Fair point. But I still wasn’t claiming it as dispositive proof.
Does this ‘myth’ also mean you disagree with DC rainmakers review of the elite direto in that it holds power longer than it should without pedalling?
1 Like
I sure he got the correct results for the model he had that day. But I know I can supertuck faster than virtually anyone I’ve raced against too. I looked at your suggestion that I tried a Neo, but his review of that said it’s accuracy was very questionable, or words to that effect.
I don’t think his review said quite that either, in his reviews of intervals he said it matches almost exactly. However, he said when spinning to zero it holds a bit more power for slightly longer than others. Perhaps it was the earlier firmware that someone implied was to blame? Either way, it doesn’t tally with my experience.
But the myth I’m referring to is not really relevant to that. From the case here and general discussion ‘microbursting’ seems to be the idea that frequently varying power between high and low (as opposed to zero) provides some additional free speed that one wouldn’t obtain at the same average wattage. Rainmakers review covers this and rates the direto as excellent, matching almost perfectly his other sources. I think maybe the Neo performed poorly here, so it’s possible there’s something with that, but I personally doubt it’s significant enough for anyone to have noticed let alone be able to exploit for an advantage.
Microbursting from my understanding is high and zero pedalling. I suspect the confusion with high and low is that most people that do it. kick in and pedal again before the Zwift power hits zero.
2 Likes
There does seem to be some disagreement on what it actually is. My interpretation was my genuine understanding of the term, given that the example provided doesn’t show zero power, I don’t think.
Doesn’t show zero but I believe those using the technique are stopping pedalling which if accurate should zero straight but it’s not which is the issue.
2 Likes
Regarding the original post (and the Castle Climb where the rider gets beaten by their opponent).
The more I look at that rider the more I’m convinced they are not microbursting but using sticky watts.
Here is their power in the Castle Climb between 19.5 and 20km where their opponent rides away from them. You can see the characteristic flat tops of sticky watts.
Here is a 2km segment from the race where the rider set their best ever 20 minute power (a TT on Tempus Fugit so no hills or drafting involved). For the 18km race the last 5km is almost all perfectly executed sticky watts.
I think this guy is using sticky watts rather than microbursting.
BTW, if anyone is interested I know of a video where sticky watts is demonstrated (not in a race situation because that would be cheating). You can see the actual rider and compare what happens in real life versus what appears on Zwift. It might help those who are still confused about what sticky watts looks like. I could post the link if someone from Zwift would guarantee that the rider won’t be “maintenanced”.
5 Likes
My final post on this (now that it’s back on track) is that sticky watts, microbursts, trainer ramp down differences, whatever… this behaviour is having a significant impact on the enjoyment of others and the general validity of the platform for racing. Its not hard to detect such a pedalling technique, and it’s therefore not difficult to penalise it within the model. Make it ineffective, and everyone stops doing it.
9 Likes
I’m thinking about this.
How do they make it ineffective. My thought is the avatar need to be able to hold momentum longer then this microburst will just be a waist of time.
There’s probably other options.
What do you guys think, how do we make this ineffective.
Pointing out issues is easy, fixing them is hard. Let’s put our heads together and think of some solutions.
1 Like
Yeah, that looks fairly conclusive. He’s gaining an unfair advantage by doing whatever he’s doing. Looks to meet the definition of sticky-watting. But whatever you call it, it’s wrong. Apologies for any and all the statements I’ve made to the contrary.
If it’s possible to tell what kind of power reading he’s using this would be extremely useful to know if it just limited to certain pedals or is it possible to do with a smart trainer. My understanding until now is that it’s limited to pedals.
If I were trying to define it, in order to identify and take action as a starter I’d suggest:
Within a 30 second period if the detected power:
Hits twice or more the users FTP on 3 separate occasions or more (or 500W if this is more available)
Hits below one third of FTP (or 100W) each time between those peaks
Has a period of low variation (+/- 5%) for 2 seconds or more during each peak OR has a period of identical power for 1 second
Possibly excluding descents where someone could legitimately be repeatedly trying to hit supertuck speeds.
1 Like
I think microbursts and sticky watts are opposite sides of the same issue.
The problem seem to be the accuracy of the power measuring device when the power is changing.
Microbursts seem to be an exaggerated reading as the power is increased and sticky watts seem to be a delay in reporting the power as the power is decreasing.
If I’m pedaling by applying steady pressure on the pedals, my power meter is very accurate.
If I start to pedal very hard, initially, there is a brief delay but then the power jumps up, freq overshooting my actual power but if I maintain the new increased pedal pressure, the power quickly stabilizes at a new level.
This would be the microburst.
As I ride along, if I pause briefly, my power will hang up at the higher level briefly before dropping, that’s the sticky watt portion.
We need definitions.
I’m proposing that microbursts refer to an over exaggeration while the power is increasing and sticky watts are a similar phenomenon while the power is decreasing.
Does this seem correct? Is this why sometimes it looks like micobursts but then someone says “no, sticky watts”.
1 Like
close enough. anything directly related to hardware (SRMs i know for sure allow it, i forget what else does) is some form of sticky watts, anything that doesn’t involve hardware is microsprinting and is probably (my gut feeling) not related to what power meter or trainer you use at all (a trainer that overshoots in peak power efforts would certainly help though, plenty of brands do including newer wahoo kickrs, early cycleops hammer generations, my old direto x used to too, etc) but is taking advantage of game physics or server latency, or whatever it is that actually allows for it, in a way that isn’t intended.
zwift believe the second thing isn’t actually an issue (myself, other racers and race organisers disagree) and truthfully it’s hard to prove independently with any kind of scientific rigour because testing it would take time, effort, at least an intuitive if not a technical understanding of what it is you’re actually looking for, and an expensive range of hardware. but people do it and they do it often
It depends who you speak to at ZHQ, in the previous mega thread (struggling to find it) they suggested they were aware of it but was deemed low priority to fix. Both WTRL and ZRL have DQ for microbursting so i don’t think it’s quite clear cut to say zwift believe it not to be an issue.
The main thing this thread shows is nobody really knows exactly and it’s very hard to quantify any advantage.
Sticky watts it’s pretty clear so easy for organisers to apply DQs but still not ideal having to notice then retrospectively apply DQs.
Zwift server end should be able to detect an excess number of flat top power sections in a ride and warn/flag the rider in the same way as they have systems to auto flag you for superhuman power efforts.
As for microbursting i wonder if they could do similar as was suggested further up. If we agree that it’s a stupid pedalling style they could flag a rider that repeats bursting an unreasonable amount as suggested when you go x times over ftp within time period.
You won’t find it as it has been removed.
1 Like
I did wonder if that were the case. Maybe we’ll get silenced here as well. Blog post instead? 
Don’t you just hate it when somebody wants to silence you??? 
1 Like
Can zwift provide to there customers : “race even” without che**ing ?
This is very frustrating
1 Like
WTRL may DQ when it’s brought to their attention but they have previously claimed to have automatic detection which is obviously untrue. There are riders sticky-watting week after week with no sanction.
I don’t think it is easy, or even possible, to detect algorithmically when a trainer or power meter is giving free watts. You’d have to know how hard the rider was pedalling, ie a second reliable measuring device. You could possibly detect large changes in power and declare them abnormal, but that would not catch all cases (unless the threshold was set low enough to also accuse innocent riders).
4 Likes
For sticky watts the cadence has to be zero; that should be pretty easy for ZHQ to detect. They could simply change things such that power is automatically throttled instead of being held constant when the cadence abruptly drops to zero. That would stop people abusing it.
1 Like