The beatflukes power rankings are actually quite different this week, after only two games. There are a lot less beatflukes, and more beatloops – in other words, more game matchups from the season are mattering in the rankings. Why? It’s one change that has a huge effect on the ecology of the graph. When a butterfly flaps its wings, baby!

1. Chicago’s upset win over New Orleans allows it a different beatpath route to Green Bay, which means that
2. GB=>CHI is now considered a beatfluke, which
3. removes the GB=>MIA link in the graph, which
4. puts MIA=>KC back in the graph, which means that
5. STL=>GB (previous removed by the MIA=>KC beatloop) is now removed by an additional SF=>SEA beatloop, which
6. busts a whole bunch of Seattle beatpath segments, which
7. removes a bunch of beatfluke Seattle wins that Seattle wasn’t getting credit for, which
8. enables Seattle to rise… and
9. KC’s weakness means that
10. KC=>SD is now considered a beatfluke…

and this basically spreads out into a whole lot of other craziness. The ecology of the NFL is different because CHI beat NO, and here’s the effect.

And, the rankings:

Rank	Team	Notes	Last Week	BeatPower
1		(Beat NE) Let’s start simple.	1	100.0 (29/29 – 0/29)
2		(Lost to IND) No real change for New England, still ranked behind Indianapolis. They climb because San Diego fell.	3	92.9 (25/28 – 1/28)
3		(Bye) Baltimore only climbs because San Diego falls.	4	88.5 (21/26 – 1/26)
4		(Bye) The tiebreaker subtly changed this week, partly because Carolina no longer has a beatloss, having shed their beatloss to Cincinnati. As a result, only the very few strongest beatwins are counted in calculating the tiebreaker. San Diego has a wider breadth of beatwins than other teams, but NE and BAL had better wins against very quality opponents.	2	96.0 (23/25 – 0/25)
5		(Beat NO) GB=>CHI is now considered a beatfluke. But it’s the beatwin over New Orleans that really enables them to climb. In the vanilla variant, Chicago is ranked ahead of San Diego.	9	84.8 (18/23 – 2/23)
6		(Lost to CHI) Other dynamics in the graph enable New Orleans to climb slightly. New Orleans loses credit for beatwins over the Giants, Tampa Bay, and Atlanta, but also sheds beatlosses to Cincinnati and Pittsburgh.	8	73.8 (14/21 – 4/21)
7		(Bye) The Jets lose credit for a direct beatwin over Houston, but shed their beatloss to Cleveland, which was holding them down. The Jets rise.	14	82.5 (15/20 – 2/20)
8		(Bye) Same exact relationships for Buffalo, but other fallen teams help them to rise. Vanilla: #13	11	69.4 (11/18 – 4/18)
9		(Bye) Same relationships for Philadelphia, but other fallen teams help them to rise. Vanilla: #8	10	68.4 (12/19 – 5/19)
10		(Bye) ATL=>PIT is no longer considered a beatfluke. As a result, PIT loses credit for beatwins over NO and CLE. In addition, DEN=>PIT=>KC disappears. Vanilla: #12	7	70.5 (12/22 – 3/22)
11		(Bye) SEA=>ARI, SEA=>GB, and SEA=>DEN are no longer considered beatflukes. SEA and ARI now have a season split. SEA sheds both SF losses in beatloops, and KC=>SEA=>DEN disappears. Seattle looks better in the ecology and rises. Vanilla: #9	20	60.9 (9/23 – 4/23)
12		(Bye) Miami sheds a beatloss to Green Bay. Vanilla: #20	16	55.3 (9/19 – 7/19)
13		(Bye) No relationship changes for Tennessee. It appears to be Seattle’s and the Jets’ rise that enables TEN to rise. Vanilla: #11	22	56.2 (7/16 – 5/16)
14		(Bye) KC=>DEN is not considered a beatfluke anymore. As a result, KC sheds beatlosses to CLE, PIT, CIN, and loses beatwins to SEA and STL. Vanilla: #15	17	41.7 (5/18 – 8/18)
15		(Bye) ATL=>CAR, WAS=>CAR, NYG=>CAR, and MIN=>CAR are no longer considered beatflukes. CAR has no beatflukes anymore. CAR loses credit for defeating TB and STL, and sheds beatlosses to CIN, WAS, NYG, and MIN. Vanilla: #16	12	60.7 (3/14 – 0/14)
16		(Bye) KC=>DEN, SF=>DEN, STL=>DEN, and SEA=>DEN are no longer considered beatflukes. As a result, SF=>DEN=>ARI disappears, as does all the combinations of SEA|STL=>DEN=>CLE|PIT|CIN . Denver lost some high quality wins, in addition to the wins over NE and SD they already had beatlooped away, so they fall. Vanilla: #10	5	46.2 (1/13 – 2/13)
17		(Bye) TB=>CIN and ATL=>CIN are no longer considered beatflukes. They shed losses to TB, ATL, and DEN, and lose wins to CAR, CLE, NO, and KC. Vanilla: #17	6	46.7 (2/15 – 3/15)
18		(Bye) SEA=>DAL re-emerges but that doesn’t hurt Dallas, and they rise a bit. Vanilla: #22	23	44.7 (6/19 – 8/19)
19		(Bye) Wins over ARI, SF, GB, and DEN are not considered flukes anymore. A bunch of matches are looped away, but STL rises a bit. Vanilla: #19	25	44.4 (2/18 – 4/18)
20		(Bye) STL=>GB and SEA=>GB are no longer considered flukes, while GB=>CHI is. This is what started all the other shuffling. Vanilla: #14	15	42.5 (6/20 – 9/20)
21		(Bye) NYG=>CAR is no longer considered a fluke. Vanilla: #18	24	38.9 (6/18 – 10/18)
22		(Bye) MIN=>CAR is no longer considered a fluke. Vanilla: #24	21	34.1 (5/22 – 12/22)
23		(Bye) Wins over CAR, ARI, PIT, and CIN are no longer considered flukes. Vanilla: #26	27	34.8 (5/23 – 12/23)
24		(Bye) Losses to TB and HOU are no longer considered flukes. Vanilla: #25	13	27.8 (1/18 – 9/18)
25		(Bye) Losses to SEA, STL, OAK, and ATL are no longer considered flukes. Vanilla: #21	18	25.0 (0/20 – 10/20)
26		(Bye) STL=>SF and SF=>DEN are not considered flukes anymore. Vanilla: #23	19	22.5 (1/20 – 12/20)
27		(Bye) TB=>CIN and TB=>CLE are not considered flukes anymore. Vanilla: #28	28	8.3 (0/18 – 15/18)
28		(Bye) WAS=>CAR is not considered a fluke anymore. Vanilla: #27	29	18.8 (3/24 – 18/24)
29		(Bye) No relationship changes for Detroit. Vanilla: #30	26	7.5 (0/20 – 17/20)
30		(Bye) HOU=>CLE is no longer considered a fluke. Vanilla: #29	30	11.4 (2/22 – 19/22)
31		(Bye) No relationship changes for Jacksonville.	31	0.0 (0/21 – 21/21)
32		(Bye) OAK=>ARI is no longer considered a fluke.	32	3.6 (0/28 – 26/28)