Of course a lot of regular winners in cat X could do fine in the cat above. The problem is the mid-pack finishers who get upgraded due to chasing on etc, who then get shelled in the first 2 mins of the harder race.
Top end of C do well in B because the top end of C rarely have to do full gas efforts for 20 minutes. For some of the them riding in the front group of C cat is a zone 3 effort. Meanwhile mid-pack riders (and especially lighter riders which is why they are the ones most likely to be upgraded while never finishing anywhere near the podium in C cat) are doing FTP tests every race just to stay with the front group.
The very nature of Zwift’s w/kg limits rewards sandbagging either deliberately or accidentally and the sooner the whole system is scrapped the better.
5 Likes
Given the strange sea levels surrounding Makuri, I’m okay with some people having some sandbagging skills.
1 Like
Calling me a liar is a bit harsh and if you’re going to resort to that I’ll decline to respond to your points.
Cheers
3 Likes
Well with your reputation that statement is always going to go down well with everyone.
But you are absolutely right.
Your category should not be defined by power but by results.
The problem is, it takes a bunch of races before ‘results based’ methods can get you into the correct category, so doing some power based seeding is necessary unless you want to see D races dominated by A racers who have only done one or two races in the last 60 days.
Who says it needs to be last 60 days ?
Don’t race - don’t drop ranking.
A great ‘results based’ ranking should take account of results quickly.
I support https://www.zwiftracing.app/
For anyone thinking they are racing out of category check it out.
3 Likes
60 days was just an example. But, as you probably know, zwiftracingapp does use power based metrics to initially categorize riders, otherwise it would be a mess. This is what I am talking about specifically. They use a modified compound score.
Yes they do use power based metrics (CS) but I’m not sure I actually support that and have commented, on their site, as such.
I would like to believe that a results based system should be good enough to stand by it results based ranking,
But I will agree that CS is not a bad starting point if there are no other results based criteria to use.
[ edit: I don’t agree that it would be a mess without a power based metric at the start - just that more work would be needed to ensure initial results prduced a quicker results based categorisation ]
I’ve followed it pretty carefully. The issue is, so many people in Zwift do not do a lot of racing in events that would actually qualify as a race (i.e. ToW does not count, ToM does not), and you do need some form of ability for results to ‘age out’ - even if 60 days is not the right number. On top of that, if you’re racing with people all about your exact same ranking it’s hard for your ranking to change in a huge way, and if ranking was the default there are also ways for people to tank their rating. So, if you don’t want D races inundated with A racers who haven’t raced much you will want to have some form of power based ‘seeding’ as there is no easy way to get an initial ranking without a decent amount of races.
1 Like
I don’t wish to be confrontational - it’s very late over here. But
So have I.
No you don’t. If you want results to ‘age out’ race and prove you are getting worse.
If you keep beating them your ranking will keep getting better. I don’t agree with their current philosophy that if you are ‘expected’ to win you don’t gain anything. IMO if you keep winning you keep gaining ranking points on those you beat.
Build the best results based system you can for honest ‘want to get better’ racers - don’t build a system to try and deal with those wanting to tank their ranking - it won’t best serve the majority.
The issue is it’s very easy for just a few riders to completely wreck a lower cat ride, so if Zwift knows you can do 4w/kg for 20 mins you honestly should not be able to blow up a D cat race no matter what system is in place.
Edit: My first sentence sounded confrontational so I edited it out - that was not my intention. Main thing is, I do hope that Zwift does something very similar to what they are doing at zwiftracingapp where they use a modified compound score to initially seed racers - this does a great job of initial seeding with very little data needed. This way you won’t get massive imbalances in the lower ranked races.
3 Likes
I second this. Tennis rankings use 1 year.
1 Like
The “fluctuation” allowance used to be a lot larger than it was this last season. I don’t know why. It might have been observed hacking of the system, complaints from racers, complaints from sponsors or orders from Zwift’s Racing Coordinator.
I agree that less than the amount is too small but with ZwiftPower dropping the old system, things are going to change next season.