NFL Power Rankings Week 3 and Projected Final Standings
| 7
|
by Derekcs
The idea behind my power rankings is simple: if every team played every other team at home and away, who would win the most games?
The rankings and projected standings were originally explained here. Basically, I take yards per play rushing and passing, third down conversion rates, sack rates, and fumble and interception rates for both offense and defense and plug them into a logistic regression system that tries to predict the probability of the home team winning. The statistics are adjusted for opponent quality (basically, if a team is 10% above average in rush defense, your rushing average will be adjusted to 90% of what it was, etc.). Each team's ranking is based on the sum of the probabilities that they would win each of the 62 possible games. Those totals are then normalized to make up an expected win total for a 16-game season. Much like baseball, the large schedule means teams are closer to .500 than they would be in an actual 16-game schedule. Please note that the prediction system favors offense over defense by a fair margin and passing over rushing by a very large margin, and though a team may be ranked higher than you'd expect, it doesn't always translate into a high projected win total (e.g. Giants).
Using the predictions, actual wins so far, and set of actual games remaining in the season, we can estimate each team's final win total. Basically, the team with the higher projected probability of winning gets credited with a win. In the case of the Colts and Saints, it doesn't mean they'll win/lose every game, it just means they're the favorite/underdog in every remaining game. The better team doesn't always win. The projected final standings are below the power rankings.
The tables seem to be crapping out here. Try this page for now: http://footballpredictionnetwork.blogspot.com/2007/09/power-rankings-2007-week-3.html http://footballpredictionnetwork.blogspot.com/2007/09/projected-final-standings-2007-week-3.html
Power Rankings
Rank Team Expected Wins
1 IND 10.7001
2 DAL 10.5789
3 PHI 10.4336
4 NE 10.3860
5 HOU 10.1382
6 JAX 9.4433
7 DEN 9.2701
8 SEA 9.1180
9 PIT 8.9880
10 WAS 8.6391
11 NYG 8.4719
12 TB 8.2504
13 TEN 8.1991
14 ARI 8.1641
15 MIN 8.0222
16 MIA 7.9489
17 CAR 7.8971
18 BAL 7.6365
19 GB 7.5572
20 ATL 7.5167
21 DET 7.3864
22 CLE 7.2844
23 BUF 7.1819
24 CIN 6.9683
25 SD 6.7951
26 SF 6.7815
27 KC 6.6613
28 OAK 6.5439
29 NYJ 6.3535
30 STL 6.1707
31 CHI 5.4655
32 NO 5.0481
Projected Final Standings Sorted by Division
AFC East
BUF 5
MIA 6
NE 14
NYJ 4
AFC North
BAL 7
CIN 7
CLE 6
PIT 11
AFC South
HOU 12
IND 16
JAX 10
TEN 9
AFC West
DEN 13
KC 6
OAK 4
SD 5
NFC East
DAL 15
NYG 6
PHI 12
WAS 7
NFC North
CHI 2
DET 8
GB 11
MIN 8
NFC South
ATL 6
CAR 7
NO 0
TB 9
NFC West
ARI 10
STL 1
SF 6
SEA 13
I would like to point out that my prediction about Detroit being 8-8, 9-7 team and backing into the playoffs still looks to be at least partially true (just not the playoff berth part with ARI and PHI projected to get the wild card spots). Unfortunately, the Dolphins part of that same prediction looks to be false right now, though I will say they've underachieved in terms of wins given their stats. I'm a saaaad Dolfan.
The raw data for these power rankings (all of the stats with team rankings for each stat) can be found here.
