2003 EGD rating distribution

EGD Rating distribution

Year:  

1996 | 1997 | 1998 | 1999 | 2000 | 2001 | 2002 | 2003 | 2004 | 2005 | 2006 | 2007 | 2008 | 2009 | 2010 | 2011 | 2012 | 2013 | 2014 | 2015 | 2016 | 2017 |

Revised rating distribution

Year:  

1996 | 1997 | 1998 | 1999 | 2000 | 2001 | 2002 | 2003 | 2004 | 2005 | 2006 | 2007 | 2008 | 2009 | 2010 | 2011 | 2012 | 2013 | 2014 | 2015 | 2016 | 2017 |

The horizontal axis shows the go rating scale used by the EGD and the revised rating system.
The vertical axis shows the number of tournament games played by players of a declared rank in the selected year.

The colored lines from yellow to blue are rating distributions, colored per declared grade.

When looking at the EGD distributions, it seems that over the years, many players graded around 1k (between 6k and 5d) have grown weaker than their rank.
In recent years, this trend seems to be reversed a bit, but this could be due to players finally complying to the rating system and demoting themselves.
But I suspect this is an artifact of the parameter values of the EGD system.

When choosing different parameter values (based on observations), the picture changes.
I find that the tweaking the rating reset policy works well to regulate rating drift (inflation / deflation).
In fact it seems that tweaking the reset policy works so well to regulate drift, that I was able to discard the epsilon parameter that was intended to regulate rating drift.
I ended up with a very liberal reset policy. Basically the system fully believes a self promotion (once). But for a period of 20 games, the promoted player's rating will be more sensitive (higher K) and their opponents ratings will be less sensitive (lower K).