Did 12 laps around the volcano today with D Miguel.
Noticed that each lap was a different time by as much as some 22 seconds.
Thought that the RPs were set at a basic pace… But 22 seconds over a short distance is quite a bit.
A robo pacer is just another rider like you. The details of where they ended up in or out of the draft, and the pack size, are going to determine the lap time. Their power is predictable and their speed is not.
This is data from a ride I did with a mate recently. First 4 laps Genie was essentially always on the front, I was able to sit in the draft at 2.7-2.9 w/kg avg. Then my mate rode a lap on the front at a pace that Genie was able to follow. 90 sec quicker and almost 20% more power needed to stay in the group just with 1 rider changing things.
Depends on how many people the robopacer is getting the draft from. With draft, the pacer goes faster, and everyone reacts and also goes faster, so the lap times are quicker and you need more power.
The above reply is right, sometimes you see some crazy shifts in lap times and power with Constance group for that reason.
This is the exact reason why the argument for robo partners on TT’s exist, alas, ZHQ doesn’t seem interested in a testing phase of this as easy as it would be to try for a week.
What it would mean is people would probably go to the *correct bot for them, as opposed to dragging a larger group around.
It would probably see lots of folks in the lower bot categories going up a bot (well, dunno about the D to C pace since it’s a larger gap than all others), which is where they likely belong anyway.
Downside of course is as pacer groups increase in size, the power required would still shift around dramatically; but in some instances, it’d likely be more stable than the current situation (or I could be wrong! But we’ll never know… until it’s finally tested.)
But this data has been shared before, for several years actually (and now again).
I have mentioned this so many times
When we tried turning off the draft for the pacers they got dropped endlessly. It does not work.
Never seen this statement before; however, easy fix… bump the paces.
Or… as stated; people just need to learn that they don’t need to be pulling the Coco group holding 3.9w/kg
At the end of the day; this is just another user basically once again asking for stable pacing bots.
(For OP though I’m kind of amazed a slow bot could be shifted around by over 20s for such a short lap, that’s rather impressive)
That was just “turning off the draft.” They need to be on a custom “Tron Trial” bike with custom cda numbers to get them to hit whatever speed they should be at. Or if that’s too much trouble, make them shorter. The outsized effect that rider height has on speed will allow you to do whatever you need to with that alone.
It’ll take a little testing but you can just run a draftees bot at Coco numbers, then lower the height 1cm at a time until it’s going 38 or 39 kph through the desert, then apply that height to all the bots. Might need more tweaking but that’ll get you 95% of the way there.
This sounds a lot like fixing something that isn’t even a problem.
There are a number of people who do consider it a problem. Mostly with the Constance group but it would be a pretty minor effort to spread it around to all the bots. I bet Eric could figure out your ideal bot-height in under an hour.
What do you think about Josh’s example? From one lap to the next, ~20% more power required due to the behavior of people (or one person) in the group. My understanding is that a lot of the complaints boil down to: a rider can’t predict how much power they will need to do to stay with the bot group even if the route is the same. Problem, or not a problem?
Whether there’s anything reasonable to be done about it a different matter. I have my doubts about no-draft pacers being a universal solution that will satisfy everyone. To achieve that you would have to have a lot more flavors of pacers.
I think there are two issues normally discussed. One is the difference between a bot depending on the behavior of the pack (if the bot is in the draft or not etc.), the other is the power swings required to stay with the bot across varied terrain (regardless of pack size) can be very big for folks who have a much different weight than the pacer.
For people who always ride with huge pacer groups there’s always someone on the front of the bot, so the first issue is primarily an issue with bots who have fewer riders. This is the scenario where bots on TT bikes would make a difference (with the caveats listed above).
For the second issue, as long as there are bots on flat routes the power swings are not too bad, but that really limits the routes they can be on, etc. This is the case where dynamic pacing affects different people differently (due to weight differences) even though it might feel great for someone who is sitting at 75kgs (same as pacer).
I have not ridden with a pacer on a TT bike, but I have done a lot of steady-paced group rides with a big pack where one person shows up on a TT bike, and you ride along with them for a while, they do some extra work but they hang with the group, and then there’s a descent, even a little one like going down into the underwater tunnel, and the TT bike rider flies out the back like they were shot from a cannon.
Ah I can see how descents would be an issue even if the pacer had some special bike to solve the power discrepancy issue on flats.
It could be said that so few ride with Constance and they are fast enough to live with it, or ride with a slower bot. The ride with a slower bot suggestion has been said by others before (not me).
Perhaps the bots need a maximum speed they reach in the draft - then they won’t exceed it. Along with maintain a steady pace message used for that bot that ran with the record rider (last year I think).
When that ran I remember the group being very steady, everyone stayed behind the bot.
A lot of people with the C grade bots do this because the pace is smoother at the front, less blocking. But they end up towing the robopacer around. Whenever the robopacer gets close they increase speed to prevent it overtaking.
This is what I was saying.
Slow bot, small lap and big difference. 
Another thing is to reduce the distance in front of the Pacer bot that it can pull draft from riders.
Also reducing the gap which riders can stay with the bot in front to get drops.
This would then prevent people pulling over the top on the front stretching the pack to a long thin line.
20 meters in front and your off the bot. 
Draft doesn’t work to pull the bot along.
I just think it’s a little nutty that you might be able to ride with any given pace partner at half a wkg below the stated level or you might need the full rated power (or even more due to group dynamics blocking and surging) and you have no way of knowing what you’re dropping in for each time. And it could also change back and forth every couple minutes.
Draft doesn’t extend that far. All it takes is 1-2 people sitting 3-5m ahead of the bot to pull the pace up substantially.
I have been doing our 100k Saturday ride on TT bikes now for a while; they are unusual in the sense that they almost act opposite from the group dynamics; it’s generally whenever I have to work the hardest, the group is at their easiest and vice versa, this is evident on larger rollers where folks push the pace dynamically anyway, but made worse from slinky action.
But on flats, it gives me an EXTREMELY CLEAR concept of any changes in pace, that might not be as evident on road bikes, but that’s due to the power difference anyways.
So where for example Tempus is obviously steady and I can tell when anyone wants to start pushing the pace more than it should be, the rolling roads of southern coast is… only as painful as I want it to be.
In the end, everything works out; I can still basically “do” flat power on either and I end up in the same place eventually… but most people like to chase to be part of the group no matter what.
Granted it’s worth remembering I’m lighter, so I don’t get flung off the front on descents; and again at my weight, I actually find if I keep my steady pace I descend equally as fast as the group on things like the descent into jungle from windmills.
All that to say, other than the pack dynamics being a little backwards feeling at times, the steadier I am on a TT, the less of an issue it actually ever is.
And this isn’t me just being slightly anecdotal. I’ve been doing this ride on a TT bike or a gravel bike now almost every weekend since the end of winter.
All of that said, I still don’t see why this needs to be more complicated than it needs to be.
So what if you’re used to riding with whatever bot…
If this is literally coming down to “people don’t like change” then this thread will get buried like every previous one has.
If James says bots were tested on Tri bikes, I’d still like to know when that happened, because it certainly wasn’t any recent version of Pack Dynamics, and I still can’t find any evidence that it did happen. Only thing I run into is Michael’s thread about some bots being on the Espada when it was released… until folks said that it was actually just a model of the Espada but was receiving draft…
Meanwhile, I will note that this group ride I’m doing on the tri bike the pace is generally upper 2w/kg, 2.8-3. I’ve been averaging 3.3 range no issue, which is also what it takes on a road bike sitting at the front of the group anyways… so as I’ve repeated, nothing is really any different, except.. the tri bike I have to HOLD that power the whole time, no breaks.
But where things get WEIRD, is … I can’t even get close to keeping up with the Coco group, but the jump from Maria to Coco is too huge, but even Maria group is weird unlike our group rides; which I haven’t figured out why.