yeah but race organisers have all been told by now, either by me or through the grapevine. the TL;DR is if you can scale a power meter and smash out a huge 3 min you can tank your estimated cp, or zFTP, or as i like to call it, “the cardio zone” and drop a cat.
you can also probably just set your weight to 150kg overnight too and race whatever cat you wanted the next day. enforcing ZP and CE at the same time in an event (mostly) prevents both of these
I’m looking forward to this coming in as even with Zwiftpower and Intervals.ICU I’m struggling to spot what has bumped me into C in Cat enforcement. Knowing my luck this change won’t have been implemented before the 60 day window times out on the ride that’s bumping me.
i think it’s 8 minutes at the bare minimum (i suspect longer) but we may as well wait for the figures to come out and you can work it out pretty quickly from there. i’ll DM you my figures on day 1 if you want
Compound Score isn’t about predicting the result - or even about grouping riders by the closest possible metric. The whole point is that it is a universal way to assign a number to a rider’s physiological capability, a normalized one. It allows you to group by power without also having to hack something together to cover for weight. As the paper shows it has a very high correlation with rider success, but that is a broad term. Of course success on AdZ is different to sprint success.
Its a great, simple to use starting point. Adjusted based on results it’s even better. Then it is just about choices for the race organiser.
If I am organising a race that finishes with a short climb, I may want to sort riders based on their 3m w/Kg. Or I may not because having riders closely matched might not really be the goal. It doesn’t really matter, as long as I’ve got the choice and race entrants know what logic has been used.
Closely matching riders is not the key to a race feeling fair.
BTW I like that short term power and longer term power are used in a 3-factor model. It’s the w/Kg part that is problematic.
I guess we can understand a technical reason, but if not the whole rundown, even just pushing a bare minimum 1 Line bit of info to ZP profiles would be problematic, such as
Category Enforcement Min = A/B/C/D ?
Yes. In the very early days of CE Zwift was definitely talking about FTP and VO2max.
Shortly afterwards CP was introduced, FTP sort of dropped, and it took a while for the chart on CE FAQ to be updated.
I think clearly some confusion between FTP and CP in early days.
The use of Compound Score is not as a race results predictor but rather a way of assigning riders that takes into account raw W and W/kg and provides an initial “level” playing field?
But people prefer to be winners! 
Naively, I wonder if this could simply be addressed by giving race organizers the ability to customize Pens based on a subset of metrics, so that even a low to mid-C like me can have a shot of being at the pointy end of a race for a change?
Thank you for your reply David and clarifying point #3 and #4 of my post.
However one follow-up on point #1:
Sharing anonymized data over a closed API is totally different than a 3rd party website publishing GDPR-protected data (which a ZwiftID + power values is not). What we need as Zwift event organizers is no different than the data sharing which you are now already sharing to for example WTRL or Zwiftpower. I don’t think there’s anyone in doubt whether it’s allowed for Zwift to share data, let alone anonymized(!) data, in a non-public way.
Yes, but such races are few and far between. What % of Zwift races have summit finishes?
Until you sort out the totally unrealistic descending speeds (with corner braking?) lightweight riders will almost always be at a disadvantage. Having to descend at/above FTP with no respite shouldn’t be a prerequisite for staying in contact with heavy riders (and especially packs of heavy riders).
There are a few possibilities. Rank based, Power profile based (the current CE system), Hybrid, etc. There could be a future where the organiser chooses between the category system he/she wants.
Actually that future is already here. I have scripts ready to launch fully customized categories in Zwift racing (using whatever parameters a race organizer wants; e.g. 10 minute effort; or only look to last 30 days, or check the Zwiftpower rank, or use the win percentage last 90 days, etc. etc.), including “pen guidance”. And it takes only a very small change to the Zwift website (one query parameter) to go from “pen guidance” to “pen enforcement” for fully customized categories.
I will send you a personal message with a full explanation.
[quote=“Anna Ronkainen [AEO], post:55, topic:596058, username:Anna_Ronkainen”]
The easiest way forward would be to rephrase the ZP consent so that it covers racing result and integrity platforms regardless of implementation (of course you’ll probably have to bother existing users once more but that’s hardly a big deal). If the goal is to move away from ZP, you’ll have to do something like that at some point anyway. (Personally I would like to see that kind of consent to take the form as something of a “Zwift racing licence”, meaning that it would simply be a prerequisite for signing up to any (public?) event categorized as a race.)
[/quote]
Or allow organizers to only allow entries for riders who have opted into such an agreement. Possibly also allow different levels of opting in i.e just information usage or also requiring something like an e-race passport with weigh in videos etc (eg I think on RGT Echelon has something like that)
Honestly I think on the legsnapper pure watts will still win out. So would be the same result.
But as James said it’s not meant to predict who will win but demonstrate that going on just WKG is flawed and why watt floors and ceiling are then needed. Compound Score is much better at determining ability and initially splitting boundaries. If this is then accompanied with results based ranking up and down.
CE seemed promising initially and I saw many riders I knew where sandbagging moved up a cat - then they all screamed blue murder and the boundaries where moved and they were all let in again and to add insult to injury could justify sandbagging as Zwift had given them the green light to do so.
For those interested in reading the study, here a link COMPOUND SCORE
I’m a mid to low C and the split-cat events, like TFC Mad Mondays, can (sometimes) make you feel more involved and competitive.
Watch this space, as we are shortly to launch a race organiser toolkit with fully customisable pens and even matchmaking after signup.
This is why I think Zwift could benefit from an open API, have the community build the tools and then implement those that are the best, it sound like a win win for the users and Zwift.
Great to hear! We are planning to bring TFC Mad Monday a step further with the fully customized categories, with different category boundaries and with checks on number of races in last 90 days, Zwiftpower rank etc.
So especially those riders who cannot win in all the current Zwift races, get their chances improved in our new category system.
I’ll DM you.
Thanks Gerard. Our club is being kept up-to-date on these exciting changes by our event coordinator Jamie Scharff 
I haven’t particpated in MM for a while. When I first raced in MM low-C I felt it was the first time I had actually enjoyed racing on Zwift. I never podiumed, but I usually felt involved and in-the-mix. Over time, however, I felt that the demographics of the low-C field was slowly but inexorably changing to heavy riders (looking at the recent results on ZP confirms this, with typical weights in the 80-90 kg region). On most courses, I found I simply couldn’t compete anymore. This isn’t any kind of criticism of the series, more a reflection of my limitations 
Sounds good, just a pity I normally do a long hardish ride on Sunday so my legs are in need of recuperation on Monday. Are the races private via your web site as opposed to entering on the Zwift web site or Companion ?
