Yeah, sounds like any wheel overlap with the rider in front should remove the draft. (Of course if you are in the 2.5th row or further back, you should still get draft from riders further ahead.) I suppose it would also make it easier to maintain a single paceline without overlaps in TTT-type situations.
This is how some other platforms manage it (VirtuPro). Zwift chose to go the speed rather than resistance route, which is understandable. The resistance method would be great for realism but a nightmare to support for anything other than top-end trainers.
Now you’ve said it perhaps the idea will get taken seriously. I accept it may be too small an effect to solve the problem, but surely it will help a bit if the second row of riders in a pack are bleeding off a bit more speed before they really hit the front.
The discussion is handicapped by not knowing precisely how the draft cone is positioned and used - is is all relative to centre of avatar or front of front wheel etc? If two riders are side by side I assume there is zero mutual draft, but how far back must one drop before they get a benefit?
(Nb for the pedants expecting high precision in the simulation, in real life the front rider derives benefit from riders behind them, albeit a very small one. So long as the speed is subsonic 
)
I currently have a 1 in 7565 success rate, so maybe.
I’m trying to work it out now using Sauce, as the draft value is real time (hundreds of times per second). Hard to get a good screenshot remoting in to my Zwift PC though.
Otherwise I’ll see if I can find that draft GUI graphic somewhere.
Yes, the idea has already been discussed internally. But as you can imagine this is a whole new can of worms to open…
Show off.
2 Likes
I would have to confirm, but I believe it’s around the center of the avatar 
1 Like
Middle of front wheel by the looks, if the graphic is correct:
I can’t imagine this would make a huge difference, but something is better than nothing. Certainly worth testing, then can determine if a small additional slowing is needed.
Deleted - drafting image already posted.
I wonder how the draft cone from multiple riders is combined. I’m not all that clued up on fluid dynamics, but I imagine simulating draft from multiple vehicles would not be simple summation, possibly some form of superposition as in classical waves which would at least be linear equations.
I rest my case
1 Like
Thanks for finding that. It doesn’t quite answer the question where the following rider feels the draft from, ie is it also their front wheel that defines their position for the purposes of draft benefit? But it does at least suggest some fine-tuning could be possible by moving the cone back (possibly also narrowing). i’d suggest there could be zero draft as soon as overlap is achieved, which is a bit too strict compared to real life but perhaps better to err in that direction than allow riders to surge through as currently.
(it’s also skewed relative to wheels, the max draft appears to be on the RHS of the bike, which suggests limited precision in the graphics.)
If shifting the draft cone means that the rider loses ~100W of draft benefit just 1s sooner when surging through (will depend on relative speeds, just a ballpark number) then that’s an additional 100J of energy loss for every overtake, compared to the status quo. That’s small beer compared to the ~1kJ lost in a bad sticky draft bump, but that was horrid and really spoilt the riding experience.
Of course it’s also possible that the draft graphic is drawn from the middle of the front wheel, and the actual draft starts from somewhere else completely.
In a TTT or in smaller groups (e.g. two riders), will both the first rider and the second rider end up having zero draft when the tip of the draft cone is moved backward? Or will the tip be moved just so the second rider gets some draft?
1st Rider behind the leading rider should have draft…IMO if the draft cone would be moved 1 meter back it would make a huge difference. We might avoid riders going left and right in a TTT and hold more steady line.
Check this video i did a test with PD4v1…you have draft numbers with Sauce. HD version in 1hour.
3 Likes
@DavidP James you mean like this? So moving out loses draft sooner and hopefully slow pack more. Currently when bumped out draft may still be present causing speed increase and churn?
5 Likes
Yes, just like that. Also, great video test from @DejanPresen, the draft % stayed fairly constant around 85% even when the reported distance varied by 3 or 4m.
Just looking for images of draft behind a cyclist and found the study by Blocken about drat behind moto’s. Interestingly the draft behind moto and cyclist is quite narrow and zwifts from broadcast image is more of a cone. Could this be part of the issue. More riders fit in the draft of a single rider?
1 Like
The problem with that image is that the draft only moved for the first Rider. I think the draft will move for all riders.
Like this:
6 Likes