PD4.1 causes odd RoboPacer behavior [KNOWN ISSUE - FIX PENDING]

I’m pretty sure @James_Zwift said he tested this and reported that the pacer got dropped (presumably with PD3).

I have no issues with the slight reduction in overall pace. I could add on 2-3km at the end of that was an issue.

However, there’s definitely something odd with the new pack dynamics in larger groups. As a lighter rider, compared with previously when well-placed in the middle just behind the front of the pack, on flats I will drift to the back much more easily, and the effort required to come back through the pack is much harder and has to be sustained for much longer. Appreciate that uphills are where I can “catch up”, but again I feel it’s appreciably harder than PD4.

In larger groups, it’s almost as if the pacer is experiencing something different to the pack, and position in the pack is now much harder to achieve and maintain.


As a heavier rider (79kg), I notice exactly the same behaviour you described above.

That tracks.

1 Like

Okay, so it was speed.

Would also like to add that I had to coffee break and while doing so, speed downhill faster than D.Bernie losing a 2.5x mulitplier, which no big deal. Came back and I was 100m+passed the pace group stuck with a couple riders that rode ahead of the group… Is that suppose to happen, even though I CB in the pack with the RP?

1 Like

I think the CB only latches onto 1 rider – whoever you happen to be riding closest to? Not sure if you can CB onto a RP itself.

1 Like

There doesn’t seem to be any way to control it. I’ve had experiences where I started a coffee break inside the pack near the ride leader and was ridden out through the fence by whoever it attached to. I’ve also had the experience of having a rider use coffee break for me to sweep them back to the group and then when a rider passed me the coffee break rider appeared to switch to them instead. That was actually nice as a sweeper relay but I have no idea how it’s supposed to work or how you would manage it. It would be ideal to be able to pick a rider to attach to with the coffee break, so if for example you are riding with a friend but some stranger is passing you, you can be sure that you stay with the friend and not get dragged up the road.


I’m thinking the draft cone is having it effects on the pack. I rode with yumi for over 3 hours today (sisters course) and almost every instance on and entrance to a hill she would shoot up and out the front by a second or so. Then she would hit the wind and fall back to the back of the pack. It would take 3.5w/kg+ to keep with her at these times of surges. It didn’t matter if someone was in front of her or not. One time she was right beside in the middle of the pack and when the hill started she shot right to the front and was a second or so in front…


Looking back at the data of the exact course and PB, It is over 10 watts of more power to stay with the bot…

1 Like

The Pacer can’t get dropped from the Pacers group ride. Do you mean that some folk rode off the front?


I wasn’t there for the test so I’m not sure what I would make of it

I’m not positive, but I think the TT bike testing might have been before dynamic watts were begun.

The hill surging has been something I’ve noticed for a while, so pre-4.1. It is guaranteed to happen on this volcano circuit on the little hill just past the switchback. The pack usually surges forward going up the switchback, and the PP falls to the back. At the top of the switchback, the PP starts gaining momentum and racing back up through the pack, then just surges up and over the next little hill as though it doesn’t exist - usually flying past riders that ease up to let the PP catch up.

4.1, I’ve noticed difficulty hanging in the pack at a constant pace. I’m finding a lot more yo-yoing, and a higher overall effort/concentration needed to keeping from slipping off the back or surging through to the front. I attribute this change in my case to the increased difficulty needed to move forward in a pack and steeper draft fall off. If you get stuck on the back, there isn’t as much draft to pull you up compared to PD4, and a much more pronounced “stickiness” that holds your position back there. To move up, you have to break the stickiness, but if you over do it, you’ll move up to far/fast and slide up to the front and lose the draft again. Then the cycle repeats as you drop off to the back again. Zwift insider has a great write up on the wasted effort: https://zwiftinsider.com/road-bike-drafting-pd41/

Last thing I’ve noticed is the “narrower” cone means that it’s much easier to lose draft in 2nd or 3rd wheel off the front because there can be small gaps between riders spread across the road that “penetrate” deeper than before.

I hope these issues are something I can adjust to. There is so much unconscious feel and anticipation in Zwift pack riding that I usually notice subtle changes for a few days. But, the new “stickiness” results in much more wasted effort than before and isn’t as intuitive, so it will require more thinking and less feeling to master.

To me, the way to give breakaways a better chance is to fix descending, not nerf pack drafting.


In a RP ride this morning, a fair group drifted away from the RP and a much larger group, and I have to say the smaller pack felt much more stable and more like previous dynamics. When the larger group and RP caught up and re-combined into the larger group, it became much more difficult again. A huge amount of effort is wasted whilst trying to maintain position in the RP pack that never previously occurred.


Thanks to everyone adding to this discussion.

@Alex_Beebe Your observations mirror mine, particularly “wasted energy” trying to maintain position.

Interestingly, I was in a D.Maria pacer group with none other than @Eric_Schlange_ZwftIn of ZwiftInsider.

I did NOT think I’d be able to keep up considering my struggle with the lower pacers but I could, sustaining 0.2W/kg LOWER than advertised power by staying in FRONT of the bot.

Eric confirmed this was always a thing, as well as being behind always requiring more effort.

I wonder if PD4.1 hasn’t so much “broken” the bots but simply exasperated exacerbated already existing quirks.


I agree…

1 Like

Very insightful and goes against my intuition. I would have thought that hanging deeper in the pack would require less effort, and I guess I’d be wrong, ha.

Each dynamics change has a bit of a learning curve to master, and I think this will be the same. I Zwift just about every day, so these changes, even though small, usually throw me for a loop.

1 Like

Why do you think descending isn’t better with PD4.1?
Group is so much slower downhill…there is no Churn if you are not pushing ‘‘NO FREE SPEED’’ :slight_smile:

There is the same Draft at the back of the group…the reason that it feels harder is because you want to move up. If you want to move up you need more power to pass riders…nothing comes FREE it’s the same IRL. When you are in front of RoboPacer it feels natural cause you relax and you are doing minimal effort just to stay in that position :sweat_smile:

1 Like

This is factually incorrect.

There are some pretty clear observations a rider might be able to sustain 0.2W/kg less than the bot while ahead of it, but need to output bot power or even above (with significant “churn” and wasted energy) while behind.

I’m describing minimal effort to maintain position, not move up.


Would love to see your Facts…do you have Strava links of this activities. Something measured by feel is not correct. You will also need RoboPacer data to compare…with Dynamic Pacing you can see RP doing 2.0Wkg but it can actually be 1.96W/kg or 2.04W/kg.

Facts are:
Riders heavier than 75kg will have lower W/kg and lighter riders will have higher W/kg for same speed.
Taller riders than 175cm will need more Watts for same speed and smaller riders will need less Watts.