Watts in W' export go negative - how is that possible?

[Steve Mansfield]

Per my other postings about recreating W', I did a ride on Saturday with the purpose of exhausting W'. Well I did that, however when I exported the W' data to see more detail, I see that sometimes the watts, which is 25s SMA go negative. Weird.

Also weird is the values for  25s SMA the exported file has are not the same as the ones my spreadsheet calculates. Is there a special way the SMA is calculated?

A snippet of the first 25 seconds of my ride is shown in the screenshot.

Cheers

Steve

[Ale Martinez]

El lunes, 15 de diciembre de 2025 a la(s) 2:48:57 a.m. UTC-3, quit...@gmail.com escribió:

Per my other postings about recreating W', I did a ride on Saturday with the purpose of exhausting W'.

The best way to preserve context is to use the same thread, to open multiple ones for the same theme is not a good forum practice.

 

Well I did that, however when I exported the W' data to see more detail, I see that sometimes the watts, which is 25s SMA go negative. Weird.

Also weird is the values for  25s SMA the exported file has are not the same as the ones my spreadsheet calculates. Is there a special way the SMA is calculated?

There is a data conditioning step first:

https://github.com/GoldenCheetah/GoldenCheetah/blob/a31a2795ac469a450fafe1d1cef0347588b08985/src/Metrics/WPrime.cpp#L172-L176

Then a moving average is computed:

 https://github.com/GoldenCheetah/GoldenCheetah/blob/a31a2795ac469a450fafe1d1cef0347588b08985/src/Metrics/WPrime.cpp#L298-L321

 

A snippet of the first 25 seconds of my ride is shown in the screenshot.

That is not very useful, if you want others to analyze your specific case you should be willing to attach the relevant files.