Not on zr.app, no, because it has already been looked at and addressed within the algorithm.
The widespread retort in this thread is because we can extrapolate to this being used for pen assignment, and what a horrible experience that will be for racers, new and old. It will be a mess. There isn’t room to put it out and then improve it, the damage would be too great and the change would take too long. It needs highlighting now and addressing now (as I think it is being).
Maybe what would have been better is a post a few months ago saying that this is what they were planning, then we could have had this debate, and at the end of June would be in a far better place at the point that actual changes are being made to the UI.
Take another run at what I wrote. I’m hypothesizing that Behavior and fitness trended towards clusters at w/kg limits for A, B, C, D over the last few years. Based on your response it sounds like that is correct and as a result ZRApp has created more categories that segment the clusters?
If you were to have ZRApp fully integrated to zwift would that mean we have more categories to enter or would you do it in another way?
Yes that’s true, we know that’s the fundamental issue of ZP categories. Racing at the bottom of a cat is rubbish. You either quit racing or sandbag the category below.
Best: Matchmaking
Decent: The organiser would set the rating thresholds for each pen
Borderline acceptable: More categories ala the zr.app ranks, with organisers deciding which ranks go in which pens.
This is why I asked people to write out the intent of the system above rather than the feature.
These are all good intentions. The inconvenient truth is that there are more categories on ZRApp than there are racers in several of the race events this morning. This would however make popular races more accessible.
I’d also add that there shouldn’t be a cat/pen selection on an race. You just join and get placed.
I do think regardless of how “rider score”, or “race ranking” is determined it would still be good for Zwift to enable flexible cat boundaries, so there isn’t always a specific “score” OR w/kg target to hit to always be a “Top A/B/C/D” racer in every race you do for the full season.
If there were matchmaking, or flexible cat boundaries which don’t give sandbaggers one specific weight, watt, ranking, or points number to float under then sandbagging to stay at the bottom of a specific thing becomes much more difficult, and your race experience will become much more varied throughout the season (sometimes hammer, sometimes nail).
Yes that already exists, it’s called category enforcement and it limits you to one pen (or above). It already supports a range of inputs, including Compound Score.
There is no need for ranks to equal pens. That’s not what ranks are for. They are not categories.
That functionality has recently been made available. we can sort pens by a number of metrics now from CS, W/kg and a few more things.
likes of tfc mad Monday use split cats with CE with custom ranges e.g upper d runs from about 2.2 to 2.7 so all those 2.5 Sandbaggers have a harder time.
Yeah, it might be that Zwift have to start varying them in their core race series or events for it to catch on. It would have to become more the ‘norm’ rather than the exception to reduce sandbagging overall - otherwise sandbaggers can just ignore the one or two events that have a flexible category I suppose. I’ve ridden the VirtuSlo events using multiple cat boundary definitions, and it really mixes things up.
Ah. The site is a bit confusing on that. Click on the “Categories” page and the tab name is “Ranks”. The terms are used interchangeably across the site.
My point was not that this kind of selection should put the same rider types always together but rather match them up across the metrics including the actual character of the race.
Eg.
I’m a punchy C rider at the sharp end of the category. I’m doing regularly a 45k race on the short Innsbruck course with the Leg Snapper as the main attraction. My fellow C riders who have better steady power profile can’t really compete for the win there. I can and do so regularly.
A new match up system could throw in some B with less punchier power profiles in there, which would open up how the race could be won.
Nonetheless I think most races with a competitive group will end up in a sprint most of the time. Besides mountain stage like races…
I think closing the pens a minute before the start and automatically shuffling the riders into even sized (or even distribution or…) pens based on whatever criteria the organizers chose would be an interesting solution to what you pose. Sometimes it’s similar boundaries as now and you’re at the pointy end. Sometimes it’s a mix of low Bs high Cs and youre fighting to stay in the pack or chase down a break made by the Bs. Sometimes it’s very high Cs to mid B’s and you’re just fighting to hang on. You don’t know in advance what the cluster you’re going to be put into is, so you just go out and race with what you’ve got.
That depends on your very idiosyncratic interpretation of “rating system”.
Even the fantastic multiplayer Elo is only an approximation to the true ranking of riders, assuming such a thing even exists.
I don’t think you have yet explained why the Elo system - augmented by various ad-hoc fixes that have been found necessary, that’s a shock - is necessarily going to do a better job than a points system.
Indeed the need to (a) add some sort of bonus (points) for the top finishers, and (b) ignore poor results which would otherwise downgrade players, is very obviously introducing substantive elements of a points system, which given your previous claims that Elo had solved all the problems of player ranking and was a plug-and-play solution, is amusing.
Indeed you might well say that the rating system is evolving into a points system.
He and many others have already, in this thread and elsewhere. Feel free to engage once you’re up to speed on what has been said, if you’re going to make comments on what has/hasn’t been said.
