This specific test was a power meter output as I wanted Zwift to respond to an exact power reading, the idea of the test was to isolate the software response as some had speculated that the Zwift physics model was at fault.
I thought the advantage of microbursting was down to the idea that surge/rest/surge/rest is overall less taxing on the individual than steady-state power. Ie. Jamie’s post #272 in this thread with decreasing HR using the MB technique. In turn this would provide more headroom to increase the average power using MB above what can be done at steady output. Put simply MB yields more avg power for given HR?
That pedaling technique would be legal in the Tour de France. Why do you think it’s not seen on the Tourmalet? Are they ignorant of the benefits?
Yeah dude. Zwifters are the real innovators. World tour will catch up soon. That’s why there’s more riders with world tour power in ZGP than there is in the world tour itself. Those boys and girls will catch up zoon enough.
i saw twitter cancelled zwift racer leandro messineo doing it in the breakaway in the vuelta san juan on eurosport earlier
Can someone explain how this erratic pedaling called micro bursting can actually yield an unfair/unrealistiv speed advantage? I understand the concept of sticky watts and have seen this in some data, but I cannot find the explanation for micro bursting.
I am very certain though that it is not a alternative pedaling style that would give you advantages IRL - quite the opposite is the case…
Random video in my YT suggested videos (from 3y ago), in which the host advocates for microbursting to avoid getting dropped @ 2:22.
hilarious
Burst Mode!!
Had a guy last night micro sprinting all race, orange figures then 0 watt then orange again, called him out on it and was met by silence
Check the comparison from 10:45.
Now assuming the 10hz frequency means you accelerate quicker but also decelerate quicker, logically the avatars should finish at identical spots.
Is this evidence of the benefit of Microbursting regardless of what kit you have? Being able to spike power to a higher value (as happens with the 10hz sample) generates more distance covered. It’s like equal and opposite reactions are not a thing in Zwift.
I can’t be certain that’s the issue here, but it looks that way.
The ~1m distance represents an extremely small amount of energy over the whole on/off cycle, whether it’s a zwift calculation inaccuracy or a trainer inaccuracy isn’t clear from the data presented but in any case surely it’s nothing like adequate to explain microbursting where people seem to be benefiting to the tune of several tens of watts, possibly over 100.
Looks like it. This is related to prior allusions that it’s something to do with how software handles inertia, acceleration and decelerations.
it does line up with what i have been trying to explain about how the game mechanics work for 3 years but i want to know if the rider on 10hz is actually doing 1000w+ like the 1hz rider and the sampling is chopping his peak or not
Indeed, I’m thinking this a contributing factor in multi-faceted issue though.
Without knowing how precisely the two versions of trainer data match up (or not) there’s really not a lot to go on here. The difference looks like it’s basically within the range of a rounding error. Clearly it’s an advantage to have a more responsive system in some race situations (which is what most of the video seems to highlight), but that’s a different issue.
I’ve called out riders on that too.
This should be really easy to test with a bot shouldn’t it? Same as how they test different bikes to see which is faster.
Hey one bot to do microburst technique the othe to do steady and see which is faster
That would surely test whether it’s a zwift issue or a trainer issue.
On a bit of a tangent :-
I think far too many people think Zwift “fitfiles” are canonical.
They are 1Hz sampled and most likely interpolated from what Zwift receives.
We know zwift clients can certainly get power data in the 4-10Hz range now - but we also know that fitfiles don’t show all that data.
Not to mention what happens when data lags across networks / t’internet - do the fitfiles use whatever Zwift interpolated for that period for the rider, or does the fitfile report samples from the data it eventually receives?
It makes assessing “micro bursting” a tricky situation with so many hardware / connection / reporting differences and unknowns.