Pack Dynamics 3.0 Update [October 2021]

I agree. We only didnt get Doppio record because we were too conservative for the first 5mins as Zone 1 and first TTT team off the rank with the new draft and didnt want to muck it up. 5mins in it was like “this feels far too easy for these pull watts” (the pull watts were our normal 2.0 watts), so we upped them significantly

BTW it was near impossible to get a course record in 2.0 if it was set prior to Apr21 no matter how many watts/wkg

1 Like

I’ve encountered the flashing avatars twice now. First was with a C. Cadence bunch the other day, where it came and went with varying intensity, but was steady and very rapidly flashing for a few minutes between Saddle Springs and the Oasis on Tempus Fugit.

Today, it happened in the Ride with Reggie event. Any time I was in the dense blob around Reggie, some riders at the very front of what was visible and most of the riders at the very back were flashing in and out while the majority of the riders were fine. If I pushed further forward or dropped back, it would abate and then stop flashing.

It definitely has something to do with the density of the bunch, having the 100 rider display limit in a very small area.

First less than satisfactory input on Pack Dynamics 3.0 .

Did a session with Angela this afternoon .

As usual not so many in the group as the other pace partners ( although more than normal ) - around 6.

Angela as usual claiming to be doing 4.2w/kg. The problem was quickly evident that this was a completely different experience than previously, which apart from hills ( where all bots seem to struggle) she used to set a decent enough pace . However today it was almost impossible to stay in the draft and not pass even riding at well below 4. That seemed to lower the speed of the bot even further of course as riders passing at much lower watts then taking the wind and so it continued getting more pedestrian . End results was a real slow group indeed . I gave up in the end and just pushed on … not even doing 4 . I suspect Angela might need her w/kg updated to cater for this new dynamic.

1 Like

Could the problem be her weight at only 65kg? Brevet is a lot heavier at 82kg.

Anybody else notice the occasional absence of the “close the gap” messages? On my ride today, they were conspicuously (but not consistently) missing in situations where I would have expected to see them, both when catching up to a group in front as well as when inside a loose group with considerable gaps between riders.

This was a bit complicated in the previous version as well, but there the main issue was that the sideways movement could throw you from full draft to >5 m behind the rider in front in an instant. Back then, at least the avatar riding position would work as a backup cue, but now even this didn’t seem to be reliable.

Hi @shooj, few issues with PD3 going up the alpe.

Windows, no steering enabled on event.


Haha, that first bit is hilarious. :rofl:

I’m gonna speculate that it’s because they’ve not ‘widened’ the usable road on the Alpe like they clearly have everywhere else, due to it being its own little world (with the segments etc). Not enough space so the mass of avatars start being pushed into one another and squashed over the edges.


Yea @Dave_ZPCMR IIRC we saw similar things in the testing on Makuri where the roads were more narrow. Avatars getting pushed into boulders and posts on the side of the road.

Some refining of the algorithm that defines the shape of the peloton (especially on more narrow roads) looks like it may be needed for further refinement.

1 Like

Am I the only one who looks at this and thinks “awesome, the graphics break instead of people getting stuck behind others in the sticky draft or because the road’s too narrow!”

I’m ok with that as a compromise; I’d rather my avatar passes through walls than gets held up. :wink:


Thanks for the report and the video. I’ll pass it along to the team.

Please confirm it was this event?

I’d much rather ride through solid rock than be stuck up someone else’s ■■■■! :smirk:


That’s the one @shooj

A possible solution for the washing machine problem in races with big packs could be that you get less draft from riders that are moving slower than you, so that when you come from the back with more speed you are slowed down from air resistance to the same speed as the bunch before you reach the front.

Hi @shooj, on the 3R Endurance Steady Ride [~3.0 // 2.5 w/kg avg] ride one Sunday the flashing of avatars was confirmed by a couple of riders, while others reported no issues (maybe those were in the front of the pack, where I didn’t have these issues as well). I have cut and converted the clip into an animated gif to avoid uploading on third party platforms; if you would like to have more evidence or the clip itself, PM me and I’ll mail it.

1 Like

I actually think that is not related to pack dynamics but to to how Zwift only draw the closest 100 riders and they focus more on those infront of you than those behind, so if you are in the back of a huge pack you will see the once behind you drawn as they get close

1 Like

ZRL TTT in London was rough. Not from a 3.0 performance problem, although the experience in the draft was different.
In France last week we found we could get quite good recovery in the wheels- this week it was super super grim trying to hold a wheel steady, and when we did the recovery was much much less than last week. And worse than 2.0 where we won on the same course with same number of riders and similar time.

Reeeeally good for PB power lol, not so good for efficient TTT racing.

So a bit of Sticky was maybe good for something???


I am having the identical experience being on the right in the bots it only seems to happen with the pacers and riding diagonal. Not sure if this has been flagged. It hasnt happened in a group ride yet.

I agree performance today in London ZRL TTTs seemed disappointing. Pack splitting sideways across the road too. There are a lot of small gradient changes in London and corners which make it hard to stay smooth, I wonder if that has something to do with it?

1 Like

Thanks for the GIF. I agree with Gerrie - I don’t think that’s got anything to do with PD 3.0 either. Looks more like the way avatars are drawn when there’s a lot of riders nearby to lower the RAM footprint.

In the ZRL TTT week 7, we had some incidents of riders being slingshot to the right. See picture. This happened just a few times and in the area where the screenshot is from.

EDIT: I’m using BT to connect and have a Sterzo that is not connected.

1 Like