Whew! That felt like a big week, didn’t it? And yet… not many of the games went counter to what you’d expect… and many of the games were between teams that were closely ranked. There’s only one game that has a huge effect on the graph, and that’s Minnesota over San Diego.
Here’s the graph:
First, I gotta say that the graph looks kind of pretty to me this week. I don’t know why, maybe it’s those curvy lines in the middle. It just strikes me as a bit aesthetically pleasing. Overall, the NFC hurt the AFC this week. The AFC is a little more split apart now in terms of quality. Two NFC teams rose into the top ten this week, while three AFC teams fell out. Three of the top ten teams are from the NFC North. Beatpaths playoff seedings at this point, with all the “screwed” teams that are ranked higher than the lowest-ranked division champion. AFC: NE, IND, SD, PIT, JAC, TEN, (Screwed: KC, HOU, DEN) And the beatloops: |
And finally, the rankings, below:
| Rank | Team | Notes | Last Week | BeatPower |
1 |
 |
(Beat IND) I never pay attention to intraday rankings shifts, but I did notice that between the early games and the late games, New England was briefly ranked behind Indianapolis. And it ended up being an evenly matched game. With Indianapolis’ injuries factored in, I’m very much looking forward to the rematch that everyone hopes will happen. |
1 |
100.0(30/30 – 0/30) |
2 |
 |
(Beat KC) Favre is being pretty savvy in how he’s motivating the team – speaking to the media about how past Green Bay teams he’s been on are better than this year’s version. |
2 |
100.0(25/25 – 0/25) |
3 |
 |
(Lost to NE) Losing to New England doesn’t hurt Indianapolis’ ranking. And now Indianapolis has one hell of a data point to help motivate them for the rest of their season. |
3 |
95.5(21/22 – 1/22) |
4 |
 |
(Beat PHI) Dallas and Jacksonville switch places – the more vertical graph gives more data to examine teams’ placements, and Dallas looks slightly better. |
5 |
96.0(24/25 – 1/25) |
5 |
 |
(Lost to NO) Jacksonville’s loss doesn’t hurt them – yet – because they had already lost their beatwin over Tampa Bay due to their loss to Tennessee. Their other victories still look strong, although their victory over Denver looks slightly less impressive. |
4 |
82.5(15/20 – 2/20) |
6 |
 |
(Bye) The Giants hold steady on the bye. |
6 |
88.5(23/26 – 3/26) |
7 |
 |
(Beat NYJ) Washington holds steady from their expected win over the Jets, although it was a close one. They’ve got a rematch with Philadelphia next week. |
7 |
84.6(22/26 – 4/26) |
8 |
 |
(Beat DEN) This game wasn’t on tv in my area this week. Good thing, it sounded like a real horror movie. I don’t think it was a surprise that Detroit was this good, but it was that Denver was this bad. Detroit just demolished Denver’s injury-depleted offensive line. |
11 |
78.8(20/26 – 5/26) |
9 |
 |
(Beat CAR) Tennessee rises a few slots after defeating Carolina. |
12 |
83.3(14/18 – 2/18) |
10 |
 |
(Beat SD) The largest move of the week goes to Minnesota for their upset of San Diego, thanks to a superhuman running back. I’ve been mystified ever since the game though, why was I so certain that a running back had already broken 300 yards in the past? I was convinced that Corey Dillon had when he was with the Bengals. Was that actually in the news at one point, and then later revised downward? I have 302 in my head for some reason. |
21 |
67.4(14/23 – 6/23) |
11 |
 |
(Lost to MIN) San Diego is ranked just behind Minnesota, having suffered the beatloss, but slips three on the loss. |
8 |
61.4(12/22 – 7/22) |
12 |
 |
(Lost to GB) Kansas City had a huge opportunity to rise – one scenario had them in the top five had they managed to defeat Green Bay. As it is, they hold about steady on the loss, sliding up one due to Denver’s collapse. |
13 |
64.7(9/17 – 4/17) |
13 |
 |
(Beat OAK) Houston improves their standings slightly after defeating Oakland. |
16 |
64.7(9/17 – 4/17) |
14 |
 |
(Lost to DET) It’s amazing how in the NFL, the balance of what makes a good team is so delicate. Denver was improving these last couple of weeks, but a couple of things go wrong and everything explodes. Denver was ranked ahead of Kansas City, but now they’re picked to lose. |
9 |
52.4(11/21 – 10/21) |
15 |
 |
(Beat ARI) Tampa Bay defeats Arizona as expected. |
15 |
46.9(6/16 – 7/16) |
16 |
 |
(Beat BAL) There’s no great reward for defeating Baltimore, although Pittsburgh sure looked dominating in doing so. But Pittsburgh hasn’t defeated any quality teams, and they’re suffering under their beatloss to Denver, who looks worse now after San Diego lost to Minnesota. So Pittsburgh slips a few spots. |
10 |
45.2(9/21 – 11/21) |
17 |
 |
(Lost to DAL) Philadelphia loses to Dallas. They rise a couple of slots due to other graph dynamics. |
19 |
35.0(2/10 – 5/10) |
18 |
 |
(Lost to TEN) Carolina loses to Tennessee but holds steady in the rankings. |
18 |
36.7(5/15 – 9/15) |
19 |
 |
(Beat CIN) The Bills win as expected, but the Detroit-to-Pittsburgh path weighs Buffalo down more. |
17 |
33.3(6/18 – 12/18) |
20 |
 |
(Lost to HOU) Oakland holds steady on the loss to Houston. |
20 |
36.4(8/22 – 14/22) |
21 |
 |
(Bye) Chicago is the lowest-ranked NFC North team, with no beatwins to any team. |
22 |
20.0(0/10 – 6/10) |
22 |
 |
(Beat SEA) Cleveland defeats Seattle, but isn’t helped very much by the win. The AFC North in general isn’t as good as people think. |
23 |
30.4(7/23 – 16/23) |
23 |
 |
(Beat JAC) New Orleans has their victory beatlooped away immediately, although the victory could always end up impacting the graph later. |
25 |
26.7(3/15 – 10/15) |
24 |
 |
(Lost to CLE) Seattle has the biggest downward move in the rankings, hurt greatly by their loss to Cleveland. And yet, they’re the highest-ranked NFC West team. Even the entire AFC West is ranked ahead of the entire NFC West. |
14 |
28.8(6/26 – 17/26) |
25 |
 |
(Lost to BUF) Cincinnati slips one ranking on the loss to Buffalo, and suffers a beatloss to them, too. |
24 |
20.8(5/24 – 19/24) |
26 |
 |
(Lost to PIT) Baltimore loses to Pittsburgh, as expected, reinforcing an existing beatpath. |
27 |
18.0(4/25 – 20/25) |
27 |
 |
(Lost to WAS) The Jets rise one notch, probably just from the NFC West looking so awful. |
28 |
4.3(1/23 – 22/23) |
28 |
 |
(Lost to TB) Baltimore’s loss does enable Arizona to shed one beatloss, which could give them more upward mobility in future weeks. But only if they win, which they didn’t against Tampa Bay. |
29 |
20.0(1/15 – 10/15) |
29 |
 |
(Beat SF) Atlanta falls three spots despite defeating San Francisco, from being leapfrogged by other bad teams. |
26 |
15.6(2/16 – 13/16) |
30 |
 |
(Bye) Miami rises one on the bye, due to San Francisco leapfrogging them downward. |
31 |
0.0(0/23 – 23/23) |
31 |
 |
(Lost to ATL) San Francisco still has a beatpath to St. Louis, so they can’t be last. |
30 |
5.6(1/27 – 25/27) |
32 |
 |
(Bye) St. Louis “holds steady” (at the very bottom) on the bye. |
32 |
0.0(0/27 – 27/27) |
[...] 1. Beatpaths has the Redskins at No. 7 overall, No. 4 in the NFC. And that’s before the win at New York. I’ve written about this site before, but, basically, their rankings are based on “wins, losses, who beats who, and nothing else.” It doesn’t care that a win came in overtime or a loss was by 45. Just wins and losses, and, by that measure, the Skins are damn near elite. Update: The Skins hold strong at 7. [...]
NFC: GB, DAL, TB, SEA, NYG, WAS (Screwed: DET, MIN, PHI, CAR, CHI, NO)
How would Washington make the playoffs instead of Detroit? I know that they beat the Lions, but the Lions are 6-2, a game better than the Redskins 5-3. We might even be ranked ahead of the Giants, depending on strength of schedule. I know, it is crazy to think of the Lions in the playoffs, but it is definitely possible. Heck, if we beat the Packers, we could win our division.
Lebkin,
It’s obvious these rankings aren’t going by win-loss records. That’s the whole purpose of this site.
TT,
I’m pretty sure this whole system of yours was designed so that in 2007, you’d have an excuse for putting the Broncos 14th. :p
But really, this team is baaaaaad. I haven’t had an opportunity to watch their games, but what happened to the Defense?
Dillon ran for 276 or thereabouts with the Bengals. Jamal Lewis had the record at 295 until last week. Nobody has topped 300 in the NFL.
Maintained beatwins courtesy of my incremental beatloop removal approach:
At 2/3 strength:
PHI->MIN
KC->MIN
GB->MIN
At 1/2 strength:
NO->SEA->SF
PIT->SF
TB->ARI
PIT->BAL->ARI
SD->HOU->KC
TB->NO->JAC
KC->SD
GB->SD
TB->NO->JAC
TB->TEN->JAC
At 1/3 strength:
PHI->DET->CHI (DET->CHI is actually 4/3 strength)
Additional arrows on the graph as a result of these beatwins:
TB->TEN->JAC
NO->JAC
SD->HOU->KC->SD
PHI->DET (the only big mover)
Redundant arrows that get removed:
WAS->DET (path through PHI)
PHI->NYJ (long beatpath)
IND->TEN (path through TB)
TB->CAR (path through TEN)
TEN->HOU (path through JAC)
NO->ATL (path through JAC)
JAC->KC (path through HOU)
KC->OAK (path through SD->DEN)
HOU->OAK (path through KC)
So you actually end up with fewer arrows, and a more vertical graph.
Just looking at the beatloops list, this whole thing would be a lot more simple if Chicago and Arizona would just figure themselves out. This season they seem to have a habit of occasionally beating a good team, and then loosing to a lousy one (good and lousy being relative to their own apparent position, of course).
Kirk,
I misunderstood what originally written in the post. I thought it was saying that Washington and New York would make the playoffs using the NFL system and that Detroit would get screwed because they’d be higher based on the beatpath rankings. Looking again, I realized I misread it, and everything is as it should be. This is what happens when I write posts too early in the morning.
The nice thing is that while we aren’t in the beatpath playoffs now, the Lions control their chances to move up. We play both the Giants and the Cowboys still this season. We could end up with beatpaths to both teams, which not only help raise our rankings, but may beatloop away our loses. A win against Green Bay would help do that as well. It is amazing to have such control over our own destiny.
I forgot that BAL->ARI would need to be added, and CAR->ARI consequently becomes redundant (CAR would have a new long beatpath to ARI through CAR->NO->JAC->etc). Also, the GB->KC arrow is redundant thanks to GB->…->TB->TEN->JAC.
I will probably stop working this out by hand in a week or two, as I expect it will become too complicated to manage. Figureing out how the MIN and CHI situations sorted out was pretty subtle already, and it won’t get any easier.
Still, I like this approach because it preserves more beatwins and thereby produces a more vertical (less ambiguous) graph.
and obviously that’s SD->HOU->KC getting added, not SD->HOU->KC->SD.
OK, I’m done spamming the thread.
Currently I have a new working version of the system that is more flexible and will allow me to plug in different beatloop removal methods. The standard method is finished and I’m currently putting together a version that uses the final scores to determine which paths get broken first. This method was used in a “what-if” earlier this year (Week 6), and someone has suggested to revisit it in more recent weeks.
Doktarr, after I complete this version I’ll see about adding an option to destroy paths based on repetition.
I posted the output results of my two runs.
http://www.twomuffin.com/BeatPaths.htm
http://www.twomuffin.com/BeatPathsWeighted.htm
The first one is the standard method for removing loops and the second “weighted” one removes the closest game from a loop first by decrementing all wins in the loop by the point differential of the closest game.
Nifty, what are you using for the graph package?
I should probably put a page up making clear the licensing blah blah. Normally this kind of thing would assume a more restrictive copyright, but I’m big on creative commons type stuff, like over on my music site. So for creating other “works” based off of what I’ve done here, it’s okay as long as there’s attribution of it being based off of beatpaths.com, and used for noncommercial purposes.
I’m not using any graph package. My program outputs the results in a text file. It lists all the teams with all of thier active BeatWins and I open up PhotoShop and build the graph myself. It also builds the HTML for me, I just have to plug in the name of the graph image after I’ve built it.
As far as copyrighting goes for programs, it only covers the code. Since I wrote my own code, it doesn’t apply. I don’t have any intention to distribute the code or methods though, so don’t worry about it. I just want to be able to participate in the discussions and run my own tests.
Nice. Let me know if you want any help explaining my approach. I sorta pseudocoded it the week 7 beatpaths thread.
And to further prune the graph in my alternate approach, you could get rid of IND->JAC and TEN->JAC, as both are covered by IND->TB->TEN->CAR->NO->JAC. Like I said, it’s hard to do it right by hand.
Doesn’t sound like a problem. If you’d like, I can put that content up here in a post on the site, maybe for the week’s “What If”.
Doktarr: I think I have the basic concept, and a framework to start from. If I can convert my tables from integers to floating point numbers to account for your fractional reduction, it shouldn’t be a problem. I probably won’t have it done until after the next round of games though.
TT: Right now I’m still fooling around with it. I’ll just post links if I have something to show. Is there a way you can make a better way to navigate the pages? I look through your old stuff occasionally and it can be a pain in the neck paging through the history looking for what I want.
MOOSE,
What are the paths in and out in your rankings? I notice that Indy has 4 paths in, but their only loss is to NE. How’s that?
In the regular rankings Indy has one path in, as does Dallas. In the weighted rankings Indy has 4 and Dallas has 21 because those were the point differentials in the games.