Beatpaths NFL Power Rankings: Week 8

We’re still in beta, but in the home stretch. Some nifty changes in the power rankings this week.

First, the beatpath graph for Week 8. There’s no change since Sunday afternoon, as NE and PIT already had beatpaths over BUF and BAL, respectively.

2005-8-Clean

(read on for Week 8’s Power Rankings)

Technorati Tags: , , ,



And now for the power rankings. We might have an adjustment in the method to determine the power rankings in the next week, but it’s already looking like the ranking system is pretty stable.

Rank Team Notes Last Week BeatPoints

1

(Bye) Indianapolis won’t move from this spot until someone develops a beatpath over them.

1

2

(17 – 0 = 17)

2

(Beat PHI) Denver is like the beautiful psycho girlfriend. Her assets and flaws are both obvious for all to see. Denver’s win over Philadelphia gives them another very long beatpath to make up for losing their beatpaths below Jacksonville.

2

1

(22 – 0 = 22)

3

(Lost to STL) Jacksonville’s loss to St. Louis resulted in them losing their beatpath to Seattle. Each time a team loses a beatpath, it makes it easier for another team to develop a beatpath to them. But it hasn’t happened yet, so Jacksonville holds steady.

3

6

(11 – 2 = 9)

4

(Lost to NYG) Washington’s spectacular loss to the Giants blew away their beatpath to the Giants and Dallas. Expect some reshuffling as the division games heat up.

4

13

(7 – 1 = 6)

5

(Bye) The Seahawks have no beatlosses, and a beatpath to Dallas. They’re well-positioned over the NFC West.

5

4

(12 – 0 = 12)

6

(Bye) The Falcons also have no beatlosses, but their wins are over teams that haven’t been very impressive.

6

12

(6 – 0 = 6)

7

(Beat ARI) Holding steady at #7.

7

5

(11 – 1 = 10)

8

(Lost to DEN) The Eagles lost to a very good team, and come out of the week with a much stronger beatpath thanks to the Chargers win over Kansas City.

8

3

(15 – 1 = 14)

9

(Beat KC) A solid win over Kansas City helps the Chargers rise one slot.

10

8

(9 – 2 = 7)

10

(Lost to SD) For the first time in weeks, there is a team between San Diego and the Giants.

9

14

(8 – 3 = 5)

11

(Beat WAS) Yes it was a huge win, but they might end up wishing they had saved some of those points for other victories. I mean, look what happened to Washington.

11

7

(8 – 0 = 8)

12

(Beat MIN) Because of Tampa Bay and Miami, these next few teams all rise a couple of notches. Carolina had an impressive win over Minnesota.

14

10

(6 – 0 = 6)

13

(Beat BUF) They were supposed to get well over the Bye, and get well on Buffalo, but it still seems like they have a ways to go.

15

9

(7 – 1 = 6)

14

(Beat BAL) The Steelers beat Baltimore due to Baltimore’s poor two-minute judgment.

16

11

(9 – 3 = 6)

15

(Beat GB) It’s hard to fault any team for having trouble with Green Bay. Green Bay sucks, but they can still compete from time to time. Still, Cincinnati is looking mighty appropriate here in the middle of the pack.

17

15

(8 – 4 = 4)

16

(Beat TEN) They won on the road. Tennessee helped them win, but the Raiders could still sneak up on people…

18

18

(6 – 5 = 1)

17

(Beat DET) Chicago makes a clear move towards supremacy in the NFC North. They also rise an extra slot due to Cleveland’s collapse.

20

21

(3 – 6 = -3)

18

(Bye) The Jets rise on the Bye, mostly because Minnesota still has to be below Tampa Bay.

22

24

(0 – 6 = -6)

19

(Beat TB) Boy, the 49ers really made out. They developed beatpaths to TB, MIN, DET, MIA, and GB. So thanks to developing a beatpath over Tampa Bay, the 49ers rise six slots…

25

19

(5 – 6 = -1)

20

(Lost to SF) And Tampa Bay drops eight.

12

20

(4 – 7 = -3)

21

(Beat NO) Tampa Bay’s beatpath to Miami forces the Dolphins to sink eight slots as well.

13

25

(0 – 8 = -8)

22

(Lost to CAR) The Vikings were pushed down, not because of their loss to Carolina, but because of Tampa Bay’s loss to SF.

21

27

(1 – 13 = -12)

23

(Lost to CHI) Detroit holds steady, despite the collapse of a couple of teams that leapfrogged them backwards. What’s a backward leapfrog?

23

26

(1 – 12 = -11)

24

(Beat JAC) These next few teams rise a couple of slots thanks to the collapses of Green Bay and Cleveland. But it really wasn’t a surprise that St. Louis won – this team is dangerous.

26

16

(6 – 2 = 4)

25

(Lost to MIA) The subjective judgments still say that Miami is better than NO – I guess they were right, although it hasn’t yet become a beatpath.

27

22

(0 – 4 = -4)

26

(Lost to NE) It still doesn’t seem like they belong this low, but they’ll need a couple of wins to prove it.

28

23

(3 – 7 = -4)

27

(Lost to DAL) Arizona’s beatlosses are still only to solid teams. Seems like they have room to rise as well.

29

17

(5 – 4 = 1)

28

(Lost to OAK) Rises two slots through no merit of their own.

30

28

(4 – 16 = -12)

29

(Beat CLE) Houston made out by defeating a team that also has another beatpath.

31

30

(2 – 19 = -17)

30

(Lost to HOU) This was probably the biggest surprise of the week. Cleveland had been rising slowly but steadily all season long. A real collapse.

19

31

(1 – 20 = -19)

31

(Lost to CIN) The graph is a really sad sight, isn’t it? Green Bay sitting all alone at the bottom of the league. It’s all because Cleveland lost to Houston.

24

32

(0 – 27 = -27)

32

(Lost to PIT) And, holding steady at the bottom, the team that almost beat the Pittsburgh Steelers. Hey Baltimore – when the clock is running down and it’s fourth down at the end of the game, don’t throw to the middle of the field, short of the first down marker.

32

29

(0 – 17 = -17)

5 Responses to Beatpaths NFL Power Rankings: Week 8

  1. DavidH says:

    Ha, emoticons rock. On my browser, the Giants’ beatpoints equation is

    (8 – 0 = smiley face with sunglassess

  2. Chris says:

    I’m a bit confused: your explanations (and the rankings themselves) seem to indicate that a team that has previously been under another team’s beatpath cannot rise above that other team even if the beatpath is later neutralized.

    Illustration: If Team A beats Team B beats Team C, than A must be ranked ahead of B, who must be ranked ahead of C. But if C then beats A, removing the beatpath, A is still locked ahead of B, who is still locked ahead of C.

    While I would agree with locking teams ahead of who they currently have a beatpath over, I don’t agree that two teams without any clear beatpath relationship should be ranked according to the beatpaths they once had. It puts a inappropriate weight on early-season games, even though recent performance might be more indicative of a team’s true place.

    Does that make sense?

  3. ThunderThumbs says:

    Yeah, your comment makes sense. I think what is difficult, though, is deciding what alternative to use. Let’s say that A->B->C used to be a beatpath, and then C beat A. It becomes a beatloop, and it means that it is ambiguous which team is better than the other. And we really wouldn’t have enough data to say that the recent performance should be weighted over early performance – maybe it was just that all three games were home games.

    Because of that, it relies on earlier rankings, which means that A->B->C would be retained. But the point is that the locks (beatpaths) don’t exist anymore, which means it is significantly easier for other dynamics to adjust the placement. Maybe C develops a beatpath to someone who beats A, for example. But otherwise, in the absence of a beatpath, there is absolutely no reason for a team to overtake another. Not that I can think of, anyway. Actually, I have already coded in one exception, where the team would replace the team just above it if and only if they beat them head to head in the immediately preceding week. But I’m thinking of getting rid of that for consistency’s sake.

    There is one other behavior that I’m still trying to nail down, though. Say that A->B->C is arbitrary, suggested only by past rankings, but not by beatpaths. Then say C beats A. I would want the end result to be C->A->B. Not B->C->A. However, right now, the end result ranking is dependent on which order I evaluate the teams. For instance, earlier this season, Miami beat Denver. Denver soon neutralized the beatpath, and started rising. But in many situations, Miami started rising with them because Denver never developed a beatpath to Miami. After a while, Miami was ranked #3 and Denver #4. That doesn’t seem right. I’m still playing with that one.

    In general though, it’s about deciding on a basic principle of rising and falling. It seems like the two possibilities are to make teams degrade automatically unless they continue to play well, making it easy for other teams to rise up the power rankings, or, to have a king of the hill mindset, where teams don’t lose their place unless a team below them develops an actual beatpath to them (or to a team above them). I think the second one is better. If you’re champ over the teams below you, you keep it until it’s outright taken from you.

  4. Miles says:

    I can’t seem to find the BeatPoints explanation. Are you gonna blog about that at some point?

  5. ThunderThumbs says:

    BeatPoints blogged here.

Leave a Reply

Your email address will not be published. Required fields are marked *