Normal: Bs are permitted to enter my C race to spoil the dynamic of the race and outcome.
Autocat: I am entered into a B race.
Normal: Bs are permitted to enter my C race to spoil the dynamic of the race and outcome.
Autocat: I am entered into a B race.
Category enforcement? The “Category enforcement” I’ve seen in ALL of the Zwift Classic races has been like this:
You’re a lightweight low B? OK, you belong in C3 with some 88kg A riders who can slam down 350 watts for 20 minutes on these flattish courses.
The one chance we had was the Muir and the Mountain stage but that got changed to YAAC (Yet Another Almost Crit). I guess Baloney is our only chance to redeem ourselves…
Mark, you’re obviously extremely unhappy with these categories - do you like the A B C D cats better? I personally don’t think it’s a lot better, also don’t think it’s a lot worse. The performance-based categorization is where it needs to go, problem that I THINK they’re trying to use ‘Autocat’ for is to categorize people who haven’t raced enough to have a stable ranking.
Currently they struggle with people with hundreds of races.
I do hope they can fix racing.
Racing is fine - I have fun with it, and did when I was getting dropped every race too. Some people are claiming it doesn’t work because they aren’t at the front - not everyone can be at the front. Any time there’s a climb in a Zwift race, people attack. Unless everyone has identical power curves, some people are going to get dropped.
I’m talking about NOT USING POWER TO WEIGHT RATIO TO CATEGORIZE - except for people without sufficient or recent enough race history.
They need to make it less complicated to get to good races, and enforce whatever categories are being used automatically - so the user doesn’t even see it, preferrably.
Its impossible to Categorise anyone unless you have sufficient data . If Zwift are trying to do this (which I think they arent by the way ) , they need to stop now as that truly is chasing a mirage.
I think your view that the new metric based categorising is for most insignificant and underwhelming in its impact is widely shared . I think even Marks feedback might even sit in that camp , i.e the expectation was this was a revolutionary thing when in fact its not , its just tinkering with some superficial entities.
This fact I think is why Zwift/WTRL can claim 1000’s are happy with it , Its the glass half full , glass half empty sort of thing . If you tick the middle option in a survey , or dont fill out the survey at all , are you happy or unhappy … How you ask the questions and interpret the answers is more important than the actual answers (or lack of ) themselves.
I hope development team have been given a more precise set of requirements to satisfy than fix racing. I must admit that from where I sit I am not so sure that is not in fact what they have been given. Maybe too pessimistic , but the evidence points pretty strongly to this being the case I am afraid .
Unless we have another route change in favour of a flatter route last minute.
I am looking forward to the Bologna route with that climb x2.
So, my take on the Innsbruckring the other day then. Just two thing before I start.
First off, let me just make one thing clear. While keeping riders anonymous as always I will still refer to individuals as examples below. This should not be misconstrued as that I have something against them or that I think they are doing something wrong. I DON’T! Very important. In the average Zwift race sandbaggers, cruisers and heavies dominate, but they are doing so because Zwift lets them. They didn’t have a say in designing the race rules and cat system. Also, it is so very easy to mix up real life racing with Zwift. Zwift is not a sport and it is nothing like real life racing. For example, Zwift makes it possible in the average race to cruise and Zwift also rewards cruising. What are the implications? That you should cruise! Cruising is the optimal way to race in Zwift, so of course you should cruise. This I have stated elsewhere many times before. So no, you are not doing anything wrong when cruising. Or sandbagging non-WTRL races. And you can’t be blamed for your natural weight. You could, however, argue that Zwift is at fault in some ways. And I do.
Now, let’s start. I have looked at results from the last round and I am still seeing what I predicted and what I saw the round before that. I will pick one race in particular. It is not unique in any way, it is just representative and a very clear example.
I do think that the races are better than the average Zwift race. But they still suffer from the same old problems and I also think AutoKitten is not doing its job very well.
So we have this one pen in one timezone. It has few participants compared to many other starts Monday, which is a good thing because certain tendencies become very visible. The race is completely dominated by two individuals. There is them, and then there is the rest of the race, so to speak. They race on their own. One of them is very heavy but a hard worker. The other one is a cruiser (it doesn’t matter whether you cruise intentionally or not - if you cruise, you cruise).
The heavy guy can push 30% higher Watts or more than anyone else in the race. Obviously he is considerably faster on the flat at the same effort level than anyone else. What is the upper bounds of his category is a little fuzzy, since AutoKitten does take other things (big sekrut!) into account than just 20 min W/kg, but it is pretty safe to assume he is on top of his cat with regards to how he was categorized. This means he is also not disadvantaged in the Legsnapper. He can stay at the presumed cat limit in spite of his weight. He is not advantaged anymore going uphill like down on the flat, but he is not disadvantaged. And he would not finish last if he were to race in the next cat. Is it a fair categorization or not? Tricky question.
But that doesn’t really matter, because eventually he doesn’t stand a chance against the cruiser in the race, who sprints the crap out of him. The cruiser, considerably lighter, has a HR diagram for this race that is basically a mix of zone 2 and zone 3, whereas the heavy guy stays in mainly orange. So while he is inarguably a good sprinter, he also has a lot of HR wiggle room when racing, for the most time underutilized potential that he can tap into whenever he chooses. The heavy guy (or anyone else in the race) doesn’t have that choice because AutoKitten didn’t let him.
The cruiser has podiumed in every single event this summer, and we’re talking 30+ races. And he is doing so from said HR zones. Could he work harder? Of course he could. But that would, over time (like after 2-3 races), have placed him in a higher cat. Should he work harder? Like… morally speaking? No, not necessarily. He can race at any effort level he chooses. Particularly so in WTRL since he had no say in his categorization at all. The question is rather whether Zwift should have put him in that category. Now, there is no previous data Zwift could have used to categorize him differently since he was previously podiuming in every single race from a zone 2-3. His W/kg and other power measures are roughly the same in his races.
So what is missing? Well, as I see it AutoKitten should factor in weight and average effort level. How to weight them is a different matter. But what does that mean? Should everyone above a certain weight be placed a cat higher? No, of course not. But weight has a value, much the same as W/kg in winning races. The value is smaller than W/kg for sure but it is still significant, and if you don’t use it in the algorithm, then heavier riders will on average do much better in races in lower categories than the lighter ones. Why should they be advantaged? The ideal would be if weight was completely randomly distributed over results lists. It’s hard to get there but one should at least try. And the argument for effort level is very much the same.
Or… you could take a completely different route and strike W/kg, weight and effort out of the picture. With results-based categorization you don’t have to worry about any of these measures in making categorization fair and fun.
At the moment I am just getting fit again and pushing 3.2-3.3 w/kg. In the past when I have been in this position and raced a C race if I had a great race and really pushed myself I could get disqualified or pushed up to B. This I am mostly fine with I should be racing in the right category. The problem is the breadth of the bands. Racing in B with a 3.3 w/kg I will get murdered, not only that I will probably get dropped the first lap as I can’t push the w/kg like the riders at the top of the category. So the race really is one lap in Zone 4 breathing out of my bottom, get dropped, spend the rest of the lap mostly riding on my own private training ride.
Honestly, I don’t expect to win, but I should be able to sit in and stick with the bunch. Getting dropped so quickly is demotivating and also I believe this does not help my development as a rider, mentally it is easier to push yourself when surrounded by a bunch and racing, rather than just doing an elongated FTP test when spat out the back.
With this WTRL series, all of the races I have entered other than the last felt competitive. I could hang with the bunch do a turn on the front, it was hard but not impossible. Ok, I still ended up getting dropped because my fitness is still not there but when analyzing the reason for getting dropped it is different. I would make a tactical mistake, not push that extra 20% to grab a wheel, not quite have the burst to keep with the group on the hill.
But every race I went into the next one feeling like, right let’s just push harder on those bumps and keep with the group. Remember to hold the wheel, keep in the front of the group, don’t get dropped. I honestly felt that I am making progress. And the training is better because I hold the group I push myself harder than I would spat out the back, I work harder and get stronger.
With the old categorization, it was really just a case of ok I am 3.3 w/kg, let’s just keep training until I get up to 3.8 then I can race again.
That is really what I mean by feels better, with autocat I don’t feel that entering a race is a waste of time or that I might as well just do a training session. I feel like I am competitive, that is there is a remote chance. In last week’s race, I was on the last lap sitting in 10th place, I feel completely dead, nothing in my legs but I am thinking if I hold on here just a bit longer I can get a place. In the end, I did not have the legs for the hill and got spat out the back, but instead of leaving feeling completely demotivated I get off the bike and think to myself, let’s keep this training going, if I push hard, next week or the week after I might be able to get that place.
Honestly, I think if I was pushing 3.8 w/kg racing in B would be fine, but at 3.3 w/kg I can see why many folks feel like it is so demotivating. I do believe this problem exists in C and D also.
Just remember as you get stronger you will move up to C2 and that will mean you will be at the bottom of the cat again.
100% got to keep progressively getting stronger and faster, but I honestly feel that with autocat it is not a case of move up and immediately get dropped until you can push the top w/kg in the category.
Is there a thread somewhere that explains why there is not just a point-based category system like physical events? I hope I did not open a can of worms with that question.
Currently we don’t know how auto cat works, there was mention of w/kg, 20min power 5 min power and race time. They test different things each week. More info here
This is an interesting one to read.
Well that really is the tl;dr version of several long threads now. I don’t disagree with any of the bits of your post that I didn’t quote either
Results-based ranking is both necessary and (almost) sufficient for fair racing. This statement is amenable to mathematical proof, under some fairly simple and reasonable assumptions (eg what is meant by “fair”).
What’s particularly disappointing about the situation is that others have been pointing this out for literally years, long before I’d even started zwifting. And I know enough people working at tech companies to know that they are not all stupid, so they have also known this for years. The reason they haven’t fixed it is that they don’t want to fix it.
Autocat is a distraction based on a false premise, that “better” categorisation based on rider biophysical metrics will fix the problems. It won’t.
Results-based ranking is both necessary and (almost) sufficient for fair racing on Zwift.
(The reason I add the “almost” in the above is that it would be possible to implement a results-based system such that it was still unfair, if you really wanted. But it doesn’t take much extra effort to avoid this.)
I fear you are right.
Whenever they mention that racing is on their agenda and that they intend to look into it Zoon (these moves are always interspersed with periods of complete silence or a “racing will remain as is for now”), even though I know it is a sucker punch I still get suckered in by these moves. Every time. Then I remember that it is just part of a bigger pattern that keeps repeating itself indefinitely, like some sickly green floral wallpaper of misery. And if that is so (and it is, you come to understand that having been zwifting for a while), then there has to be an underlying reason. And, as you say, it is not very likely that this reason is that they just plain suck and completely lack the skill set to extract thumbs out of orifices. No, it has to be something else. And then you suddenly remember that this is, after all, a company where even the CEO and founder has been caught sandbagging on occasion. Then it dawns on you (again):
W O R K I N G A S I N T E N D E D .
Which translates to:
T H E Y S I M P L Y D O N ’ T W A N T T O .
And then you bury your childish dreams of fun and fair racing, a little deeper every time. But you keep pestering them anyway. Because you feel cheated on. And you are cheated on. In every race and beyond.
It breaks my heart to see people getting cheated. A weak spot of mine. People getting mistreated, I just can’t take it. They never lost any sleep over it. They’re a different breed.
I’m with you on this. I’ve had some races where I haven’t had a chance and some where classification has been ok. I know some people’s experiences have been better and others worse than mine.
I’m also surprised there’s no ability to grade the races and comment on them, like you get with Futureworks activities. If you leave it for people to only feed back if they want to, you’re only likely to hear from an unrepresentative minority.
Good point, why isn’t the autocat work sitting in the Futureworks department?
A coupla things:
I’ve started a thread for Stage 7: Twice up The Baloney
And all this yapping about The Autocat and how the new cats are going to do this or that is just that: Yappin’. I’ve done one series of racing with race entry cat enforcement on Zwift and it was amazing. At the time I was at the pointy end of C and for the first time, I managed to hang with the lead group for the entire event as we were all C cat racers. It was obvious there were no A’s or B’s in the event because the most were putting down powerwise was within the C cat limits.
Instead of fooling (NOT the word I was thinking of) with Autocat, let’s give the current category system a chance to work with entry cat limits. Thus far my experience with Autocat has been just like having a bunch of sandbaggers in the race because it’s so fouled (NOT the word this time either) up.