I wrote the following on my BGG profile today:
As of Oct 26th I am returning to rating games but with a relatively strict adherence to a sort-of-flipped Poisson Distribution with a lambda of 4. This distribution is also appropriate for modelling the number of soldiers killed by horse-kicks each year in each corps in the Prussian cavalry.
proportion of games that get(rating)=(lambda^(11-rating)*e^-lambda)/(11-rating)!
This is more positively skewed than a regular distribution because I am a positive person.
I reserve the right to change my lambda as I see fit. You know, as I get older and crankier.
I also will not rate a game unless I’ve played it 7 times. I also choose to change that number (probably up) as I see fit.
You can see those ratings here.