Power rankings are subjective fluff, fluctuating wildly week to week based on subjective overreaction. Power ratings, on the other hand, are math-based and considerably less impressed by one result or several results. They are also the source of the weekly betting lines.
As always, a simple reference is the Jeff Sagarin NFL ratings. His numbers are used by oddsmakers and are still freely available online. Other major power ratings sources eventually sensed their value and cloaked the numbers, requiring subscription.
Sagarin updates his public model following the Monday night game. That will provide an indication of how much the Dolphins "gained" from last week. Entering week 3, Sagarin had Miami 12th in the power ratings, 6.58 points behind #1 Buffalo:
Sagarin has notably been far less bullish on Buffalo than other respected sources. For example, TeamRankings.com is an excellent reference point on numerous sports. They are math based and use both terms...ratings and rankings.
TeamRankings.com has numerous ratings categories. The one they tout and prefer is the "Predictive Rating," because -- as they emphasize -- it incorporates a preseason rating as opposed to relying on recent results alone. That is the single greatest misconception and ignorance of the typical fan, that current results are all that matters. The reality is exactly the opposite. Applying preseason evaluation dramatically improves clarity in all sports. College football and basketball ratings are an amazing resource. Everything tends to drift back to the beginning. Regular season results can actually deflect from true ability level. Nate Silver always emphasizes that he uses preseason ratings to help fill out his NCAA bracket. I cannot imagine wagering on college sports without severe loyalty to preseason ratings.
The NFL has greater fluctuation year to year because the talent level is more balanced. A 3-point shift in power rating doesn't take you to a dramatically different level in college football. But it will in the NFL. You'll leap frog as many as 8 teams. But this is how TeamRankings.com explained the input of preseason numbers:
"Designed solely for predictive purposes, our current ratings improved upon our original ratings in several ways, most notably by incorporating a preseason rating (or "prior") for each team. Using preseason ratings improves the overall predictive accuracy of the system, although the impact of a team's preseason rating decreases over time."
NFL Predictive Rankings & Ratings from TeamRankings.com, your source for pro football computer power rankings.
www.teamrankings.com
The Dolphins are now up to 6th in that Predictive Rating. Buffalo is 1st and 4.5 points ahead of second place Kansas City. The Bills are 7.2 points ahead of Miami.
If you look closer at other ratings categories it explains why they aren't convinced by the Dolphins:
NFL Luck Rankings & Ratings from TeamRankings.com, your source for pro football computer power rankings.
www.teamrankings.com