This seems like the right place for this. Back in October the AP put out this study.  They claimed that it showed no SEC bias in their poll. I found this on reddit where a user did a good job of showing that their study was shit.