Summary of my doctoral thesis of the new Pack Dynamics 4.5
Extremely easy to stay in wanted position within the Peleton. Not moving forward or backwards in the same manner as today’s PD.
When moving up through the peleton you don’t get the slingshot moving in to the front. Not even when coming fast from behind in super tuck. You just smoothly tuck in among the first part of the peleton. Also you don’t loose speed after hitting the front, or if it is the peleton that’s not moving as fast as today, resulting in that you don’t drop to the back of the peleton as easily as of today’s PD.
Got the impression that the Peleton moves slower when no one is pushing, so bringing back a break away requires higher effort from the peleton. Really consistent effort with higher power in the front, than the break away.
Easier to get dropped when there are attacks, so more risky to hang around at the end of the peleton.
Did not.notice any difference in the sprint, but I was top -5 during the last 600m, so sprinting moving through the peleton was not evaluated.
Good experience, will most likely result in more attacks and break aways. Felt too easy to find a sweet spot within the peleton without moving backwards or forward… (too easy due to I am quite good at it today… )
Why didn’t you finish the race? The numbers you were doing and others not finishing are very important for this test. The Breakaway can’t happen if someone will do 4Wkg/300W for 1 hour.
I have done both tests of PD v4 on this course…pictures attached. On 21st dec the breakaway of 6 riders was able to stay away because we stoped the Pack from chasing…yesterday every move was shutdown as soon as there was a few seconds of gap. The climb from the desert to Titans had almost the same time.
To me it was easier with the new version to control my position and to follow the moves. It was hard but it didn’t feel like iTT as before.
@James_Zwift and @DavidP we would love to Test PD v4.5 at our VirtuSlo Tour Fight for GC Monthly series. Stage 1 starts next Tuesday We have a various course selection with more equal Pens. Please consider…i think we can get great feedback.
Was meant to be on an easy day and planning on sitting in, couldn’t find the course so was hoping it was flat. Of course, I got sucked in to trying stuff and my legs definitely weren’t feeling good, only just hung on over Titan’s Grove the first time. It got to the second time up there and I pulled the plug from the main group, and there was no-one for 1:30 behind me so just cooled down and called it a day. It’s on Strava if the numbers are of any use?
For me it’s perfectly fine. @James_Zwift if you can turn it on, it’s a thumbs up from me.
We also want to start to have steering enabled for all events, so if you accept to have that on would be great too.
My experience was good in the 6:30pm UCT C race. Was a small field of 11, took it easy until Titans first go and I went hard to be first over to test supertuck with zero draft, another rider was with me. I couldn’t tuck when behind him and could when in front, this was good. The rest of the riders couldn’t catch on the decent. We both swapped turn in the desert and we were holding the chasers. There was one big effort from chasers sustained at 4.5wkg but we lifted to match and they couldn’t close. At one point the chasers messaged to be calm they outnumbered us, but this wasn’t as much help as previously.
One guy again tried to bridge, we let him as 3 would be better, we kept swapping off. Unfortunately at this point some riders left the race as they decided they couldn’t catch. We were now 3 vs 3 and 1 way off the back. Attacked again on titans and dropped the guy who had bridged. He still held off the other 3 until the end. 2 of us still took turns and had a sprint finish.
Was a tough race with HR at threshold from Titans lap 1. Harder than this course normally would be. I think that was down to being committed as the break was possible whereas normally no one really tries to get away as it’s always brought back by the pack and just ends in a mass sprint.
If @DejanPresen gets this for his series it will be an awesome test as his series is already great with custom categories and committed, experienced racers.
I rode most of the B 18:30 UTC race (from half way into Titans 1st lap) in small groups of 2 or 3. Its hard to say for sure but it seemed easier to take turns in a small group without accidentally overshooting (and autobraking is actually useful in a small group).
(BTW looking forward to results based cats so I don’t have to ride against people out of my school never mind class.)
seems good. i only lost 5s going down titans grove + the rollers instead of 50. not sure if PD made any difference on the flat but downhills are much better. thank you for the work
David I loved the new pack as they have commented the feeling of going in a group is more real and pleasant I do not even want to think with the direction connected. About what they say about the visual bug my opinion would be not to disconnect anything and try more, I personally yesterday I did an event of my club with the pack activated and I did not notice such a bug. Now that I’ve read it here, I’ll keep an eye on the next event. The only thing I could say negative is that the peloton is a damn locomotive and make a breakaway is very difficult if not impossible.
Game NEEDS to be visually accurate and predictable in order to understand the race dynamics. Without physical feedback visual queues are all we have. looking forward getting this visual tweak in game so super tucks are represented accurately.
Kudos for the good work here David!
Little bit of a side question as compared to the others in this thread, but…
…if you’re with a pack and without steering, you sometimes get forced off to the side of the group by the normal (random) interaction of the riders. You basically have no ability to determine whether you will be directly behind others or not. Are you still getting as much draft out there to the side in the Zwift pack as you would be if you were directly in the middle, behind the other riders? Or are you being (arbitrarily, since you can’t really control your side-to-side positioning) penalized with less or no draft?
I trust others will correct me if I’m wrong, but my sense is that there is less draft on the edge of a pack. My sense is also that, if you are on the edge but then accelerate relative to the pack, you tend to move towards the middle as you move up. In any event, I don’t think of being moved to the side as a “penalty,” since any location within the pack seems random, so it seems reasonably fair to everybody (setting aside the case when you’re moving up).
Something I was thinking about from a recent (non-PD4.x) race is that the same riders were attacking the climb through Italian villas every lap and getting a gap, but the pack would catch them within a couple km. IRL it seems like the elastic snaps after a few bouts of these, the break get established, and then it’s a question of whether the remnants of the peloton that didn’t make the break can reel them in before the finish.
Instead, what happens in a typical Zwift race is the selection happens out the back — pace is forced on the climbs, elastic between the bunch and the break never fully snaps because the break is too easy to catch, but the peloton just thins out through attrition of riders that can’t burn unlimited matches going out the back.
So, these comments make it sound like we are getting closer to my impression of a real life racing experience.
I was thinking about this in the middle of the night…
As long as you’re tinkering with super tuck, what do you think about disabling it below a certain draft level rather than any (?) amount of draft? I’m thinking about situations where a group is strung out, and I’m trying to catch a rider a few meters ahead. IRL, at high speed, there should be enough draft benefit that I can catch the rider in a tuck (i.e., while coasting). As I get closer, and the draft benefit increases, I have to sit up to avoid crashing, or I can steer out of the draft and stay in the tuck.
I don’t know how the rider placement algorithm would work in the situation of “steer out of the draft to overtake in the tuck” but it seems like you’re going for a system where the rider automatically sits upright when the draft is high, so you’d have to break through the barrier of the rider ahead to regain ability to tuck again.
Anyway, I thought it would be worth considering some maximum draft level below which super tuck is allowed…maybe this needs to be smoothed (not sure how volatile the draft figure is) over 3-5 seconds?
We have a various course selection with more equal Pens. Please consider…i think we can get great feedback.
