He hasn’t questioned it. He says it happens, which is why double draft isn’t widely used. It’s not the cause though, hence trying to fix the washing machine effect to be able to enable double draft in the future for a much better race experience.
Yup, I commend David for engaging as much as he has on this thread. It’s clear he wants to do the correct thing and has put a good amount of thought into it. When there’s such a wide array of knowledge on the subject it’s hard for someone to really completely both make everyone understand the nuances and agree to an approach. I’m actually pretty excited he’s thinking of adding a bit of UX to show when you might be able to save a few watts in the draft and do hope they can land on something that makes racing a bit more interesting than it is today (one blob until the sprint for any race without a large hill, with very minimal ability to have a small group break out effectively ever).
That said, I don’t understand much about how to approach this given the inputs Zwift has available, and the physics model in use, I also have no actual IRL race experience, and very little IRL drafting experience, so I haven’t engaged personally on this thread, but I am happy to be a part of any further testing when an event shows up I’m able to make 
2 Likes
This post touches on something I’m not clear on… is the aim to replicate outdoor riding? Or is it to provide the best gameplay/experience for the end user?
If it it is outdoor riding, are we losing power ups over night?
If it is the best experience in gameplay, what’s the issue with riding through someone virtually?
I also notice there are repeated queries on the sticky watts that have been ignored and put down to ‘urban legend’ - rather than dismissing them out of hand how about addressing them directly?
5 Likes
In David’s first post on this thread he mentioned the following:
“The main goal of PDv4 is to make packs behave a bit more like IRL and try and bring possible tactics into races, for example, give a bit more chances for breakaways to succeed.”.
I don’t think that it would imply reducing any gamification, or removing any gaming elements such as power-ups however.
1 Like
Totally understand the sentiment with this post, although I disagree a little bit with the conclusion. It can’t be easy juggling all these variables and I don’t envy @DavidP, but it reinforces the need for more communication/transparency. Consider this thread so far - lots of thoughtful feedback offered/questions asked, but then we discover that a critical game mechanic to get PD 4 working correctly wasn’t actually working (from what I understand). I’ve largely sat on my hands with regards to feedback (a whopping 22 posts in nearly 8 years 
), mostly because I trust Zwift to get it right. But it is frustrating to see little headway made on critical game mechanics over that time frame, especially when combined with some other notable missteps.
I know time and resources are limited, but a community test environment that can peel off feedback into a more productive format and allow the devs to push changes out more quickly would be a wonderful addition. It would also help knock down the mythology that surrounds some of Zwift’s game mechanics that permeates nearly everyone’s views to some degree.
4 Likes
It’s hard to provide detailed feedback without being familiar with exactly how the game physics work.
I have many years of experience with IRL group riding and racing. I’m very happy to provide feedback in the next test event.
As I’ve stated numerous times here, IMO the biggest issue in Zwift racing is the physics of groups. It destroys virtually any chances of dynamic racing. Anything, that goes towards improving this situation is a step in the right direction.
The end goal has to be an effective system that requires the chasing riders to match or exceed the power (w/kg) of the breakaway rider, to catch them. As it currently functions, it’s wildly optimistic. A breakaway rider can be doing a full 1 to 2w/kg more power than a chasing pack. A pack which, zero of the riders are even matching the breakaways efforts. It’s not how it works IRL. IRL the riders on the front of the peloton have to match the breakaway. Usually, for an extended period. This normally results in frustration at riders sitting on. Hence, the riders on the front give up, people repeat this and the break gets away.
The riders on the front of the peloton actually have a 3% advantage over the breakaway from the air mass behind them pushing them forward. I’d be inclined to engineer this out in Zwift to make attacks even more likely.
If these changes fix pack speeds and real draft can be enabled, Zwift racing could become very exciting.
Thanks for the feedback @DavidP
Looking forward to trying potential solutions.
9 Likes
Hi James, Is it possible to make this happen. If anyone can compare the change in pack dynamics it would be the people who ride in the PP packs every day. Same PP, same riders at the same times. Not everyone races and it is likely that most people who ride PP in Zwift also do club rides IRL. One day of PD4 in Makuri Island would allow everyone who wants to experience PD4, the opportunity. I don’t want to enter a race but want to see how the physics have changed and would like to comment on the feel of the changes.
I’m not making anything available with PD4 until David’s changes have gone through development and release.
4 Likes
Re: the 10% power and overriding auto-braking
-
As a preliminary matter, does the auto-braking get applied based on not having a path to pass a slower rider?
- I would assume this means multiple riders in front or next to so that there is no “open path”
- this seems important, as I should be able to pass if I have a higher speed (from power/draft-based speed boost combo)
- I wonder if this explains any of the issues people are describing towards the back of groups?
- it would seem with a “strung out” pack, I should usually have a path to pass riders in front of me.
-
Would it make sense to simply slow my speed to the rider in front “immediately”, and then if I keep generating enough power/draft to have a higher in-game speed for X secs, I can “nudge” through or past the riders in front?
It would seem this would go far to stop churn because I would have to re-accelerate from auto-braked speed to the speed the power/draft model determines I “can” go. It also allows for a tweakable “nudge” system.
It does seem like what you described would also achieve something similar, I just wonder if it would be better to compare the output/speed of riders in question to each other rather than rely on the primary rider to alter their output to initiate the pass (disable the auto-breaking).
This would create an interesting tactical situation. It would always be possible to move up through the group, but it would require a consistent application of higher power to move “row-by-row” forward. Alternatively, I could wait for the group to stretch out and I might be able to slip forward easier relying more on draft-boost than power.
… someone did this today, sticky watts, behind me in the ZRL race … didn’t help them 
1 Like
They weren’t using sticky watts if it didn’t help them. They were probably using a trainer where it doesn’t work.
Can you please describe how auto braking feels for you?
I feel intermitent resistance increase in pedals. Like an ABS brake. Is this auto braking? I cannot figure out if I have a problem with my setup.
Is this auto braking enabled in Tour of Makuri? Can I disable it?
It’s not enabled anywhere. Sounds like a trainer issue.
1 Like
nah they did catch me up or go ahead, but i had enough power to leave them anyway
I may be missing something here. If auto-braking is slowing people down to keep them from moving up through the pack, why not just use the IRL solution?
IRL you don’t just get to move up through the pack. You go to the side and move up the side. Out of the draft (or mostly out of the draft.)
So auto-brake and show some indication so people can sit up if they like but if they don’t sit up then move them to the closest side and they can move up just like IRL, or if reduced draft means they slow down, then they drop back.
Effectively it (like IRL) means you have to work hard to get to the front as you will be working your way up the side. Or you can sit in, and as space opens up you can move up slowly. Again like IRL.
This probably also requires that people at the front get moved to the side and shuffled back like IRL when they drop their watts.
When this conversation comes back to the fore… Can we use the following type of imaging to show the change…
This is the current draft, this is how it will change etc… Even if it just density of colour to show where there will be more/less draft
4 Likes
I would love to be able to enable that in general just to see what is happening with the draft. Just out of interest.
A little bit - but there’s a pretty big penalty for weight on CdA as well.
Any news of PD4? Is a new test planned in coming days?