NFL Score-based "Power Rankings"

[MagicHef]

Several years ago, I made up a speadsheet to determine some sort of NFL rankings actually based on objective data, rather than opinion. I thought that others on here might be interested, and have suggestions for improving it. What I came up with is a system that determines the difference between what a team scored against any opponent, and the average that the opponent gave up to other teams. It also does a similar score system for defense, but multiplied by -1 so that a higher score is always better. For example, when the Chiefs beat the Browns 16-14, they got an offensive score of -4.75 (the Browns gave up an average of 20.75 points in other games) and a defensive score of 3.5 (the Browns scored an average of 17.5 points in other games). The offensive and defensive scores are averaged for all the games throughout the season to give a total offensive and defensive value, which can be added to create a total, which should represent the total number of extra points that a team should score/prevent the other team from scoring. Zero represents league average.

There are obviously problems that arise from using scores, such as most teams not coached by Belichick tend to let up some when enjoying a large lead, so the end score may not be indicative of the true ability of a team. Also, safties, defensive touchdowns, and special teams touchdowns become positive scores for offense, so the defense/offense balance represented by the scores may not be completely accurate.

Here are the scores for 2010. These are regular season only. The number before the team name is rank, the number after is score.

Offense
1 NE 11.68
2 PHI 4.84
3 ATL 4.68
4 SD 3.86
5 IND 3.43
6 OAK 2.80
7 NYJ 2.49
8 GB 2.43
9 PIT 2.34
10 NO 2.30
11 DAL 2.07
12 DET 1.91
13 HOU 1.43
14 NYG 1.09
15 BAL 0.81
16 CIN 0.40
17 TB -0.21
18 DEN -0.58
19 KC -1.05
20 CHI -1.29
21 TEN -1.33
22 JAC -1.57
23 BUF -2.41
24 SEA -2.64
25 CLE -2.98
26 SF -3.36
27 MIA -3.58
28 MIN -3.83
29 WAS -3.89
30 ARI -4.33
31 STL -4.38
32 CAR -8.07

Defense
1 GB 7.11
2 PIT 6.90
3 BAL 4.54
4 SD 4.02
5 CHI 3.83
6 NYJ 2.80
7 TEN 2.62
8 ATL 2.30
9 MIA 1.99
10 KC 1.81
11 NE 1.75
12 CLE 1.21
13 STL 1.17
14 MIN 1.02
15 NO 0.75
16 NYG 0.48
17 TB 0.36
18 SF -0.14
19 IND -0.31
20 WAS -0.36
21 PHI -0.91
22 OAK -1.01
23 HOU -2.20
24 DET -2.22
25 CIN -2.45
26 JAC -3.25
27 CAR -3.89
28 BUF -4.02
29 SEA -4.38
30 DAL -4.43
31 ARI -6.11
32 DEN -6.38

Total
1 NE 13.43
2 GB 9.55
3 PIT 9.24
4 SD 7.89
5 ATL 6.97
6 BAL 5.35
7 NYJ 5.29
8 PHI 3.93
9 IND 3.12
10 NO 3.05
11 CHI 2.54
12 OAK 1.79
13 NYG 1.57
14 TEN 1.29
15 KC 0.76
16 TB 0.14
17 DET -0.31
18 HOU -0.77
19 MIA -1.59
20 CLE -1.78
21 CIN -2.04
22 DAL -2.36
23 MIN -2.81
24 STL -3.21
25 SF -3.51
26 WAS -4.26
27 JAC -4.82
28 BUF -6.43
29 DEN -6.96
30 SEA -7.02
31 ARI -10.44
32 CAR -11.96

First thing that I noticed:
How exactly did Seattle win a playoff game?

[Kraus]

First thing that I noticed:
How exactly did Seattle win a playoff game?

Because the Saints can't tackle.

<iframe width="425" height="349" src="http://www.youtube.com/embed/GOZkeFXvBPA" frameborder="0" allowfullscreen></iframe>

[NaptownChief]

Only problem I see is the formula only had Denver finishing 29th instead of 31st or 32nd ROFL

[MagicHef]

Only problem I see is the formula only had Denver finishing 29th. ROFL

They were dead last defensively, which is certainly accurate.

[NaptownChief]

They were dead last defensively, which is certainly accurate.


Which proves the formula is on the right track.

[kysirsoze]

Only problem I see is the formula only had Denver finishing 29th. ROFL

That and it has KC finishing 3rd in the division. No objective scale is going to be perfect, though, when it comes to actually trying to determine the quality of teams. Too many factors that go beyond stats. Looking over the list, it's not too far from reality.

[MagicHef]

That and it has KC finishing 3rd in the division. No objective scale is going to be perfect, though, when it comes to actually trying to determine the quality of teams. Too many factors that go beyond stats. Looking over the list, it's not too far from reality.

Yeah, I was really surprised by Oakland's offensive ranking.

[Extra Point]

Sign Hali, and let there be no SD @ #4, at least on D.

[Pioli Zombie]

"You play to win the score-based power rankings!" Herm

[Extra Point]

Yeah, I was really surprised by Oakland's offensive ranking.

Look at last year's games with OAK. That, my friend, is scary.

Points is points, wins is wins. Have to think that, reduced down to lowest terms, this system carries some weight.

[MagicHef]

it has KC finishing 3rd in the division.

According to this, both west divisions are terrible. AFCW winners were ranked 15th, and NFCW winners were ranked 30th. Meanwhile, SD finishes 4th in offense, 4th in defense, 4th overall, and misses the playoffs.

[MagicHef]

I can do this for next season, and do an update after each week. I won't be able to start until after week 2, though, because I need at least 2 weeks to compare.

[loochy]

I can do this for next season, and do an update after each week. I won't be able to start until after week 2, though, because I need at least 2 weeks to compare.

just like the chargers

[MagicHef]

Looking over the numbers again, the defensive rankings really did a superb job of predicting postseason success. #1 defense won the superbowl. #2 defense lost in the superbowl. Of the 6 teams that won a playoff game last season, 5 of them are in the top 6 on here. It's just the Chargers and Seahawks mucking things up.