Discussion about this post

User's avatar
Doge of NM's avatar

Get the point, but the linear fits look weak. Perhaps model with a robust linear fit such as the Theil–Sen estimator and show the confidence intervals. These linear fits look like least squares which are skewed by the outliers in the data, which I imagine makes it seem like there is 'more progress' then there really is.

Expand full comment
Dmitri's avatar

Did you know that there is no NEGATIVE PROGRESS. Next time just write PROGRESS , that is what intelligent people would do.

Expand full comment
1 more comment...

No posts