Exactly right. It’s not technically difficult, it’s just not the direction they’re going. They’ve already got a ‘vision’ based mainly on what takes the least amount of work (there is no other logical reason why you would ever take the ZP ranking system).
It feels like a few guys who aren’t skilled in this particular area (no offence, but it’s a specialist task) doing the best they can to deliver the ‘results based categories’ that everyone has been asking for, when actually the company should be properly investing in the actual skills and experience necessary to do the job properly, designed end to end, instead of managing a barebones dev team and directing all the cash at marketing.
It feels like to get anything done at ZHQ you have to take the pace partners approach.
Just thought it was funny seeing how a typical Zwift D cat race with category enforcement ends up with some interesting power numbers and it all being shown on one of youtube’s biggest cycling channels.
I think it has mostly to do with priorities and the way they work. Changes to racing is not a priority since Zwift already have an offering and it is only used by a minority of the subscribers. It does not directly result in new subscribers either.
It is obvious when you see the release notes each month. You can see what racing improvements is competing with.
Yes, we got better TTT starts, which is only used by a small subset of the racers each week. Someone (WTRL) must have done some lobbying there.
It seems like there are no dedicated UI resources for each team, but more like each team has to compete for the resources. That is not the best way to work if you want to push new features from different teams in parallel.
E.g. even if the “competition” team and the “data science” team has resources to implement race ranking from their side, they have to compete for UI resources and therefore have to just ask for the minimum UI changes for it to even be considered implemented.
My presumption based on different posts was always, they have done the work for rankings in the background and they are leaving them sit there for a few months to build up data so on launch it was fairly accurate - And then pen assignment needs reviewing\changing to ensure this can be linked to rankings…
Im still in shock its going to take 6 months to deliver a ranking system and the work hasnt started.
Then add in, hardware limits coming before it in the priority stack and I do wonder what their view of users wants is and how this is formed. I do wonder if many of this is driven be a loud voice internally.
More pens helps if there are enough participants. I think though the same issue will exist though that shouldn’t necessarily exist. It seems Zwift really wants every race event to be open to every level of ability. Not sure if just to avoid a too cluttered events calendar or what. There are plenty of individually listed A, B, C, or D -only group rides. Not so much of this for the race events.
Maybe the new pens will emulate the number of choices in current pace partner roster
Why does everyone want close racing but mostly themselves to be in the front group? Then just do a group ride with a sprint at the end. It’s also not a bad thing to be a certain ‘category’ but you know you won’t be in front at certain types of routes. Racing will be significantly worse with everyone in front group and a sprint at end…
Personally I moved from D to now B since I started at the end of 2019. Especially in B has been quite some time to be near the front (on certain routes) but being dropped is earlier in well attended races I’d be among the same riders every week. Despite ending in group 3 there was competition. It’s not very different from now being in a front group. Races where virtually no one drops off are actually the least competitive and boring.
So, too many categories won’t necessarily be an improvement. Well attended races are most important. Getting a categorization (hopefully on results) does not need to fit the route. Better if not. I know which race I won’t be ‘competitive’ in the ZRL. However, I wouldnt want to ride at a lower division just because my 20m power is modest at best as I prefer to compete with same people on a regular basis.
Arguments given against zwiftpower ranking are sensible but I don’t think it’s a bad start if a longer ranking window is chosen (a year), the downward momentum is tackled in the formula (rankings now are significantly lower than a year ago although not necessarily an issue) and not too many pens are created. Even if a bulk will be in the 500s relative changes there still matter so if you win you move up. Moving up I see as just dividing signed up racers by rank and put them in roughly equally sized groups. Mentioned downsides are mostly smoothed over this way.
I don’t think everyone necessarily feel like they need to be in the front group. I think it is the distance up to the front group (in level) that makes riding in a group at the back feels meaningless. I often see riders who gets dropped, quit and I can’t blame them. TT’ing to the end in a non-ITT race is boring at best.
If one knew that, “even if I was dropped this time, there isn’t much I have to improve to be able to hang on with the best a bit longer”. More categories/pens will make that distance smaller. It is a psychological thing.
I’m sure there are people who gets excited about TT’ing to the end, or fighting to not be last, but I’m not one of them. It doesn’t trigger anything in me.
I would. I’m using Zwift to get the adrenaline going while I at the same time get some training. I guess we have different reasons for racing on Zwift.
If more categories are added, I think more work has to be done to consolidate or steer the event organizers to focus on having races serve a subsection of the entire range of cats instead of the full range.
As an example, looking at today’s races from 1pm and 2pm local time. There are 12 of them. Of them, 10 include down thru D cats. Signup numbers though for D currently look like:
3, 5, 3, 1, 5, 0, 4, 1, 2, 1
Can’t help but think that more Cats aren’t going to help all by itself. OTOH, if there were instead fewer (eg 8) races, but instead 2-3 races each of separate A/B, B/C, and C/D split races that each in turn has more sub-cats, you’d consolidate the entrants to fewer races with more participants in more tight races.
The ZwiftPower ranking system is fine as long as it is massively changed?
They are choosing it so they don’t have to do additional work. If they have to do additional work, then they may as well implement an effective system.
A longer window is not a massive change is it? Nor is removing the built in downward trend.
Having said that: A major problem is that Zwiftpower is not integrated (still…) so any non standard (e.g. primes) race degrades the quality of the rank/data. Having experienced Zwift my hopes of something proper are low to non existent.
one obvious issue is that it is very easy to find races that are low in rank quality and for those who don’t race for rank (me, although i do have a good one since i did the zwiftinsider races as an A last week), their rank will never reflect their actual ability if they are not actively looking to improve it
Yes but that is not a problem if they remain in low ranked races. Occasionally there will be people with a high rank and if they get beaten this person will also jump up in rank. Someone can always consciously stay low on whatever metric but that will just be a few people.
Congrats on being second in As by the way. I could only witness it from the middle of the pack unfortunately! But those races also negatively impact current ranking but there is no holy grail and we should stop hoping for them. Most things on results measures over a longer window, also current ZP rank, is better than currently. As it solves the main complaint currently: Person A always beats me but stays in category whatever while I am bumped up.
I support a Race Ranking system and ZwiftPower is the only one I have spent any time trying to understand. I am sure there are better systems. I believe any ranking system based on results must be better than the existing one based on wkg. Unless there is a sudden change in Zwift’s Race Scoring plans I believe we will end up with the ZP/USAC system.
What I am keen to attempt, through this thread, is to outline where I believe The ZP ranking system does work whilst also learning more about where it doesn’t and what modifications might help improve it.
It is easy because you are able to race in a power based category and are not restricted to race only against those within a similar Race Score range. On a dedicated ZP ranking system you will almost certainly be forced to race against those with a similar race ability.
Can any system identify your actual race ability if you are feeding it false information?
Clearly if you want to retain a poor ranking keep racing poorly - others have said previously they have no problem if racers just want to finish in the back of the pack time after time.
However a Race Scoring system, based on results, should promote you if you win or place highly.
I say should because under the ZP system you need to be racing in a race where the Race Quality is within 10% of your own Race Score to be sure to gain an improved Race Score result. (It is often written that you need to beat better ranked riders to improve your ranking. IMO that is not strictly correct)
If I may use your own Zwiftinsider tiny race experiences as examples of how the ZP ranking does and also doesn’t work.
In your first 16 B Cat tiny races you were never out of the top three podium places but in only one of those races did you gain a very small ranking improvement. Here ZP didn’t work because your own ranking was so much better than those you were racing against.
In your first 3 A Cat tiny races you came 2nd, 5th and 6th and all three of these races gave you a ranking improvement. IMO exactly how it should work as you were racing in a much more appropriate race for your racing ability and not based on a wkg category.
IMO the ZP ranking issues we see now are mostly due to a Pen category system not putting together racers of similar racing ability. This would not be the case when pens are based on Race Score.
