With at least 6 and (for half the league) 7 games in the books we have a decent sample size to draw off of in evaluating teams. There's still a razors' edge between leading your division at 5-1 and looking at a formidable hole sitting 1-5, but we have some indicators as to which teams owe more to friendly bounces and strength of schedule than to actual talent.
There are 17 teams above .500 going into week 8, 5 more are at or within a game of break-even
5-1: Jets, Patriots, Steelers
5-2: Ravens, Titans, Giants, Falcons
4-2: Texans, Colts, Chiefs, Bucs, Seahawks
4-3: Redskins, Eagles, Bears, Packers, Saints
3-3: Dolphins Cardinals
3-4: Jags, Raiders, Rams
So who's legit and who's due for a crash? 3 measures should give some clues, Pythag wins (point differential), DVOA and my personal opinion. They all have their pros and cons. Pythag wins are simple and objective, but don't include significant factors like schedule. DVOA factors in a lot more information, but it's run through a black-box and can't account for in-season injuries. Opinion is subjective and while it can account for any factor you'd want, the data is spotty and incomplete while the processing is "problematic". Somewhere between the 3 there's a good assessment to be had.
Pythagorean wins is a concept taken from Bill James' statistical work in baseball. He found that plugging runs scored and runs allowed into a formula reminiscent of the Pythagorean Formula he could very accurately predict a teams record that year.
Pythag wins are calculated (Points Scored^2)/(Points Scored^2+Points Allowed^2)
Makes sense, doesn't it? Scoring more than you allow generally leads to wins, while the opposite is the mark of a losing team. The real usefulness of Pythag wins shone through when again and again statgeeks found that a teams' Pythag wins the previous was a better predictor of a teams wins the following year, than their actual win total the previous year! Very good evidence that a Pythag wins are a more accurate judge of team strength than their real win total. The reason would likely be that a team with a small difference in points scored vs points allowed was in many close games, which makes it more likely their wins and/or losses came down to luck, compared to a team that consistently blew out, or was blown out by opponents. This explanation would fit with FO's findings that rolling over inferior teams is a more reliable indicator of team strength and narrowly beating out quality teams.
Tampa Bay, 4-2 actual, 2.2 pythag wins.
Houston, 4-3 actual, 2.7 pythag wins.
New England 5-1 actual 3.8 pythag wins
Kansas City, 4-3 actual, 3.9 pythag wins.
Chicago, 4-3 actual, 3.8 pythag wins.
Philadelphia 4-3 actual, 3.8 pythag wins.
Look out for:
San Diego, 2-5 actual, 4.1 pythag wins.
Detroit, 1-5 actual, 3.1 pythag wins.
Green Bay, 4-3 actual, 4.2 pythag wins
Don't buy: Tampa Bay, Houston, New England,
NY Giants, 5-2 actual, 4.0 pythag wins.
Jacksonville, 3-4 actual, 2.0 pythag wins
Arizona, 3-3 actual, 1.6 pythag wins
AFC: Tennessee (5.2-1.8), Pittsburgh (4.4-1.6), NY Jets (4.3-1.7), Kansas City (3.9-2.1), Indianapolis (3.8-2.2), New England (3.8-2.2)
NFC: Atlanta (4.3-2.7), Green Bay (4.2-2.8), NY Giants (4-3), Seattle (3.3-2.7), Chicago (3.8-3.2), Philadelphia (3.8-3.2)
The Outsiders Say:
Tampa Bay, 4-2 record (T-8th), -14.9% DVOA (25th).
San Diego, 2-5 record (T-25th), +13.9% DVOA (8th).
Dallas, 1-5 record (T-28th), +2.1% DVOA (17th).
Chicago, 4-3 record (T-11th), -11.2% DVOA (24th).
Washington, 4-3 record (T-11th), -7.3% DVOA (23rd).
Pittsburgh, 5-1 record (T-1st), +27.8% DVOA (2nd).
Tennessee, 5-2 record (T-4th), +25.9% DVOA (5th).
New England, 5-1 record (T-1st), +26.2% DVOA (3rd).
Indianapolis, record 4-2 (T-8th), +12.3% DVOA (10th).
NY Giants, record 5-2 (T-4th), +27.9% DVOA (1st).
Look out for:
San Diego, Dallas,
Philadelphia, 4-3 record (T-13th), +25.0% DVOA (6th).
Kansas City, 4-2 record (T-8th), +26.2% DVOA (4th).
Miami, 3-3 record (T-18th), +5.9% DVOA (13th).
Tampa, Chicago, Washington,
Arizona, 3-3 record (T-18th), -37.6% DVOA (32nd).
Jacksonville, 3-4 record (T-20th), -25.5% DVOA (28th).
Baltimore, 5-2 record (T-4th), 8.4% DVOA (12th).
AFC: Pittsburgh (11.7 mean wins), New England (11.3 mean wins), Kansas City (11.0 mean wins), Tennessee (10.4 mean wins), NY Jets (11.1 mean wins), Baltimore (10.2 mean wins)
NFC: Atlanta (10.8 mean wins), NY Giants (10.7 mean wins), Seattle (9.3 mean wins), Green Bay (8.7 mean wins), Philadelphia (9.5 mean wins), New Orleans (8.7 mean wins)
Tampa Bay. They've allowed more points than they've scored, FO has them with one of the softest schedules to date in the league. They don't do anything particularly well and have both the leagues' least effective pass rush while also being weak against the run.
Chicago. While Pythag says 4-3 is about right, the Bears needed two wins gift wrapped for them to be over .500 against a very mediocre schedule. Throw in Cutler being sacked once per 8 dropbacks in an offense calling a 61-39 pass-run balance and they're going down.
Jacksonville. Worst point differential in the league at -79, wretched DVOA. Pencil in another win over Indy, but that's probably it for them.
Tennessee. As much as it pains me, they are a dangerous team. Sure VY's "progression" is from walking disaster to game manager and Chris Johnson is cracking under the heavy workload they're leaning on him for, but that defense has really turned around. 18 turnovers so far (2nd in the league), 25 sacks (tied for 1st) and the #3 scoring D. Their pass D has been fantastic. 3.95 ANY/A against. How ineffective is that? Not only worse than Vince Young now, it's half a yard per dropback worse than 06/07 Vince Young. The closest performance to that this year so far is Brett Favre at 4.1 ANY/A. They can really play the pass.
Pittsburgh. Another dominant defense, but the Steelers have hope of a real passing attack coming along. Both FO and Pythag buy into the hype. Ick.
Kansas City. We all saw firsthand that they have a dangerous running game and a legitimate defense. Cassel isn't anything to write home about, but he's competent enough.
Look out for:
San Diego. The almost seemed like they were trying to give that game away against New England, but how many teams could give the Pats a short field off turnovers 3 times and still have decent shot to send it to OT? They'll have McNeil and Vincent Jackson back for the late run and their special teams just can't possibly continue to be this astoundingly awful the rest of the way. Regression to the mean can be a good thing.
Green Bay/New England/Indy All in the same boat. Great passing Os, defensive struggles and major question marks over whether they can adequately replace vital parts subtracted mid-season. Any of these 3 could fall off the map, or reemerge as a big favorite.
Detriot. They won't be mounting a playoff run, but they look awfully competent for a 1-5 team. Shaun Hill looks like the man who led the 2-6 Niners to a 5-3 second half in 2008, sabotaging what had looked certain to be a top 5 pick going to the New England Patriots. The two glaring weaknesses that doomed them in previous years, no pass pro and even less pass defense, have been greatly improved. The Lions have allowed 11 sacks in 6 games, which doesn't sound THAT improved, until you factor in that they are throwing the ball a league high 43.2 times per game. On the other side of the ball they are actually slowing teams down through the air thanks largely to their pass rush actually being a factor. Led by Suh with 4.5 in 6 games they are on pace to out-sack last years' D by 19!
Jacksonville, Chicago, Arizona and Tampa obviously
Baltimore. They aren't bad by any means, since they don't do anything poorly, but can a team be good not doing anything well? By DVOA: #15 pass O, #14 rush O, #16 pass D, #19 run D, #10 STs and they're the 2nd most consistent team in the league (4.1% variance). Balance is great, but can a team be that much better than the sum of it's parts?
Houston. The Texans have big wins over IND and KC bookending their schedule to date, but close scrapes with WAS and OAK along with stompings at the hands of DAL and NYG make them hard to peg. FO's numbers bare this out extraordinarily well. The Texans are DVOA's top offense with the highest variance of any offense in the league, and the league's 30th defense with the least variance of any defense. The Texans are going to play turrible defense, whether the offense shows up to bail them out is a crapshoot. If DVOA is to be believed, no offensive unit in the league has outperformed the collection of passing games the Texans have faced to date. If you can't stop the pass, no game is ever over, and if you can't count on your offense to perform, well you could drop any and every game.
AFC: Pittsburgh, Kansas City, New England, Indianapolis, Tennessee, San Diego
NFC: NY Giants, Atlanta, Green Bay, Seattle, Philadelphia, New Orleans
Absolutely Legit: Pittsburgh, Tennessee, NY Giants, Kansas City
Potential Contenders: Indianapolis, New England, Green Bay, San Diego
Clearly Dangerous: Atlanta, Philadelphia, NY Jets
Entirely Competent: Baltimore, Seattle, New Orleans
Decent Overachievers: Houston, Chicago, Washington, Tampa Bay
Still Fighting: Miami, Detroit, Oakland, St. Louis, Denver
No Good: Cincinnati, Minnesota, Dallas, Cleveland
Truly Awful: Buffalo, Jacksonville, Arizona, Carolina, San Francisco