execute my race plan and adapt to in-race developments in a sensible way.
My preference would be for mass-start events, where you hang on as long as you feel is wise (judging this is a key race skill in its own right) and then have a ârace-within-a-raceâ with whichever group you end up with. There are obviously no safety issues unlike IRL, though in the interests of full disclosure, I once fell off my turbo, after not mounting the bike correctly.
My preferred ranking system is ZR.appâs vELO regime, as that feels intuitively logical: to improve rating you need to produce materially improved 5 minute power or beat higher ranked riders; if you lose to lower ranked riders, your rating will likely go down. And the terrain-adjusted pre-race ratings appear (based on unscientific analysis) quite good at predicting results when riders are racing, rather than in the race for some other reason e.g. sandbagging, training, making up numbers in team event etc.
The openskill-based system Zwift has implemented produces too many counter-intuitive outcomes (results vs pre-race score; and score changes vs results) for my liking, notwithstanding that the system has stood the test of time in its use in computer gaming.
3 Likes
Computer game match making is considerably better than Zwifts fixed consistent split categories they use - I think this will also be a failing of the OpenSkill deployment unless rectified.
1 Like
Iâm not a computer gamer so I donât know how closely matched players are in an online game, but I assume they will be reasonably closely matched as there are relatively few (as the examples given in the TrueSkill / OpenSkill literature all relate to small numbers of players - happy to be corrected on this!) and there are bazillions of X-Box users out there.
Conversely, if the ZRS 0-1000 is essentially roughly equivalent to 0-3000 on vELO (3,000 being about as high as you can get with plausible numbers outside of the likes of Ganna and peak Wiggo) then a range of ZRS from 560-690 is equivalent to circa 1650 to 2050 on vELO, which covers a very large range in ability (4.7w/kg vs 5.2 for a 75kg rider) so I think the task of creating âroughly equalâ ability pens is impossible, hence my suggestion of mass starts (and my reference to racing being about losing) with vELO being for bragging rights at an individual level, and a good basis for matching ability for team events.
2 Likes
On the first ride 5. nov. 24 on Zwift Companion app, your name on the pictures says «E. Markman» and not «B. Markman», as your later rides from 6. nov. 24 does. It also looks like your avatar has changed from male to female? Donât know if that is the case, but mabye itâs something to look into?
Graham Weigh often rides in the pen A Tiny Races, despite being a Category Enforcement D racer. A quick look at the results I posted 9 days ago in the Tiny Races thread shows he lost 6 points from race 3 to race 4 without reaching 85% of his seed score, after finishing last (67th).
On 7th Decemeber, I lost 3 points for finishing 78/97 in race 3 in pen E.
Suggests racing in higher pens gives more of a points drop to me.
1 Like
On this ride she has 194w mean power for 26 minutes⊠so she has a lot of power. She would be a very very strong C.
yes, when 5min power is a lot lower than 20min power, it will happen. but once they do a race with a 5-10min hill, theyâll be bumped.
are you going to be providing us our actual seed score along with the current race score and the progress to get to that score? I am currently trying to figure out my score as looking at previous races where I have raced 350-520 pen when I was around 450 they are now showing sometimes close to 150 points higher than before v3 was recalculating results and this doesnât seem right to me as on 14/11 I finished a F365 race and score shows 554 with a down arrow, raced ZRL the following week, which supposedly doesnât count towards ZRS (but they did affect ZRS at the time) then on 21/11 my score is showing 557 with a down arrow, how can my score change like this without actually racing?
I think Zwift recalculated all the scores based on racing since 25th Sept (might be a different date). Thus up and down after any particular race on Companion App etc. means relative to what the score would have been before that race had it been calculated on the new method. Confusion only enhanced by them doing this twice in a week!
Im sure if you post on Zwift Insider article around your issues with Zwift racing score, you will get the same personal support and experience from Zwifts Director of ProductâŠ
Or perhaps you wontâŠ
2 Likes
Will Petr finish around 22nd in the 1710 GMT pen D?
Iâm sure he will finish higher than than if putting a proper effort in, I think those bad results from v1 seed score have lowered his vELO substantially.
Petr is more than capable of doing that little KOM in ~2mins30secs each time, it wonât even threaten a seed score increase.