1223777 - [e10s][telemetry] regression in FX_NEW_WINDOW_MS

Status: RESOLVED INVALID

(Firefox :: General, defect)

Description

[Jim Mathies [:jimm]]

Attachment: index.png

https://github.com/vitillo/e10s_analyses/blob/master/aurora/e10s_experiment.ipynb

Comment 1

[Roberto Agostino Vitillo (:rvitillo)]

(In reply to Jim Mathies [:jimm] from comment #0)
> Created attachment 8686010 [details]
> index.png
> 
> https://github.com/vitillo/e10s_analyses/blob/master/aurora/e10s_experiment.
> ipynb

Note that the effect is not statistically significant.

Comment 2

[Jim Mathies [:jimm]]

(In reply to Roberto Agostino Vitillo (:rvitillo) from comment #1)
> (In reply to Jim Mathies [:jimm] from comment #0)
> > Created attachment 8686010 [details]
> > index.png
> > 
> > https://github.com/vitillo/e10s_analyses/blob/master/aurora/e10s_experiment.
> > ipynb
> 
> Note that the effect is not statistically significant.

Thanks for pointing that out.

The probability that the distributions for payload/histograms/FX_NEW_WINDOW_MS (parent) are differing by chance is 0.65.

Closing this out for now, if and when we have a confirmed regression we can file again.