Also, as a counter to this descents used to be a nightmare as someone much lighter than the bots.
I’ve ridden a lot with Maria in makuri this week and she’s still dropped on every rise. Would be even worse with static power.
This is how it all started. If it was only slowing the bots on the downhills then it would have been fine.
In my opinion the fix to this is to have the bot slow down before cresting the top of the climb, but that would take a lot of programming to have the bot look ahead.
So the fix was to make it super hard for heavier riders to stay with the bot to the top of the climb then the bunch will be smaller and only the strongest will go down with the bot and they will be strong enough to keep up.
1 Like
Try yumi bot then? You’ll have to work more on the hills.
Is the better question then, why is it taking months to make a simple UI change?
It’s nuts that it’s one or the other.
Pace Partner - Steady
Pace Partner - Variable
Both. Otherwise known as options…
Problem solved.
1 Like
You must be new.
7 Likes
The UI change would be easy to mock up, it’s having to build the logic behind it then test it and also remember it has to go into the stream of existing work (ie, what’s called a release).
That’s why it takes time. If you do this something else probably doesn’t get done.
I tried this out today. Rode with Miguel on a gravel bike (me, not him 
). Three main learnings:
-
The plan fails completely if there’s much route that isnt tarmac. You end up pedalling at about 1.0wkg as poor Miguel drags his Tron through the gravel.
-
On 99% tarmac, on Three Little Sisters, I averaged pretty much exactly the same power as I would if I’d ridden on a fast bike with Maria … only achieved in a very different way.
On the flats I was about 0.6wkg higher than Miguel. On the climbs, I was maybe 0.2wkg higher than him. As a result, on the flats I was a bit higher than if I’d been on a fast bike with Maria, and on the climbs I was a bit lower than if I’d been on a fast bike with Maria. So, on a route with some climbing, it all pretty much cancelled out. Bottom line is that the effect of the gravel bike is higher at higher speeds, which I guess is what you’d expect as the main difference is rolling resistance.
It’d be interesting to have another go at it on Tempus Fugit and see if the bias towards flat and fast riding causes a meaningful uplift in the wattage.
- And relevant to this thread … riding on a gravel bike with a lower RoboPacer actually resulted in a steadier ride. Less power needed on the climbs, more needed on the cruisy bits.
2 Likes
How about not having the robo pacers on the tron bike?
Maybe put them on Giant TCR with some modest wheels? That could also adjust the pace.
1 Like
I have stopped using the robopacers at all now, I used them for zone 2 training which was the bulk of my time on zwift. I dont think I am alone, there are less than half the usual number of people using them at the time I normally do. Since this was the big thing keeping me using zwift I am now looking around at other options like rouvy or wahoo to see if they have something similar.
5 Likes
I usually see quite a lot of people with the robot groups. Constance yesterday had 12 or so, one of the blue pace groups had a very large number (we passed them), the coco group was also huge. The slower ones also had 40 or so.
You’ll still get your zone 2 training even with a smaller group or riding alone with TT bike.
On the flat routes to avoid the surging on the hills. 
Isn’t there some irony in the idea of wanting to ride with a large group of people for the social interaction, but wanting everyone to behave the same, no surprises nor deviations from the norm?
I don’t think so. If your goal is to have a social experience around a zone 2 ride, then that’s your goal.
Maybe. Can’t you chat or text with others on the road, just cruising anywhere you want during a training ride set at zone2 wattage?
Not sure what you mean. It’s hard to get a consistent ride on the road due to stop lights etc. Zwift is a great place for long sustained Zone2 efforts from that perspective.
In terms of calling people, it’s more effort to set up a call with people you know vs an always-on group ride as you have with pace partners. The barrier to entry is very low which is great.
Sorry, I just meant any Zwift ‘road’ or route. Not outside.
Some worlds have many hundreds to thousands of people just riding around, all the time. If you want to be social and chat while you ride, there seems to be lots of opportunity if that’s what’s being sought, and if you’re not looking for a naturally dynamic group ride. Just solo or do a training ride… all of those people will still be riding around.
Ah gotcha.
That said, you can make the same argument about a dynamic pace group - why not just go out in the worlds as they are and do your thing with people just riding around in a dynamic way? In general most of the people riding around the world are actually doing a dynamic pace, so you don’t really need a consistent pace bot to get a dynamic experience with most folks riding around.
The answer is (both in the dynamic, and static case), it’s much more convenient to jump in with an always-on 24hr group that has a pre-set pace rather than going out and looking for a pre-set pace randomly with people riding around. It’s consistent, it’s always there, and you know what you’re getting into in advance so you can use it as one of your planned rides.
1 Like
Part of this could be that it’s the northern hemisphere’s summer, so many people will ride outside now that conditions are good. After September/October, numbers everywhere will pick up again.
1 Like
The problem with the dynamic robopacers isn’t that they are dynamic, the problem is that they are too dynamic. Ramp rate is too severe in both directions. On Maria this morning, I had to coast on virtually every decline because robopacer went from 2.2 on the flats to 1.7 on the declines. And this was the flat route. I can’t understand why they don’t try playing with the ramp rates, it would be so easy to adjust and get feedback.
I rode with Maria on Big Foot Hills for 90 minutes yesterday and didn’t coast once. It took some concentration, though, as the RP radius was less than 10 metres. It possibly helped that I was tired after weights the day before, so I was having to overcome heavy legs to keep up more than having to back off.
IIRC, they did adjust the ramp rates based on feedback back in Futureworks days, and where we are today is the result.
Edit: Square brackets 
2 Likes