Zwift Chat

Page may contain affiliate links. Please see terms for details.

Whorty

Gets free watts from the Atom ;)
Location
Wiltshire
I think it was lifting 33kg batteries out of a frame. Awkward leaning forward whilst trying to pull up on battery. I thought it was a bladder infection initially, but test confirmed no proteins or white cells 😳

Bladder infections are nasty. Not going into details, but my wife suffered a lot due to her illness and spent far too much time in hospital, so glad for you it's not that serious :hugs:
 

alex_cycles

Veteran
Location
Oxfordshire
He has 238 races listed, a D cat with a zFTP of 141w but only one in the last 90 days.

Yeah only the last 90 days counts for anything. But if he's still at 2 after that result finalises, something is wrong.
 

Whorty

Gets free watts from the Atom ;)
Location
Wiltshire
Had a good ride today on the 16:10 community Titan Grove course. 100 starters in the 180 to 350 pen. Standard frenetic start. I was hanging on to the lead group of about 60 riders for the first minute, then remembering we were heading to Titan climb I let them go and got picked up by a few stragglers. Fought my way up the climb and ended up in a group of about 8 riders - we stuck together to the KoM where I was shelled. I found a few riders of my pace at the top and we rolled through sprint after the Italian village together, but the group broke up on the climb to the esses. I couldn't hold wheels and ended up getting dropped, having dropped 2 others from the group. I rolled in alone.

Decent power numbers for me, similar to the Cyclopathy ride last week.

Note to self though, check the course before a race. I knew we were going up the Titan climb, I didn't realise we were heading up the little KoM - my legs were not expecting that :wacko:
 
Last edited:

Whorty

Gets free watts from the Atom ;)
Location
Wiltshire
Was that Downtown Titans? I did that one last week in a Ladder race and got spat on Hilly KoM which is one of my favourites when fresh.
I wasn't fresh - I was dead. :laugh:

That's the one haha - i convinced the group I was in to hold off a small chasing group through the descent from Titan and to chase 2 others up the road. We worked pretty well together through the tunnels. I was thinking, the climb up past the start/finish banner on the sea front might be hard but I should be able to draft, then we turned left up the KoM :wacko: oops, I really should have left a bit in the tank and not pulled so much on the front :laugh:
Not my best time up there for sure, but at least I managed to hold a couple of wheels so I didn't descend alone and could work with others so that was a bonus.
Quite a nice race course if fit, hills nice and punchy and enough rest on the flatter sections.
 

alex_cycles

Veteran
Location
Oxfordshire
That's the one haha - i convinced the group I was in to hold off a small chasing group through the descent from Titan and to chase 2 others up the road. We worked pretty well together through the tunnels. I was thinking, the climb up past the start/finish banner on the sea front might be hard but I should be able to draft, then we turned left up the KoM :wacko: oops, I really should have left a bit in the tank and not pulled so much on the front :laugh:
Not my best time up there for sure, but at least I managed to hold a couple of wheels so I didn't descend alone and could work with others so that was a bonus.
Quite a nice race course if fit, hills nice and punchy and enough rest on the flatter sections.

We did it once in ZRL against Hivey and crew. I think we both got some FTS points.


View: https://youtu.be/NXMu5zb-LfQ?si=5sV3GDP7z5iJNENc&t=562
 

bridgy

Legendary Member
Location
Cheddar
Exactly, but there are a number of inconsistencies with the write up that it's not totally clear (https://zwiftinsider.com/zrs-faq-deep-dive/) - they talk of the seed and not allowing anyone to go below 15% of the seed but they also talk of tweaking the seed in future and possibly allowing the floor to drop lower.
They talk about future possibility of adjusting all sorts of things later if needed, including the way they calculate the seed score - a very sensible approach and not what I would call an "inconsistency"?

I really don't know what you mean about "allowing the floor to drop lower" though - can you show me the quote you're referring to?

To me, a seed is a starting point. It doesn't change. Simple. But they talk like the 'seed' is actually a moving beast based on possibly 90 days, possible up to a year depending on the data.
It's definitely based on 90 days, that's been made clear previously and confirmed in earlier articles (and nothing in this article contradicts that). So you don't need to say "possibly".

The year thing is something they're talking about applying to those who have no 90 day data - a good idea to help mitigate a lot of the problems you've been complaining about with riders returning with no recent data and therefore being seeded too low.

So it's not really a 'seed' it's the current baseline prediction based on 30 second, 10 minute and weight of the rider.
Don't get caught up on the word "seed" and your personal definition of that word. It's just what they have chosen to call the 90 day baseline

So their model is continuosly tweaking our seed?
Your seed score is based on your best 30s and 10min power in the last 90 days. If you set new best power levels, it goes up as those are now your best numbers in the past 90 days. If your current best power levels become over 90 days old, it adjusts downwards based on whatever is now your best power numbers within the last 90 days. So it's not necessarily (or likely) changing "continuously", but it can and will move. If I do my new best 30s and 10min power in a race tomorrow, and don't improve on them over the next 90 days, my seed score won't change for 90 days until those drop off.

Surely all that's needed once a rider starts to ride with their seed is results based movements?
The previous categories were based on 90 days too. Think about it - if someone starts on a really low score based on that one snapshot in time they choose to use, then do A+ numbers in their first race, it could take ages for them to progress up through the categories - with people complaining about them every race no doubt - when you can easily just put them closer to where they should be straight away based on their new power numbers.

It can also help to mitigate against sandbaggers. Without a 90 day baseline it would cause chaos with people over and under seeded all over the place.

Again, it mitigates exactly the kind of issues you've been complaining about.

And this is where there is still a lot of noise - Zwift are not clear what is going on, Eric (to me) didn't clear anything up, and riders are confused and frustrated.
It can be complicated I'll agree with that, and there's details Zwift are choosing not to tell us - but a lot of your confusion seems to be misplaced or based on not reading the articles properly to be honest. What is it specifically you're not clear on? (There's definitely things I'm not clear on just like everyone, I'm just not sure, other than the 90 day thing which is hoepfully cleared up, what specifically you're confused about)

Maybe they should stop calling it a seed and call it something else, but if it's moving daily based on power numbers then it's not a seed.
It really doesn't matter what word they call it, just what it means. And it means a baseline based on the last 90 days

It's not moving daily unless something really weird's going on like you're increasing your best 30s or 10min power numbers every single day.

Anyway, race time, let's see how far my ZRS can drop today :biggrin:
Sounds fun!:rolleyes:
 

Whorty

Gets free watts from the Atom ;)
Location
Wiltshire
For those who have been on Indievelo, is there any way to see all races in a day? Or am I seeing everything? That's a big gap between 13:40 and 15:00 then races spread quite sparsely from then on.

1728989114302.png
 

Whorty

Gets free watts from the Atom ;)
Location
Wiltshire
..snip
Don't get caught up on the word "seed" and your personal definition of that word. It's just what they have chosen to call the 90 day baseline

.... snip
As an ex statistician by education (albeit some years now), and having built a career around data modelling, data warehouses, data mining (lots of data stuff lol) and more recently AI and machine learning etc, I read the words that Zwift have given us and yes, I do question it. It's Zwift using the term 'seed' not me - a 'seed' is something that starts a process/model and doesn't change afterwards. I have a lack of confidence when people use terms that are incorrect. Seed isn't my personal definition.

I guess my personal problem is, I can see what they've done, technically I can understand it and 'see' their models, and it's far too complex for what they are trying to achieve. If my team at work spent this long to build something that is showing so many faults (yes,, I get that's my personal view) then they'd be getting a talking to.

I'm sure it will settle down at some stage, the noise will stop and people will resign themselves to where they've landed.
 

bridgy

Legendary Member
Location
Cheddar
As an ex statistician by education (albeit some years now), and having built a career around data modelling, data warehouses, data mining (lots of data stuff lol) and more recently AI and machine learning etc, I read the words that Zwift have given us and yes, I do question it. It's Zwift using the term 'seed' not me - a 'seed' is something that starts a process/model and doesn't change afterwards. I have a lack of confidence when people use terms that are incorrect. Seed isn't my personal definition.
Have you never heard of seeding in sports like tennis, (or virtual bike racing)? It's ranking players based on their performance in a previous set time period (eg the last season in tennis, or in ZRS in Zwift, the past 90 days). https://en.wikipedia.org/wiki/Seeding_(sports)

The word "seed" has multiple different definitions depending on context and Zwift can use it however they like. You may as well argue Zwift's use doesn't describe a plant embryo (this is such a pointless issue to get hung up on, honestly!)

I guess my personal problem is, I can see what they've done, technically I can understand it and 'see' their models, and it's far too complex for what they are trying to achieve.
Wow, that's turned around quickly - yesterday you were very confused and nothing was clear. I'm glad you now fully understand it. They haven't actually confirmed a lot of things about the modelling they're using so that's even more impressive

If my team at work spent this long to build something that is showing so many faults (yes,, I get that's my personal view) then they'd be getting a talking to.
What faults? (I know there are faults, I'm just interested in what you perceive them to be)

I'm sure it will settle down at some stage, the noise will stop and people will resign themselves to where they've landed.
Hopefully!
 

Whorty

Gets free watts from the Atom ;)
Location
Wiltshire
Have you never heard of seeding in sports like tennis, (or virtual bike racing)? It's ranking players based on their performance in a previous set time period (eg the last season in tennis, or in ZRS in Zwift, the past 90 days). https://en.wikipedia.org/wiki/Seeding_(sports)

The word "seed" has multiple different definitions depending on context and Zwift can use it however they like. You may as well argue Zwift's use doesn't describe a plant embryo (this is such a pointless issue to get hung up on, honestly!)


Wow, that's turned around quickly - yesterday you were very confused and nothing was clear. I'm glad you now fully understand it. They haven't actually confirmed a lot of things about the modelling they're using so that's even more impressive


What faults? (I know there are faults, I'm just interested in what you perceive them to be)


Hopefully!

ok :cheers:
 
Top Bottom