Yep, did that both laps, first lap with great affect… Was ~50th and off the lead group on lap 1, by the time we reached end of lap 1 I very briefly had my nose in the wind.
On lap 2, it might simply have been others ramped their power up on that long descent with ~1Km or so to go and I either didn’t react quick enough and/or hard enough.
@DavidP it’s all gone quiet on feedback and there are now lots of events using PD4 so that must be a good thing. Do you think it’s good enough to launch? I’d love to see it rolled out now.
Tuesdays VirtuSlo race was PD4 and I never got autobraked once. Hard race again but everyone was trying to make a break so it was lined out from time to time. But all good. Decent of the petite KOM seemed good and small groups stayed as small groups, no avalanche of riders.
Thanks for the feedback. Yes, I think so too.
We are just aligning internally to decide on dates related with RoboPacers testing and full roll-out. So it’s close 
Hi David, i had a few races in the past week but i didn’t have any time to write a reply here…i made some quick comparison for VirtuSlo 4Endurance Race on Sunday on London Pretzel with PDv4 and from January with PDv3 and Double Draft ON.
I compared the 2 rides and data is very good.
Flat Part around London 30Km
PDv3 + DD - 42:34min (32 starters)
PDv4 + DD - 43:04min (60 starters)
Box Hill Descent
PDv4 Group of 4 vs Group of 12 is 4s slower.
PDv3 Group of 15 vs Group of 4 is 15s faster
But on Tuesday we had a crazy fast race at VirtuSlo Tour Fight for GC Cat B with 45km/h average on the flat part until Petit KOM.
Live Stream
Looking forward for the official release…downhills are way more interesting now.
The descent was great, I was in a group of 4 and 2 chased back on. DOWNHILL!
Hello team,
Is there a problem with the kilometre counter with PD4?
In the following races with PD4 cyclists have commented that the event ended when the distance to go marker still had some km to go.
Do you know of any problems?
Thanks
No, that is nothing related with PDv4. It’s a known problem with some routes, that we hope to fix soon.
Gotham Grind and Beach Island Loop have issues?
Beach Island Loop
Today after 115km they had 2.3km at the finish banner
1 Lap spot on banner at 0m from July 2022
3 Laps from 5 months ago spot on 0m at the banner
This must be a new problem with this routes but sure not related to PD4 cause this has been a problem for a long time.
Roadie for 20 years, 3 month user of Zwift, on Zwift every other day at MINIMUM. C Cat.
Just tried out PD4 for the first time today and being in the pack looks and feels smooth but man… this thing punishes you HARD for resting your legs for a second or trying to catch up to a blob from less than 5 meters away. I have never been dropped so hard, so mercilessly.
Maybe it has something to do with me being 6’5 and over 100kg but I really hope this code is factoring folks like me who need to average 280+ watts to get by.
There is obviously something wrong at the back.
I’m 6’2, 100kg, and also in C. Done 4 races with it so far and much prefer it to PD3. Haven’t had any issues in the pack, but have noticed the pack moves to shut attacks downuch faster, which may lead to delayed reaction if you’re at the back.
IMO, its actually an advantage to bigger riders like us, as we’ll have more of a chance to utilize our higher raw wattage on the flats, either to attack or chase back on if we get dropped on a hill.
Yes, it should be a greater advantage, as I have the ability to put go big on the wattage and breakaway if I choose to.
I took at a closer look at my metrics on Strava and those near me and my new theory is that PD4 tries too hard to bring order to people’s positioning in the pack. If you reduce your power in a pack in PD3, you slowly begin to fall back. It doesn’t look pretty but there’s no surprises. In PD4, if you reduce your power, it visually looks like you’re OK because an invisible rubber band is holding onto you. Lose too much power and the rubberband snaps as you swiftly get rejected from the blob and shot out the back.
I responded with a 600watt burst of power and it did absolutely no good at all. No draft to grab onto.
I’ll give it a try again in the near future but that’s my take.
Very strange - I haven’t seen that in the PD4 stuff I’ve done. I did not notice any difference in backward motion in the pack, although I did notice that the much cushier ride of double-draft sometimes lulled me to sleep a bit and if I did drift out of the cone the difference in power required was a quick wake up. The actual power required when exiting draft cone (on side or back) was the same as PD3 but the difference was greater because double-draft was a lower baseline.
If I’m interpreting this whole thread over time correctly, the actual PD4 changes are: 1) doubledraft is on and 2) if you are moving from a high draft to lower draft situation (like, hitting the wind closer to the front) and you are reducing power then the system will check your speed with an auto-brake. Oh and 3) you can’t supertuck unless you are in clean air
I don’t think there were any changes to how the system tries to avoid collisions or order people or change backward motion or how it models draft (that is: size and shape of cone) or anything, so I’m not sure how that works with your hypothesis unless I’m misunderstanding something.
All that said, I’m a C as well and the VirtuSLO races on Tuesday afternoons are PD4 if you want to give it another PD4 race a go and say hi. I’m usually there because I’m trying to get a feel for PD4 myself. Cheers
@DavidP Are you allowed to elaborate briefly on how the “auto-breaking” is coded for PD4?
I see auto-breaking the biggest downside of PD4 frankly. I understand that it is a necessity to curb the pack speed, but in my opinion it makes riding in a pack too “surgy”. Let me try to explain what I mean.
Before PD4, if you were coming with speed through the pack, you could come off the power and “sail” to the front. Especially for heavier riders, this happens a lot when the road slopes down. I used to be able to use this mechanic to my advantage in races as an ~80kg rider. To me, this is similar to “coasting” in a real-life scenario and gives a brief moment of recovery after a hard climb.
With auto-braking, however, I am immediately forced to the back of the group if I am shooting to the front of the group and come off the power. There is no “sailing”. Therefore, I surge, everybody surges. Going downhill in a pack is a clown fiesta with PD4.
Is there a fixed-amount of power loss applied when auto-braking kicks in? Or is it just reducing the speed to the speed of the pack? What are the variables taken into account when deciding when or how much auto-braking to apply?
I think I had already done that before. Essentially it is a combination of factors like speed (speed higher than those in front of you), draft (within a certain range), power (reducing your power), distance to the bike in front of you (it needs to be close), all hitting certain threshold that trigger speed reduction.
Before PD4, if you were coming with speed through the pack, you could come off the power and “sail” to the front. Especially for heavier riders, this happens a lot when the road slopes down. I used to be able to use this mechanic to my advantage in races as an ~80kg rider. To me, this is similar to “coasting” in a real-life scenario and gives a brief moment of recovery after a hard climb.
Well, what you describe is essentially the problem that PD4 tried to solve. That is what causes the super efficient packs that speed up more than they should. Remember that in real life you can’t just float through the pack, so that’s a bit of what we try to achieve here to a certain degree.
Therefore, I surge, everybody surges.
Why would you do that? If you want to stay in the pack just relax in the draft 
Yesterday i noticed a supertuck BUG again when watching riders from outside event. Do you know if this was also fixed or has the BUG returned with the update?
Check a few riders at the back going into supertuck.
The bug was definitely fixed. This is something related with fan viewing I guess 
I’ll look into it when I can.
The double draft is a big factor here. It became more apparent that the draft cone of the pack is too short, and dangling at the back is very scary if you are just a few meters behind. It feels very unnatural as this is the place that should be one of the places in the pack with the best draft effect.
We are slowed down to the speed of the rider in front when approaching a rider at the back if I remember correctly. I’m not sure what the criteria are for this to kick in. Could it be that this slowdown kicks in too early/too far behind/too little speed difference? I remember we saw this problem in one of the earliest ZI Tiny test races. If the speed difference is big, we should be slowed down progressively IMO. It seems as if this slowdown kicks in for the first rider we could possibly be slowed down by, even if that rider could be moved to the side (and no slowdown should happen). I would have steered clear of that rider if there was some gap up to the rest of the group.
The draft cone was not changed, and it’s not different between regular draft or double draft. The length of the draft cone is around 8 bike lengths and naturally is stronger closer to the bike providing draft and weaker the further away from the tip of the cone.
If the speed difference is big, we should be slowed down progressively IMO
If the speed difference is more than 3km/h there is no slowdown whatsoever.