I saw KC Joyner's chat touched on the issue raised in this thread so thought I'd pass it along. Basically, he says the bad decision metric is just one measure of QB performance. I don't put this in to start a debate about whether Brady is or is not one of the best QBs in the league but just to shed some light on this topic.
Steve (Richmond): KC, If the bad decision metric doesn't either correlate to overall QB performance (as you've stated above), or to a team's performance (SB QB's), then of what use is it as a statistic?
KC Joyner: (3:55 PM ET ) The best way to answer your question, Steve, is to say that the bad decision metric is simply one way of measuring a QB's performance. It was never intended as an overall barometer of QB performance. The closest thing I have to a single metric for QBs is what I call BADSYPA. It combines yards per attempt, success percentage and bad decision percentage. Ben ranked 3rd in that category, so his overall numbers were great (as I said in the column). He still needs to improve in this one area, though.