I'd be interested to gather a list of characteristics of ptij Qs that may affect votes (e.g. post length, # of wordplays/sourceplays, relevance to Purim/Esther/Megillah, whatever else folks think of), see if we can find a subset of them that do in fact
significantly influence scores and how, and then see how each fared in various years, so we know what about Qs has led to a decline in score. @IsaacMoses