Following the Six Nations 2017, I read this interesting webpage explaining how World Rugby's rating was calculated (also thanks a lot to L'Equipe for pointing out the webpage). I was then wondering how much the ranking was sensitive to the method of computation, and in particular, I was eager to test the ELO method applied to chess for instance. I am grateful to this Github repo which provides with the Web database for rugby (and more) matches: this avoided me to write text miners in order to extract the results from the World Rugby webpage.

Recomputing the ranking since 2003, and trying alternative methods of ranking:

Ranking as on Feb, 19, 2017.