Here’s the post-MFN beatpaths graph for Week 9. No rock-papers-scissors for IND/DEN/NE! Last week’s power rankings were 8-6, and we’re 72-56 for the year, or 13th in King Kaufman’s Panel O’ Experts.
Splits: WAS/DAL
Beatflukes: CAR=>TB, CAR=>NO, CAR=>BAL, MIA=>CHI, STL=>DEN (yes, very weird that Carolina has three beatflukes – a very tough team to define)
Loops:
ARI=>SF=>STL=>ARI
ARI=>SF=>OAK=>ARI
PIT=>KC=>SD=>PIT
MIA=>TEN=>HOU=>MIA
BUF=>MIN=>DET=>BUF
BUF=>GB=>DET=>BUF
DAL=>HOU=>JAC=>DAL
HOU=>JAC=>PHI=>HOU
HOU=>JAC=>TEN=>HOU
JAC=>TEN=>WAS=>JAC
CAR=>CLE=>NYJ=>DET=>ATL=>CAR
SEA=>DET=>ATL=>CIN=>KC=>SEA
MIN=>SEA=>NYG=>PHI=>SF=>MIN
SEA=>NYG=>ATL=>CIN=>KC=>SEA
SEA=>NYG=>TB=>CIN=>KC=>SEA
KC=>STL=>DET=>ATL=>CIN=>KC
CIN=>CLE=>NYJ=>DET=>ATL=>CIN
Poor Carolina. 4 wins, but 3 are determined to be flukes, and the other is beatlooped away. At what point do a number of flukes cease to be flukes?
And looking at the graph, Indy has a path to all but 3 teams, CHI, SEA, and DET. That must mean something too, but I’m not sure what.
It is amazing that my Lions managed to “beatloop away” all of their loses but one with a single victory over the Falcons. It makes Detroit look a bit less pathetic in the graph. Though honestly, it is probably more a testament to how inconsistently Atlanta is playing than skill on Detroit’s part.
So the SF=>MIN loss is actually quite costly, as it takes away their win over Seattle.
I think it’s fair to say that Carolina is the first team to be “screwed” by these dispassionate rankings. It’s just a weird fluke that 3/4 of their wins are beatfluked away, but still. The real problem is that 62 yard field goal that allowed Tampa to beat Philly. Take that game away, and the CAR=>TAM game is no longer a fluke, and the Panthers take a position in the graph that looks much more in line with the general perception of their team. Of course, that’s the weakness of a system that relies solely on W/L results to figure out which games were flukes.
MIami beat Chicago and obviously Chicago is still higher on the graph (because of their wins over teams that beat Miami). Can someone tell me what the loop was that cancelled out Miami’s win over Chicago? (i.e. Miami => Chicago => x => Miami, what is x?)
Normally there would have been a beatloop with MIA and CHI. Two of them, actually: MIA=>CHI=>GB=>MIA, and MIA=>CHI=>BUF=>MIA. However, Chicago had an alternate beatpath to Miami through SEA=>OAK=>PIT. So, MIA=>CHI was considered a beatfluke, and the other beatloop segments (CHI=>BUF=>MIA and CHI=>GB=>MIA) were restored to the graph.
So is this how teams can skyrocket with 1 win? Carolina has 3 beatflukes, but if they win the right game or 2, not only would they get credit for those 2 wins, but also the 3 “beatflukes” wouldn’t be flukes anymore, and they would get credited for 5 wins they werent getting credit for over the span of 2 weeks..
Re: 6. Yes, sort of. I’m going to explain the Carolina situation a bit more fully in a new blog entry, coming up in a few minutes.
This is basically like an ecological model, where everything is in the balance, with critical mass points that can make things topple to one side or the other. If the balance point changes a little, then big changes can occur. We witnessed that a lot near the end of the last season with San Diego in particular.
Sometimes there is a not a very big difference between “This is a very good team that is poorly disciplined” and “This is a bad team with flashes of brilliance”. A team can have a couple of good games that can get them over their poor discipline, and all of a sudden beatpaths will see them as a good team. Or, a team’s flashes of brilliance can be overcome by their overall suckitude, and beatpaths will send them down the rankings.
Carolina seems to be kind of like that – a bad team with a couple of superhumans. I almost wonder if it would be good for them to trade away Steve Smith for some great draft picks, so they can relish their suckitude, get more better draft picks, and build their team for real. Right now Smith is good enough to keep them from sucking, but they might not ever get really far, while Smith will keep them from being able to rebuild.
It’s almost impossible to get equal value in a trade involving a player like Julius Peppers or Steve Smith, simply because there are each at the top of their position. Although Carolina is a flawed team, it’s tough to argue that they should trade away and try to re-load. They are in OK shape with the salary cap, and would probably be best served by simply trying to build carefully around their superstars on either side of the ball, while both of them are still in their primes.