Week 6 Power Rankings


Calculated Power Rankings, Unit Strength, Strength of Schedule, and Estimated Final Record following week 6 of the NFL season.  I use an engineering method of iteratively solving for best fit to the data (scores) to create the ranks and related applications that corrects for opponents played.

See earlier power rankings posts for a more detailed calculation methodology.

Strength factors for each unit of each team are calculated, what they are expected to score or give up to an average opponent, on offense and defense, equivalent to scoring offense or defense (NFL.com stat) but opponent adjusted. The calculation is basically making each unit of each team a variable, and solving for each variable iteratively, what would most closely create the NFL season.

The net difference between the factors is then considered team strength, the teams ranked by this for power. The numbers are very similar and represent the same thing as scoring offense or scoring defense (ppg) shown in NFL.com stats, however corrected for strength of opponent by iteratively solving the polynomial of actual results, with the solution being the best fit.

Early in the season this was smoothed with a preseason projection, it takes several weeks for the data to “calm down” and give reasonable results, since then I’ve completely dropped all preseason projections from the calculation, however I’ve added hot/cold calculation.

To acknowledge the hot/cold factor of teams, I am running 2 versions of the calculation, from the beginning of the season, and the most recent 4 weeks. The recent calculation is then combined with the season long calculation at a 1:4 ratio, to tweak the results for hot and cold teams. It is also provided separately for information purposes. I expect this short term calculation to fluctuate wildly and over time highly diverge from the primary calculation, right now the only difference is the week 1 games.

The previous rank is in parenthesis, ranked by net expected scoring against an average opponent, recent calculation only rank in brackets (the main rankings should trend toward the bracket)

Week 6 Power Rankings:
1. New Orleans Saints (1) – 18.1 [8]
2. New England Patriots (5) – 17.6 [1]
3. Atlanta Falcons (3) – 15.7 [2]
4. Indianapolis Colts (2) – 13.2 [3]
5. Denver Broncos (6) – 11.2 [6]
6. Baltimore Ravens (7) – 8.6 [7]
7. Minnesota Vikings (9) – 8.5 [9]
8. Chicago Bears (4) – 7.7 [4]
9. Arizona Cardinals (19) – 7.1 [5]
10. Miami Dolphins (8) – 5.5 [18]
11. Green Bay Packers (17) – 5.0 [11]
12. San Francisco 49ers (15) – 4.6 [12]
13. New York Giants (13) – 4.4 [13]
14. Seattle Seahawks (10) – 2.9 [10]
15. New York Jets (11) – 2.4 [21]
16. Cincinnati Bengals (14) – 1.8 [17]
17. Pittsburgh Steelers (16) – 0.6 [15]
18. San Diego Chargers (18) – 0.5 [16]
19. Dallas Cowboys (20) – 0.4 [19]
20. Philadelphia Eagles (12) – 0.3 [20]
21. Houston Texans (22) – -1.7 [14]
22. Buffalo Bills (23) – -4.7 [26]
23. Jacksonville Jaguars (21) – -8.1 [22]
24. Cleveland Browns (25) – -9.2 [23]
25. Kansas City Chiefs (28) – -10.0 [25]
26. Detroit Lions (24) – -10.4 [24]
27. Carolina Panthers (27) – -11.6 [27]
28. Oakland Raiders (31) – -12.7 [28]
29. Washington Redskins (29) – -14.2 [30]
30. Tennessee Titans (26) – -16.3 [31]
31. Tampa Bay Buccaneers (30) – -17.5 [32]
32. St. Louis Rams (32) – -19.3 [29]

NE’s win might have caused a bit of an overreaction, but both teams were at the opposite end of the spectrum to begin with. Some more upsets this week, but really it is some middling position shuffling. The pattern still shows, even stronger, 10 very bad or worse teams and a lot of average to above average teams. There are very few below average teams.

Top 5 Offenses:
1. New Orleans Saints – 36.8
2. New England Patriots – 31.6
3. Minnesota Vikings – 30.0
4. New York Giants – 27.8
5. Indianapolis Colts – 27.4
Ranked by expected scoring against an average defense.

Top 5 Defenses:
1. Atlanta Falcons – 10.6
2. Denver Broncos – 11.8
3. New England Patriots – 14.0
4. Indianapolis Colts – 14.2
5. New York Jets – 15.2
Ranked by expected points scored against by an average offense.

5 Worst Units:
1. Washington Redskins Offense – 6.6
2. Oakland Raiders Offense – 8.2
3. St. Louis Rams Offense – 9.7
4. Tennessee Titans Defense – 31.7
5. Tampa Bay Buccaneers Offense – 12.2
Ranked by the greatest negative deviation from the average opponent (21.2 ppg)

Who’s Hot:
1. New England Patriots – 11.9
2. Atlanta Falcons – 10.7
3. Indianapolis Colts – 9.9
4. Arizona Cardinals – 8.8
5. Chicago Bears – 8.6
Teams whose expected net scoring agaisnt an average opponent varies the greatest (positive) between the full season run, and the most recent 4 week run.

Who’s Not:
1. New York Jets – -13.9
2. Buffalo Bills – -11.3
3. New Orleans Saints – -7.7
4. Philadelphia Eagles – -6.6
5. Miami Dolphins – -6.4
Teams whose expected net scoring agaisnt an average opponent varies the greatest (negative) between the full season run, and the most recent 4 week run.

As the season goes along, I expect the whose hot/not lists to become much more interesting, to help ID teams with bad starts playing good ball, or vice versa.

I added another application to my program, basically a strength of schedule calculation. I chose to use the power rankings as the data to create the calculation. The average weekly ranking of all opponents will be what is used to rank the teams. Two methods are tracked, what the opponent was ranked at the time of the game, and the most recent ranking of each opponent, taken to be the most accurate. Right now I’m only reporting data based off of the most recent ranking, though watching the other as well. I think this will give a little different picture than pure W-L as to schedule difficulty, to try to find a better way.

The number in parenthesis is the ranked difficulty of the future schedule. The number reported is the average power ranking rank of all opponents played thus far. Teams are ranked by the overall difficulty of schedule.

Strength of Schedule:
1. Detroit Lions (5) – 12.2
2. Carolina Panthers (2) – 20.4
3. Tampa Bay Buccaneers (1) – 21.7
4. St. Louis Rams (3) – 16.0
5. Chicago Bears (10) – 14.2
6. Buffalo Bills (14) – 13.8
7. New England Patriots (15) – 13.5
8. Baltimore Ravens (11) – 15.3
9. Cleveland Browns (25) – 12.2
10. New York Jets (17) – 14.3
11. Miami Dolphins (20) – 12.4
12. Tennessee Titans (21) – 13.7
13. Cincinnati Bengals (22) – 14.0
14. Pittsburgh Steelers (4) – 20.3
15. Minnesota Vikings (8) – 18.5
16. Green Bay Packers (13) – 17.8
17. San Francisco 49ers (26) – 13.0
18. Houston Texans (6) – 20.2
19. Atlanta Falcons (28) – 11.8
20. Jacksonville Jaguars (16) – 18.3
21. Indianapolis Colts (19) – 17.2
22. Oakland Raiders (23) – 17.0
23. Philadelphia Eagles (9) – 22.4
24. Arizona Cardinals (29) – 14.8
25. Seattle Seahawks (31) – 14.7
26. Dallas Cowboys (18) – 20.2
27. New York Giants (12) – 22.2
28. Kansas City Chiefs (24) – 19.2
29. Denver Broncos (27) – 17.8
30. Washington Redskins (7) – 25.7
31. San Diego Chargers (32) – 13.2
32. New Orleans Saints (30) – 19.2

I also added another application that uses the schedule programming, used to create the strength of schedule, current records, and the latest power ranking run, to determine an estimate for final records.

Estimated Final Record:
Denver Broncos – 15-1
New Orleans Saints – 14-2
Indianapolis Colts – 14-2
New England Patriots – 13-3
Minnesota Vikings – 13-3
Atlanta Falcons – 13-3
New York Giants – 11-5
Baltimore Ravens – 11-5
Arizona Cardinals – 11-5
Cincinnati Bengals – 10-6
Chicago Bears – 10-6
Green Bay Packers – 10-6
San Francisco 49ers – 10-6
Seattle Seahawks – 9-7
San Diego Chargers – 9-7
Miami Dolphins – 9-7
New York Jets – 9-7
Dallas Cowboys – 8-8
Pittsburgh Steelers – 8-8
Philadelphia Eagles – 8-8
Jacksonville Jaguars – 7-9
Houston Texans – 6-10
Buffalo Bills – 6-10
Oakland Raiders – 4-12
Cleveland Browns – 4-12
Kansas City Chiefs – 4-12
Carolina Panthers – 3-13
Detroit Lions – 3-13
Washington Redskins – 2-12
Tennessee Titans – 1-15
St. Louis Rams – 1-15
Tampa Bay Buccaneers – 0-16

Estimated record uses the schedule matrix and power rankings, and assumes that the team higher in the current rankings will beat the team lower in the current rankings a certain % of the time, throughout the remainder of the season. These projections will change weekly with the power rankings and actual results.

The code captures the net expected points for each team, and compare them (basically equivalent to the line for a game, what the expected outcome of the game is), for every game remaining during the season, using the most recent data run.

If a team is favored by 14+ points, I assumed that they win 100% of the time.
If a team is favored between 7 and 14 points, I assumed that they win 80% of the time.
If a teams is favored between 0 and 7 points, I assumed that they win 60% of the time.
If a team is not favored between 0 and 7 points, I assumed that they win 40% of the time.
If a team is not favored between 7 and 14 points, I assumed that they win 20% of the time.

This is then used to compute wins and rounded to the nearest integer. (plus a correction made so that there are 256 game winners).

Power Ranking Performance:
Week 1 record – 14-2
Week 2 record – 11-5
Week 3 record – 12-4
Week 4 record – 11-3
Week 5 record – 7-7
Week 6 record – 9-5
Missed Last Week – Cin-Hou, Mn-Balt, Phi-Oak, Sea-Ari, NYJ-Buf

Advertisements

Tags: , , , , , , , , , , , , , , , , , , , ,

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


%d bloggers like this: