Four games into an NFL season is a fairly significant sample size to start drawing conclusions. It used to be the exact quarter mark of the regular season, but now we have a 17th game.
Either way, I think we're starting to see some trends that we can rely on as real, and not dismiss as small-sample-size outliers or anomalies. Let's take a look at some teams in the Power Rankings I think are out of place: