I'm still going with the Natural Power Rankings, because if last week is any indication, I'd do better predicting games using this than my own intuition.  I would have called the Browns win over the Ravens, the Giants' win over the Eagles, Buffalo's win over the Jets, and Oakland's win over Miami, all of which I got wrong in my week 3 predictions.  Of course, I wouldn't have gotten Tampa Bay or Arizona's wins correct this way. 

"What are Natural Power Rankings?" you might ask if you didn't read my blog last week or the week before.  They rank the teams 1-32 based purely on who beat whom, and by how much.  You know how you can have huge, long debate about whether the Giants or the Lions is better? 

"The Lions have a better record."

 "Yeah, but look who they've played." 

"They just beat the defending NFC champions." 

"The Bears suck right now." 

"The Bears have played tough teams." 

"So did the Giants-besides, San Diego isn't that tough right now." 

"They've played tough teams too." 

"They just lost to Kansas City!"

 "Kansas City has only lost to Chicago and Houston, who are both good." 

"We're going in circles.  Look, it's quite simple.  The Giants beat Philadelphia.  Philadelphia beat Detroit.  The Giants are better than Detroit." 

At this point the argument has reached a logical conclusion, though it would most likely continue with irrational statements from the Detroit fan.  So basically what I've done is take such arguments for every team, thus ranking them according to the following arguments:

1. If A beat B, A is higher than B.

2. If A beat C by more points than B beat C, A is higher than B.

3. If A beat C by the same margin as B beat D, and C is ranked higher than D, A is ranked higher than B.

4. If rank still cannot be determined, margin of victory is weighed against quality of opponent (by Natural Power Ranking), and whichever is more significant determines the higher ranked team. 

An attempt to do Natural Power Rankings after week 5 gets particularly interesting with Arizona's win over Pittsburgh, Kansas City's win over San Diego, and Atlanta's win over Houston.  The first creates a huge series of circles involving the NFC West and AFC North.  The second and third create more of a web. 

But here are the circles:

1) Seattle>Cincinnati>Baltimore>Arizona>Seattle

2) Seattle>San Francisco>Arizona>Seattle

3) Cleveland>Baltimore>Arizona>Pittsburgh>Cleveland

4) Pittsburgh>San Francisco>Arizona>Pittsburgh

5) Cincinnati>Baltimore>Arizona>Pittsburgh>Cleveland>Cincinnati

Now no matter what happens, these seven teams will be consecutive in the Natural Power Rankings.  But who goes above whom?  Well, one thing's certain-it'll go Cleveland-Cincinnati-Baltimore, since Cleveland has beaten both of those teams.  But we're still stuck trying to figure out if Seattle and Arizona belong above or below those two.  So I'll have to go by score differences in each game involving teams in the circle. 

Pittsburgh +41

Seattle +20

Arizona +4

Cleveland -7 (but automatically over Cincinnati)

Cincinnati -2

San Francisco -16

Baltimore -18

And here's the web:

1) Houston>Carolina>Atlanta>Houston

2) Houston> Kansas City>Minnesota>Atlanta>Houston

3) Kansas City>San Diego>Chicago>Kansas City

In 1) Houston is +3, as is Atlanta, and Carolina is -6.

In 2) Minnesota is +18, Houston is +7, Atlanta is -11, and Kansas City is -14

In 3) Kansas City is +4, Chicago is -1, and San Diego is -3

So that sorts itself out all right-Carolina will go above Kansas City for having a closer score to Houston.  Houston will be above Atlanta for having a higher score in the 2^nd part of the web.

All right, with that figured out, here are the rankings:

1) New England
2) Green Bay
3) Dallas
4) Indianapolis
5) Tennessee
6) New York Giants
7) Jacksonville
8) Washington
9) Philadelphia (check it - they've only lost to higher teams)
10) Detroit
11) Denver
12) Oakland (practically deserves to join the muddle and bring Denver with it, due to the last-second field goals, but a win's a win)
13) Pittsburgh
14) Seattle
15) Arizona
16) Cleveland
17) Cincinnati
18) San Francisco
19) Baltimore
20) Buffalo
21) Tampa Bay (and they've only beaten lower teams)
22) Minnesota
23) Houston
24) Atlanta
25) Carolina
26) Kansas City
27) Chicago
28) San Diego
29) New York Jets
30) Miami
31) New Orleans
32) St. Louis

I italicized the tangles to emphasize that they're more or less interchangeable.

So, if the rankings hold true, these teams will win next week.  I'll compare this with my later predictions to see if NPR proves more reliable again.

Arizona

Tennessee

Carolina

New England

Washington

Jacksonville

Houston

NY Giants

Pittsburgh (crapshoot - Pittsburgh and Seattle are consecutive teams in a tangle)

Indianapolis

San Francisco (same situation for SF vs. BAL)

Denver

Green Bay

Dallas

Okay, the only one of these that's at all interesting is Washington over Detroit, which I'll probably be picking anyway.  But I will pick some upsets, so the overall score will be different. 

I think I am approaching the point where this will become impossible.  If any of the winless teams prevail this week, or the undefteated ones lose, it'll get so messy that it probably won't be worth the trouble.