ZwiftPower Critical Power Issue

Hey everyone,

We wanted to provide you with a quick update about ZwiftPower.

After we solved the issue with the power metrics and .fit file processing queues, it’s come to our attention that some users are seeing matching power values for 15s, 30s, 1m, and 5m power intervals. It can also be that the 20m power is lower than the overall average power for an event.

Here is an example:

If you’re seeing this, please provide us with a link to your ZwiftPower profile or a specific event.

We currently are investigating this issue, and while we aren’t sure of the cause at the moment, we will do our best to keep you updated as we make progress.

Thanks for your understanding and patience.

ZP - 834206 (won’t let me post link)
Hi, last race data looks fine but all 6 previous had higher ave than 20m max. The 20m max and 20m ave from the zwift and garmin fit files were both higher than the zp recorded 20m.

Also 2 races unprocessed despite the backlog now cleared so does that mean they won’t be processed?

Thanks :+1:

The other thread really should not have been locked; there are still delays in fit file processing. Do we just create a “Zwiftpower .fit file processing stuck?” v2?

1 Like

I think you were a little too fast in closing the other thread, so now we will have to spam this one instead with the remaining problems…

The fit-file queue seems to have emptied, but I still have a handfull of races with only live data. I have checked, that they are all public on Zwift. They are all either ZRL-races or WTRL-TTT’s.

I took the most recent one (last thursdays TTT) and tried using the “Request update” link on the event. That finished in less than 15 minutes, but I still have a green lightning bolt on that one.

So it seems there are still fit-files, that aren’t beeing pulled into the system correctly. One event that has this problem is ID 1459532 - that’s the one where I tried the request-update…

2 Likes

Hi @xflintx thanks for the much improved communication and working to fix these issues.

I haven’t got any examples of repeated power data for the time intervals (thankfully) but have one race where 20min power is lower than overall. I have read people trying to explain how this is possible previously (and do see it on power curves on Strava etc. too sometimes) I think mathematically it is possible on a shorter race, although a minimal issue would be good to get a definitive resolution.

I do also have one race that the .fit file still seems to be stuck for as per some of the other guys above. The NYE WTRL TTT is stuck on live data for some reason, looking at the race results it seems about 75% of people in that event still only have live data green bolts.

Thanks Again!

Sam

Hi,

I have this issue here.

Event id: 1375248
ZHQ FutureWorks Crit City Race - Anti-sandbagging
21 December 2020 17:30

Regards Dan

From a few exchanges with Sticky, it is not uncommon for the fit files for TTT events to get stuck and get not processed correctly, leaving you with only the green arrow. I have a few instances of those over time, myself as well. I think it comes down to the fact that we need waiting in the PEN. Not very scientifically, but I could somewhat correlate issues with my files to when I moved a bit in the PEN (to get closer to the front). The last few TTT I have not done that and have got blue arrow results… For ZRL scratch races, that does not apply so I am not sure what is happening there.

Here my 20 min power is lower than my average for the race (James Eastwood, cat B)

Mathematically that is possible of course, but it does seem a little odd.

Hi, not sure if it’s related to the issues with the Fit files or metrics, but I noticed that no KOM segment for VenTop was calculated for the following event:
…events.php?zid=1424580
(Can’t post actual link)
I was meaning to post this in the .Fit file processing thread but that was closed.

If you look at your power for the race it’s shows higher power at the start and end.

20 minute power is probably the last 20 minutes which includes the middle section and end of race.

Full race average also includes the start so will be higher.

From your profile on ZP click on the last icon on the row for the race and it’ll show you the power/power curve for the race and you can see it going up after 20 minutes. forum won’t let me post a link but the url so end up finishing with

racereport.php?id=3189883963136402734

1 Like

Hi Chris, I have the same issue as @gloscherrybomb for that Veganuary race on 2nd Jan (though I was in A and he was in B). I have a higher overall AP than 20 mins AP. My hardest 20 mins was definitely the start of that race. There were loads of A+ men and I was at threshold for 8 minutes before the elastic snapped and I was dropped and settled in with the detached group.

I had my pedals connected to both my garmin and zwift. The garmin uploaded the file to training peaks so I have the power peaks from there. (My second garmin which was meant to dual record the trainer switched off so unfortunately I don’t have the dual)

Comparing peak powers on TP to ZP (20min / 5 min / 1 min / 30s)
Pedals to garmin (TP) 287W / 321W / 352W / 396W
Pedals to Zwift - ZP numbers 282W / 308W / 327W / 347W

Overall AP is 285W on both ZP and my garmin recording.

Not sure what to suggest beyond maybe uploading the Garmin file to the ZP analysis page and see if it shows anything.

Event 1409816 (Zwift Fondo Series 2020/21, Zwift Gran Fondo, 2021-01-03 09:00 UTC) seems to be stuck on green-bolts for most riders. What’s the best way to request the analysis be re-run?

There’s definitely an issue now I have looked at the numbers of the same file in intervals.icu. Today Tofu Tornado has my 15 secs at 6.1w/kg. intervals.icu has it as 9.3 w/kg.

I understand how this can happen mathematically, but it’s not the case here (and is rarely the case in Zwift due to the fast starts, the first 20 mins is often the fastest). Double checked this one in intervals.icu too and its an error on Zwift Power’s side.

1 Like

Agreed. Yes it is possible but that’s not what has happened in my races - the data from both zwift or garmin fit files do not match the zp 20 min max numbers. ZP is repeatedly wrong for me.

Event
events.php?zid=1465026

My Profile
profile.php?z=999852

It is not letting me post links.

This has corrected itself now. I’ve requested an update to the older events to see if they get resolved too.

This has now happened to me as well. avg is 3.9w/kg and 20m is 3.8w/kg. Just looking at others in the same event and it actually looks like it happened to several of them…

ZP Event https://zwiftpower.com/events.php?zid=1477209
My profile https://zwiftpower.com/profile.php?z=2627867