It’s not a bug – it’s an undocumented feature.
Hi, I was analyzing my data and noticed, that there is something wrong with the 5 and 15 seconds power values, as they appear to be quite the same as the values for 30 seconds power... I compared the data with strava and zwift itself and the values there for 5s, 15s and 30s are totally different (strava: 30s - 748W, 15s - 871W , 5s - 950W, link: https://www.strava.com/activities/1938258060), than on ZP (for the very same activity it is: 30s - 716W, 15s - 757, 5s - 759W, link: racereport.php?id=54961639600111872). For the 5s power , there is almost 200W difference, which is huge. I have discussed this on the Zwift.pl facebook group (which is basically the 'Team Poland' facebook group) and there are several guys experiencing the same issue (the power curve getting flat from ~30 seconds). This seems to be some kind of new problem, because (not my statement, as I'm new to zwift and zwiftpower) one year ago it had been all right. Could you have a look?
I had this also today where if you look at the power profile curve it gets chopped off for power under 30 seconds. Happened with all other racers though it seems like the steeper the curve the more power/time gets cut off.
pvdbg wrote:
Sat Jan 05, 2019 7:13 pm
I had this also today where if you look at the power profile curve it gets chopped off for power under 30 seconds. Happened with all other racers though it seems like the steeper the curve the more power/time gets cut off.
If you look at the blue bar at top it indicates .fit files are still being processed. Hopefully that will be more accurate.
Processing taking a long time today with the Tour on.
Ok thanks for pointing that out. I never noticed it before.
I noticed that the short power numbers reported are off quite a bit even after the uploading is complete. I have compared the zwift files in TrainerRoad and Golden Cheetah so maybe there is problem between Zwift and Zwiftpower?

I for some reason can't attach images but in my last 2 races the 15s power is low by about 140w on Zwiftpower, 1m is about 50w lower, 5min is about 10w lower, and 20min is within a couple so its not all off. Just seems odd that the shorter the measurement the more inaccurate?

I also looked at the race I did before these two and all the numbers match. But also looking at races a couple months ago there are also discrepancies like above.

Just want to let you know that something odd is happening.
ZwiftPower ignores the first 5 seconds of your fit file as it often contains power spikes (waking a power meter up from sleep). I can see in at least one of your races you hit over 1000 watts at the start.

Tip: You only need around 450-500 watts gradually easing off to stay with even the fastest riders at start. A full blown sprint is a waste of energy.
Ok thanks for confirming. It might be new but Zwift actually doesn't start their power recording right away either. It begins about 5 seconds in once you pass the first banner, thats when the clock starts.

I do the sprint for training. Deplete W' from the beginning and try to recover throughout the race to sprint again at the end. More racing to train than to win.