The picks were 10-6 this week. Biggest movers are Minnesota (down 10) and San Francisco (up 7). Playoff picture in the AFC is starting to look more stable, with Pittsburgh grabbing the final spot (albeit as a division champion). The NFC wants Detroit and the Giants in as the wildcards. Most of the impact on this week’s graph was due to the Washington@Minnesota game.
| Rank | Team | Notes | Last Week | BeatPower |
1 |
 |
(Beat MIA) So much for a twistedly poetic outcome against Miami. New England against the Giants will be the game of the week, although Indianapolis against Tennessee actually comes close in terms of respective rankings. |
1 |
100.0(31/31 – 0/31) |
2 |
 |
(Beat CAR) Dallas is just trying to play out the season at this point – most of the press has seemed to be of they sort that comes when sportswriters are bored. Wasn’t too much of a big deal made out of that Owens/Romo thing? |
2 |
95.2(20/21 – 1/21) |
3 |
 |
(Beat HOU) Indianapolis pretty much swamped Houston, who was otherwise looking like a good team lately. |
3 |
93.5(21/23 – 1/23) |
4 |
 |
(Beat DEN) San Diego rises, not so much because of defeating Denver (again), but because Washington’s victory over San Diego enables San Diego to finally shed their beatloss to Minnesota, which was getting pretty rickety. San Diego’s beatloop situation is pretty interesting at this point. One with IND and JAC – all top playoff teams. One with GB and CHI. A split with KC. And a really interesting long beatloop with DEN, BUF, WAS, and MIN – all teams that have been inconsistent this year. |
7 |
81.8(15/22 – 1/22) |
5 |
 |
(Beat OAK) Jacksonville swamps Oakland and is looking pretty dangerous. I’m really impressed with this team, it just seems like they way they are set up gives them a lot of solidity. They don’t seem to need to worry about inconsistency very much – I’m not sure what elements they need to add in to raise above Indianapolis and New England, though. They beat San Diego already, but are ranked behind them because San Diego beat Indianapolis (as fluky as that victory appeared to be). |
5 |
80.4(16/23 – 2/23) |
6 |
 |
(Beat BUF) Since Minnesota and Green Bay fall, it enables the Giants to climb ahead of them, but they also manage to rise above Tennessee because of slightly better tiebreakers – San Francisco looking stronger is probably the explanation. |
9 |
82.5(15/20 – 2/20) |
7 |
 |
(Beat NYJ) Tennessee hasn’t looked too impressive the last couple of weeks, but they aren’t suffering under any damaging beatlosses, especially due to their beatwin over Houston, which protects them from the Denver beatloss. They even split the series with Jacksonville this year. When your worst beatloss is to the #4 ranked team, you’re not doing too bad. |
8 |
70.6(10/17 – 3/17) |
8 |
 |
(Lost to CHI) It’s odd that Green Bay gets swept by Chicago. Even odder that Green Bay swept Minnesota, who swept Chicago. This is a rare double beatloop, which I’ve seen in baseball a lot but I don’t think I’ve seen it before in the NFL, at least not in the years since I’ve started doing it. Green Bay loses a lot of quality beatwins due to the Chicago losses, and this makes them look a bit weaker. |
4 |
72.2(10/18 – 2/18) |
9 |
 |
(Beat STL) It’s starting to dawn on sportswriters that Pittsburgh isn’t the dominating force that people wanted to believe earlier in the season… but they’re still pretty good. They’re the #6-ranked AFC team this week, and in the top ten, so they’re not screwing anyone out of the playoffs. |
11 |
75.0(11/16 – 3/16) |
10 |
 |
(Lost to CIN) The loss to Cincinnati didn’t affect anything this week, and Minnesota’s fall and Detroit’s smaller fall leads to Cleveland rising a couple this week. |
12 |
68.8(10/16 – 4/16) |
11 |
 |
(Beat KC) Detroit already had a beatpath to Kansas City, so the win doesn’t help them – and, Denver’s collapse in the rankings leads to Detroit falling slightly. |
10 |
53.3(5/15 – 4/15) |
12 |
 |
(Lost to SF) There isn’t very much beatpath data for Tampa Bay – the vagueness leads them to kind of be slipping around in the rankings a bit. Their loss to San Francisco removed some of their middling beatwins, but there wasn’t much reason for them to fall in the rankings, and San Francisco climbed significantly. |
14 |
53.6(3/14 – 2/14) |
13 |
 |
(Lost to IND) Houston is another team that gains benefit from Detroit and Green Bay no longer having beatpaths to them. The loss to Indianapolis doesn’t impact anything. |
15 |
56.2(6/16 – 4/16) |
14 |
 |
(Beat BAL) Seattle basically holds steady after the expected Baltimore victory, aside from Minnesota and Denver leapfrogging them from the wrong direction. |
16 |
52.9(6/17 – 5/17) |
15 |
 |
(Lost to NYG) Buffalo with an expected loss, also being leapfrogged by Minnesota and Denver. Incidentally, the beatpath that led to the Giants being favored over Buffalo was destroyed this week, so now the Giants have a direct beatwin over Buffalo. |
17 |
47.1(6/17 – 7/17) |
16 |
 |
(Lost to WAS) Minnesota’s beatwin over San Diego was really propping Minnesota up, and Minnesota suffers the biggest drop of the week. |
6 |
58.8(5/17 – 2/17) |
17 |
 |
(Beat CLE) Cincinnati splits the season series with Cleveland – it normally would have had no effect, but the flattening of the bottom half of the graph from other matches has led to some different tiebreaker behavior – a lot of the teams in this area are fairly clumped together. |
22 |
32.4(3/17 – 9/17) |
18 |
 |
(Beat GB) Chicago’s win hurt Green Bay but doesn’t really help itself much – too many other tough losses. |
18 |
40.0(2/15 – 5/15) |
19 |
 |
(Lost to SD) Couldn’t watch the game, but I checked the stats halfway through the third quarter and was stunned to see something like three first downs for Denver. I don’t know what the heck was wrong with this team this year. The injuries and personnel switches, and even the playoff eliminiation… it just doesn’t explain everything. There has to be some serious kind of negative mojo going on in that locker room, because the strong performances like against Kansas City strike me as the sort of thing that can happen when a team is temporarily able to lift themselves above a tough emotional dynamic that still exists no matter how much they try to pretend it doesn’t. I don’t know if it’s a poisonous player or some seriously bad coaching chemistry (they should be responding to Bates better)… but it just seems there’s an untold story here. Anyway, Denver slips because they lose their beatwin over Buffalo from back in the first week of the season. |
13 |
30.6(1/18 – 8/18) |
20 |
 |
(Beat MIN) Washington upsets Minnesota and creates an interesting beatloop with SD, BUF, DEN, and MIN, which explains all of the big moves in the graph this week. This was the game of the week in terms of beatpath efffects. Washington is helped by it, even though it moved some of the other teams more. |
24 |
53.3(3/15 – 2/15) |
21 |
 |
(Beat NO) Philadelphia actually holds steady after the New Orleans victory because Cincinnati and Washington leapfrog them. But they are ranked ahead of New Orleans now, and they do shed their beatloss to Seattle – they have more room to rise in the future. |
21 |
41.7(2/12 – 4/12) |
22 |
 |
(Lost to PHI) New Orleans has a slightly weaker beatpath below them, and they fall slightly. |
19 |
35.7(1/14 – 5/14) |
23 |
 |
(Lost to DAL) Carolina also loses some of its beatpath below it, due to San Francisco’s win, and falls a few slots. |
20 |
33.3(3/15 – 8/15) |
24 |
 |
(Beat TB) San Francisco sheds a few beatlosses, which helps it rise considerably, and also hurts some of the teams that had relied on the long beatpath to San Francisco through Atlanta. |
31 |
26.5(1/17 – 9/17) |
25 |
 |
(Lost to SEA) Baltimore loses to Seattle as expected and basically holds steady. |
25 |
23.5(2/17 – 11/17) |
26 |
 |
(Lost to JAC) Oakland slips after the big loss to Jacksonville. |
23 |
18.8(1/16 – 11/16) |
27 |
 |
(Lost to DET) Kansas City slips a little more… every week it’s a little more. |
26 |
13.3(0/15 – 11/15) |
28 |
 |
(Lost to TEN) The Jets hold steady after the expected loss. |
28 |
12.5(1/16 – 13/16) |
29 |
 |
(Lost to PIT) St. Louis doesn’t have as many teams below it this week. |
27 |
11.9(1/21 – 17/21) |
30 |
 |
(Beat ATL) Arizona creates a beatloop with Atlanta and manages to rise a couple of slots. |
32 |
15.2(0/23 – 16/23) |
31 |
 |
(Lost to NE) Miami can’t create some really cool headlines. |
30 |
2.5(0/20 – 19/20) |
32 |
 |
(Lost to ARI) And Atlanta is now the lowest ranked team in the league. It was close though, wasn’t it? What was the explanation for their signs of life? |
29 |
4.5(0/22 – 20/22) |
My graphs are up:
http://www.twomuffin.com/start.php?page=BeatPaths.htm
http://www.twomuffin.com/start.php?page=BeatPathsWeighted.htm
http://www.twomuffin.com/start.php?page=BeatPathsDoktarr.htm
I hope for anyone that thinks including point spread to help determine rankings takes a look at the weighted graph this week and sees how poorly the ratings turn out. KC, while not a good team, is not the worst in the league and DEN isn’t the second worst. CHI gets the tiebreaker in the CHI=>GB=>MIN mess which lifts them to 8th. SD’s horrible loss to MIN earlier in the season holds them down to 13th. It just doesn’t look right.
Meanwhile the juxtaposition of the standard graph and Doktarr’s shows an amazing difference here at the end of the season. The standard graph is a jumbled mess in the middle as many relationships have been severed due to contradictions. Doktarr’s on the other hand, is extremely tall and narrow. There’s a section of five teams in the middle that the algorithm states are unambiguously ranked 17th through 21st.
Also, if Green Bay beats Detroit in Week 17, there will be another double beatloop: DET->CHI->GB->DET. It really underscores how odd Chicago’s victories over the Packers were. (Not that I’m complaining, since I’m a Bears fan!)
I have a quick question about how beatloops are resolved. Let’s say that there are two beatpaths caused by a single victory, such as MIN->NO->ATL->MIN and SD->NO->ATL->SD. Under the current system, both beatloops would be removed. But what if NO beats ATL again? Would both of its victories be removed since there are two separate beatloops? Or would only one of its victories be removed, leaving it with a path to ATL?
Only one victory gets removed in this case. The way it works is that you take all loops and reduce the win count of the loops by 1 at each path length. So when we get to loops of size 3, all of those relationships would be erased except for NO=>ATL which would have its strength reduced from 2 to 1. If the NO=>ATL beatwin appears in a beatloop later when considering size 4 or greater, it then would be eliminated.
In the weighted graph the number of wins is irrelevant, and two wins of 1 point each are less powerful than one of 3 points.
Doktarr’s algorithm is similar to the standard in that the NO=>ATL path would start off at 2 strength, however it uses a weight system to resolve loops. The first pass in this case would find a minimum weight of 1, just like the standard, but it would apply it to the NO=>ATL twice since it appears in both loops, and taking away both wins.
So the short answer is: In the standard algorithm, one win survives. In the weighted algorithm, it would depend on the point spreads of the games. In Doktarr’s algoritm, both wins would be removed.
I kind of feel like you should never be surprised by what happens in divisional games. See, for example, the Titans and Texans against the Colts in recent years, the Dolphins against the Patriots (before this year), etc., etc.
I don’t really have any data to back it up, but it seems like those kind of weird, never-shoulda-been-that-close (or actual win) games happen in divisional games.
Long story short, it’s not entirely surprising to me that the Bears beat the Packers twice. Go Bears!
Ha ha, Kenneth I’d tease you for rationalizing in favor of the Bears, but I agree with you. The weird NFC West results of the last few seasons should back that up (although it seems more sane this season).
Regarding your Denver comment… they had two players killed in the offseason, including the awful Darrent Williams shooting last new year’s… That might be part of what you’re seeing here.