RoboPacers - An Open Discussion

RoboPacers
121

David, you are missing the point. I don’t want a flat power profile.

Basic science: If I am the same weight as the RoboPacer I need to put in the same amount of power as he/she is doing. As far as I understand, the RoboPacers are actually increasing the power by 20% (?) on inclines, which I am fine with.

If I weigh 33% more than the RoboPacer, I will have to put in a lot more effort to get up the inclines. For most people I guess that makes sense…

As I said: Try it out. If you come back and tell me you couldn’t feel a difference, I don’t believe you tried. :wink:

You could say that that’s just reality, and it sure is. But Zwift is a game. And especially for RoboPacers - who should be able to pace me at the pace/power profile I am interested in - it makes sense to make settings which normalizes the power profile between heaver/lighter riders. For me, I’m perfectly ok just changing my weight from ride to ride, although it would obviously be better if I could change the “local” difficulty for RoboPacer weight.

Now, I could change the normal difficulty settings instead (and I may do that just for testing), but I have to admit that I don’t know the exact details on how that affects the ride. Changing my weight gives me a very accurate way of changing the power profile of a RoboPacer ride to within a few watts.

122

Of course there’s a difference. Same as in real life. In real life what would the solution be if you can’t keep up with a group, assuming you can’t shed 20kg overnight? :wink:

A) Find a different group ride at a lower pace; or
B) Get faster over time until you can hang with the stronger riders

I’ve been trying to suggest solutions to A) by pointing you to Miguel and B) by suggesting teleporting btwn RPs as a way of doing interval training.

:slightly_smiling_face:

123

David, you still don’t get it… The problem is that all the RPs are the same weight. It has nothing to do with the real world; it’s a game… If I wanted the real world I would find a group of 100 kg riders to ride with. Just using another w/kg does not help when the weight is way off. It’s basic physics and it’s ok you don’t want to test it out for yourself. The teleport solution is a non-solution; if I want a pacer I don’t want to TP to him/her all the time.

I could give you some examples but that would require some math, and for some reason I think you would just say “Lose some weight”. And honestly that’s just ignorant.

124

At no point have i said “lose some weight” or anything like it.

I’m out.

2 Likes
125

It is pretty much unsolvable. As the rider weight diverges from the pacer weight the efforts see-saw more. With 1.6W/kg pacer (on the flat) then on the flat I’m at 1.3-1.4W/kg, downhill freewheeling, and uphill (when the pacer goes to 1.9W.kg) I’m at 2.2+W/kg.

Speed would have the same problem - the ratio of hill speed to flat speed for a large rider is lower that for a smaller rider.

Nerfing the speed/power requirements also seems to be unattractive - these are segments you ride without the pacer and I for one would like them to be comparable.

I wish I had a magic solution I could give Zwift to make everyone happy but I don’t think it exists. Keeping it real is probably the best available option

1 Like
126

Exactly Poul. Except it is actually solvable, by either having a RoboPacer of your own weight or - even better - being able to tell Zwift before each start that you needed X watts average on this ride with a given RoboPacer (and then just letting Zwift change your “weight” while cycling with the pacer). But you’re right that the last option would mean that speed/power would be messed up in comparison to when you did it without the RP; this however is just like saying there should be no difficulty setting in Zwift (which there is).

Just to give some numbers:

Assuming I would go with David’s suggestion, and just ride with Miguel (since I can’t shed 25 kg overnight), here are the numbers I get from “The computational cyclist” (can’t insert URLs, google it)

Incline (%), Miguel (75 kg, 135 W) ,  Me (Watts, trying to keep up with Miguel)
0                 28.73 km/h                   145 W
2                 18.64 km/h                   167 W
5                 10.32 km/h                   175 W

Obviously they are not exactly equal to Zwift numbers, but just to give an idea about the problem. I am assuming that the RP is using a constant power, which is not correct either, but given the fact that RPs actually increases the power on inclines (as far as I understand), the problem just get worse.

So: A rider at 75 kg would simply follow Miguel, no problems. A 100 kg rider has to put in a lot more watts, (wow, we didn’t see that one coming!), in fact around 30% already at 5% incline. And even though that is (or should be) obvious, the problem here lies in the RP weight.

For me personally 145 W is just plain old boring right now (maybe when I use longer rides and need zone2 training). On climbs it would be ok, but on descents I would more or less do nothing.

Then I could use Maria instead. Let’s try to put in the numbers:

Incline (%), Maria (75 kg, 165 W),  Me (Watts, trying to keep up with Maria)
0                 31.09 km/h                   175 W
2                 21.39 km/h                   202 W
5                 12.41 km/h                   213 W

Again these are numbers without Maria increasing the power uphill. When completely flat, I can easily follow Maria and get a decent workout. But already at 2% inclines she get above my FTP. Not by a lot, and obviously this is nothing new (if I wanted constant power I could just use erg mode), but again: This is just to highlight there is a significant and highly non-linear impact on the power, when your mass deviates from the pacers.

Which is very easy to observe in-game if you just vary your own weight. It is understandable that people at around the 75 kg point does not notice this. But it makes a huge difference when riding with RPs and you’re heavy.

Currently the only way to mitigate this (imo) is to change your own weight to be close to the RPs. But it would be so nice if this could somehow be integrated in the game more seamlessly.

127

Except it is actually solvable,

You give two options - customisable robopacers, and nerfing. I ruled out nerfing and even you admit it would be messed up. However it is not like the (misnamed) difficulty setting as that does not change the amount of energy you need for a given course, it just changes the effective gearing of the bike. So I still rule out nerfing as a solution.

Your other solution is customisable robopacers. Let us have a look at that. Weights on zwift vary from 50kg to 130kg, a range of 80 kg. Power ranges from 1.0W/kg to maybe 4.0W/kg or more. If we made them fully customisable then using 0.1w/kg and 1kg steps you end up with 2400 Robopacers. OK - you can make that gaps bigger to get fewer pacers but you end up with a lot of robopacers. Even selecting them would be a pain. And most would be unused, and for most of the rest it would be a very small group. I, then, respectfully diagree - I don’t think personally customisable robopacers is a viable solution.

1 Like
128

  1. Have fewer pace partners to increase the peloton size In each group.

  2. Have a personal power adjustment bar (similar to a ftp bias in workout mode) that would allow riders of different capacities to ride together. This would also allow a static pace on hilly routes for people wanting a zone 2 ride.

  3. Add a zapper to the front of the pace partner group so you can keep the group together. Add a similar function/scooper at the back of the group that would pull people along for a set period of time if they fall too far behind.

129

On rolling Hilly routes like Titans Grove it’s easy to drop outside the drops game envolope despite dynamic pacing, maybe further increase it’s size on grades as dynamic kicks in.

2 Likes
130

Zwift knows your weight.
Zwift knows your height.
Add a toggle for “Match RoboPacer Height/Weight”.
Ride with the RoboPacer, and Zwift automatically changes your height and weight to that of the RoboPacer.
The changed height / weight will remain that of the RoboPacer until you are more than twice the distance of the Drops Multiplier bubble, to allow for the occasional drifts outside the bubble.

1 Like
131

RGT has them, but you specify the number of bots to ride with, and you ride as the only human (unless you invite others). They just start a new server job when someone requests a pacer ride. Presumably Zwift prefers to keep the social aspect of riding with groups of humans instead of bots as the main reason for not having customizable pacers.

132

I would like to see a “draft effect” metric beside the power/cadence/HR metrics while in the RP pack to better understand how the draft works and how to position yourself to modulate the amount of work you want to do. This metric would only appear while in the RP bubble. This would provide incentive for zwifters to ride with a RP and would be another point to focus on during longer rides.

133

I’m a super heavy rider (~140kg) so even the 1W/kg RoboPacer Taylor is problematic due to her simulated weight of 75kg(?).

As far as I know, all RoboPacers simulate a weight of 75kg. This leads for me to overshoot in descents and getting dropped on ascents - or at least very uneasy action to stay in the group in both situations. I assume riders with a significant lower body than 75kg may experience the issue the opposite way.

For me the use of RoboPacers would ideally be to get mentally into “the zone” and spin the legs for some time without having to think about custom workouts, route selection or pacing.

2 Likes
134

It was perfect without dynamic pacing and the pack dynamics overhaul.

1 Like
135

Nah. Pre-dynamic pacing, anything other than a flat course was completely unrealistic.

3 Likes
136

It was not unrealistic…many did/could just not hold the prescribed pace…

1 Like
137

Because those people had their trainer difficulty at some setting other than 0 ?

1 Like
138

Feature request, allow discarding a power up for 10xp. I don’t need that aero when I’m doing a Z2 Jacques ride.

2 Likes
139

i use robopacers all the time. the main draw for them, i think, is the predictability. you get a group ride at any time you want to start that is roughly the same experience each time. i tend to ride at a time (8:30pm EST weekdays) that isn’t super popular for group rides, so pacers often fill the gap.

i use the pacers to try to ride for longer at paces i am not proficient at. for example, i can ride with yumi for quite some time, but i often try to ride with jacques to see how long i can hang on. note that if you pick a bot at the edge of your ability like this, the dynamic pacing on hills is really nasty – i find i can stay with jacques a lot longer on flatter routes than hillier routes.

i also use maria for zone 2 riding.

it might be cool to have a “bot pack” that rides at slightly different speeds. sort of like the roboRacers mentioned above – they’d be pushing at different effort levels, but maybe come back together once per lap. you could try to stick with the faster ones as long as you can, and then fall back to the easier ones if you get tired or dropped. you could also work together to try to get one of the slower bots caught up to a faster one.

i liked the pacers more when they looked like humans. not sure why. it’s totally irrational. but there it is.

2 Likes
140

Sauce for zwift!

I don’t use it, and I personally don’t like the idea. But it has all these features, plus way more if you want. It’s paid, btw.