So, are you suggesting that a 50kg woman would have the same amount of muscle as a 100kg man? a 50kg woman putting out 200 watts would be 4 watts per kilogram, while a 100kg man putting out the same 200 watts would only be 2 watts per kilogram. It’s definitely not easy for a woman to maintain 4 watts per kilogram!
I guess the thing is, if they implement 50kg pace partners that doesn’t help anyone who is well above the current 75kg pace partner weights who also have issues with the current 75kg bots, particularly in power swings on any hill. I would expect anyone who is really heavy would see very large power swings during hills etc.
In Zwift’s initial implementation they took the approach of “midpoint weight of all Zwifters” from what I understand to try to have a reasonable compromise that would work for as many people as best it could. This sucks for folks on both the extreme light and heavy end (for different reasons). They kept them dynamic which amplifies these issues.
If they were to add a new set of bots they probably would want to sit back and think about how to make the combination of both bot sets cover as many Zwifters as possible again so this is inclusive to both heavier and lighter riders (so they don’t need to add a 3rd set for heavier riders etc.).
So instead of 50kg bots and 75kg bots maybe they would do something more like 65kg bots, and 85kg bots - (just examples, but where the numbers would be derived from average usage in Zwift, they could look at male/female weight midpoints etc, as well as the type of Zwifters they would like to attract to the platform). It might be that Zwift is concerned more than we are about ‘average non-cyclist’ weight for instance because they are trying to attract new folks to the platform who might not be the average cyclist’s weight but want a good experience with the platform, I don’t know.
There is definitely an issue of how to make the experience clear and usable when the only difference between the bots is weight because there would then be a lot of bots, and they’d need a good way to succinctly explain the difference between Yumi x w/kg at y weight, vs. Yumi x w/kg at z weight. Maybe just show the power output Yumi 2.9 w/kg (217W), or weight Yumi 2.9 w/kg (75kg).
Then today they have split the 75kg bots between hilly and flat, so would there also need to be a second set of lighter bots on both hilly and flat? Would it be ok for everyone if there was just one set that moved between hilly and flat?
I have not suggested that. I merely used numbers at both ends of the weight spectrum to showcase that none of the w/kg or raw watts are good enough to predict performance but one is less reliable than the other. That despite light riders needing more w/kg to account for the lower weight, they are rewarded with the ability to output a LOT less raw wattage (more so than what real world equations would predict). That’s a heavy bias towards us light riders that Zwift built into the physics.
If they stop describing everything using w/kg it’d be a good start
Just throw a wrench into everything and make the bots taller / shorter… 
The Halloween bots were still the same names though right; they just appear different?
Unless they did something different this year; I don’t pay that close attention to single day events / it’s easy to miss.
That’s a separate issue that needs to be addressed also. But no use holding up or blockading lighter robopacers because the other isn’t also addressed at the same time.
Remember, it is being asked to have ADDITIONAL lighter robopacers. That doesn’t prevent other changes being made or make the existing problem of 75kg robopacer acceleration worse.
It’s always possible later to remove the increased pace that the robopacers do uphill, provided enough people ask for it to be reversed.
It would have been nice if they had argued so strongly against it back when it was introduced as they are now.
They were additional bots. And the argument was that we can’t have additional bots because of “UX” so how did we have those? We had it because those were just done without consultation.
Just get the change done and then work on the next thing in adjusting the power levels done by the robopacers uphill so they don’t accelerate too much.
At the moment they do accelerate too much and people are raising this as an issue. That needs to be another separate request. I know I’ve asked for that before but it got shot down.
I’m not sure what the best route is.
I’m 75 kg and seldom ride with the pace partners now days.
The hill accelerations and decelerations may be partially at fault.
I’m not entirely sure why a slower pace partner wouldn’t solve the issue but if those who are affected state it doesn’t feel like a good answer, then I accept it.
I also don’t know why all the partners couldn’t be 50 kg and people would ride a faster PP and curse the hills!
Would PP on non draftable TT bikes help the situation?
One last comment, I hope we don’t try to fix this with Pack Dynamics.
I have stated in another thread that the ideal pack dynamics for Pace partners is different than for racing.
I’m also not 100% sure on this one either. One point made was that for some they have a set of typical people they are riding with and those all ride with Yumi, or Coco etc… But I’m not sure if what happens is if you go down a pace partner as a light rider if it’s too far down, in which case the experience might be made better with using a gravel bike with a lower partner?
For folks looking for a lighter bot, if you went down a bot (i.e. from Yumi to Coco etc.) what has the experience been like?
I expect this would make things a bit better because there would be less variability in pace between pace partner rides with fewer folks vs. a huge crew which will invariably have people ahead of the pacer. The issue would be that you might only have to go a fraction of the advertised pace given you would be drafting and the pacer would not, so they’d need to figure out how to resolve that I would think.
Yeah, the pace partners often seem to make quite big jumps ahead quickly inside the blob, probably due to whatever oddities of the Zwift pack dynamics. Putting the pace partners on TT bikes (or other measures to same effect) would certainly make things more stable.
What I still don’t understand is that I seem to have to push more W/kg (as well as more W as I’m not lighter than the bots) than the pace partner just to maintain position in the blob behind the pace partner whereas riding in a smaller group ahead of the pace partner conforms to my expectations much better.
1 Like
I think this might also be a pace dynamics thing, where the folks behind the pacer have incrementally increased their power to get back to the bot, and to overtake you need to go above their power etc… so maybe there’s more brake feathering or whatever to make moving up a bit more difficult. It definitely ‘feels’ like the bot can just shoot up the blob more easily, but it’s probably a pace dynamics quirk. For me personally on my pacer rides the overall power average seems to make sense even though it feels harder when trying to get back to the pacer rather than riding ahead if them.
I believe our experiences as lightweight riders on Zwift are quite similar. We both can sustain a high wattage output or of around 4w/kg or more for extended periods (if we want to keep talking about w/kg) allowing us to ride with almost any robo pacer or group we choose. However, for most lightweight females, the reality is different. What do you think the percentage of light females as strong as you on Zwift is? I imagine it’s not very high.
We should standardize on terms. Wattage is wattage. w/kg is power to weight. So high wattage is different than high power to weight correct?
I would guess the drops multiplier is the least of worries for Cat Allen given her enormous amounts of riding. She could probably buy the entire Zwift application with her level of drops and XP. 
I also think the riding in front of the robopacer is much easier and smoother, but it can also end up speeding up that group more than some people want so I’m careful to avoid that.
In the Constance group who are usually riding when I am, the deliberate steady 5.0w/kg in front of the bot by certain regular people gets called out very quickly with those people asked to ride away on their own or slow down and stay with the group. That’s good because it keeps things better for people who actually want to ride in a group. These are all top level racers who can go very fast if they want, but they are at least respectful of riding in a group. It’s nice riding with them for that reason.
2 Likes
In a small group (like Constance) that is possible, but with Yumi and Coco on flat routes you can’t avoid people in front of the pacer. If you’re with a pacer and 100+ people the pacer is pretty much always in the draft in my experience regardless of what one person decides to do.
Having the pace partner on a TT bike, so they’re faster than a road bike for the same output but don’t get dragged along by the draft, should keep their pace in a good spot.
Smarter dynamic pacing too, they shouldn’t be holding the increased effort for a long climb.
And change their messages to sound more like a stubborn shop ride leader.
“Stay with me, no need to chase”
“don’t stretch the pack”
“let’s help the people at the back”
“if you want to ride faster do it yourself”.
4 Likes
That’s true as well, and that’s probably applicable for all robopacers of all weight types. But first we need some lighter robopacers to supplement the existing ones as a start which is what is being asked for here.
I like the idea of those messages, maybe they could call out particular rider names too. 
Really? That’s exactly how most Zwift group rides work. Endurance pace on the flats, tempo/sweetspot on the climbs. 
I have never done a zwift group ride that followed that pattern. The group rides not a big part of my zwifting so maybe I just haven’t come across those groups. It’s always been the leader trying to maintain a steady effort at the target pace, and usually some consideration of keeping the group together.
Even so, organised group rides have everyone starting at the start with an event description and a ride leader letting everyone know what to expect and whether the pace is going to change.
Pace partners are a 24/7 group that you can join at any time, I would think that most people are expecting a reasonably steady effort. There is a natural surge to keep up the momentum as you start up a hill, and a natural lapse as you get onto a descent, which is all the dynamic pacing should be trying to mimic.
Particularly uphill I feel like pace partners keep up the surge long after the momentum has faded. It becomes a sustained effort at a zone or two higher than people are expecting.
If it’s most, then which ones, in particular?
The ones I’ve been in (such as 3R) are always good at using the electric fence on on the climbs, particularly say the Bologna climb to make sure everyone stays together because otherwise everyone takes off and the group is split.
Now, we still want lighter robopacers to supplement the heavier ones, please.
That pattern would emulate an IRL ride, where the mantra is ‘work the hills, coast the downhills’