All of the numbers below come from Chris Cox at NFL-forecast.com. His app uses the win probabilities from the ANS team efficiency model to run a Monte Carlo simulation of the remaining NFL games thousands of times. Based on current records, our estimates of team strength, and knowledge of the NFL's tie breaking procedures we can come up with some pretty interesting predictions of how each team will fare come the end of the season. If you want to use a different model or just fiddle with the numbers by hand, go ahead and download the app yourself.
Week 12's biggest movers
Capitalizing on Washington's anemic passing performance, San Francisco ended its losing streak. Because Seattle has the West essentially wrapped up, the Cardinals and the loser of the South divisional race are San Francisco's primary rivals at this point. Until recently they also had to worry about the second place team in the north, but the top 3 teams in the division all going winless last week effectively ended that concern. Based on all this, the 49ers jumped 23% to a 67% playoff probability.
Consecutive losses have dropped the Jets from a 43% chance to just 9%. After week 10 they were the favorite to secure the AFC 6th seed. The Dolphins, Steelers, Ravens, Titans, and Chargers all had at least one win in the past two weeks to get to the same 5-6 record. They are all more likely to make it in than the Jets according to the simulations, in part because the model thinks the Jets are the 26th best team in the league.
Pittsburgh has been slowly climbing back into playoff relevance. 3 weeks ago we gave them odds of less than 1%, but a winning streak, greatly improved efficiency numbers, and weak performances from other teams in the middle of the conference all contributed to their current 25% chance.
High leverage game of the week: GB @ DET
By a large margin, the Green Bay-Detroit matchup is the most important game of the week. Neither team has any real hopes of getting in as a wild card, so the divisional title race will likely determine the only team from the North to get in. Both teams had more than a 40% playoff odds swing at stake in this game. Detroit ended up embarrassing Green Bay on both sides of the ball, earning 561 yards from scrimmage to the Packers' 126 (the largest such differential this year). With their 30 point win, the Lions jumped to 72% likely to make the playoffs and the Packers fell to 15%. Before the Rodgers injury we had Green Bay's odds at 83%.
Arizona and Philadelphia, for different reasons, have a lot on the line in their meeting. While the Eagles are in a tight race divisional race that they must win, the Cards know they are likely going to have to beat out the Panthers or 49ers for a wild card game. 25+% probability swings are looming over both.
The Numbers
These values do not account for Thanksgiving's game results.
AFC EAST |
Team | 1st | 2nd | 3rd | 4th |
NE | 97 | 2 | 0 | 0 |
NYJ | 0 | 38 | 34 | 28 |
MIA | 2 | 40 | 29 | 29 |
BUF | 1 | 20 | 36 | 43 |
AFC NORTH |
Team | 1st | 2nd | 3rd | 4th |
CIN | 93 | 6 | 1 | 0 |
PIT | 5 | 57 | 21 | 17 |
BAL | 2 | 21 | 42 | 35 |
CLE | 0 | 16 | 36 | 48 |
AFC SOUTH |
Team | 1st | 2nd | 3rd | 4th |
IND | 97 | 3 | 0 | 0 |
TEN | 3 | 89 | 8 | 1 |
HOU | 0 | 7 | 57 | 37 |
JAC | 0 | 2 | 36 | 63 |
AFC WEST |
Team | 1st | 2nd | 3rd | 4th |
DEN | 90 | 10 | 0 | 0 |
KC | 10 | 87 | 2 | 0 |
SD | 0 | 3 | 86 | 11 |
OAK | 0 | 0 | 11 | 89 |
NFC EAST |
Team | 1st | 2nd | 3rd | 4th |
PHI | 66 | 32 | 3 | 0 |
DAL | 34 | 60 | 6 | 0 |
NYG | 0 | 7 | 72 | 21 |
WAS | 0 | 1 | 20 | 79 |
NFC NORTH |
Team | 1st | 2nd | 3rd | 4th |
DET | 47 | 35 | 17 | 0 |
GB | 36 | 31 | 33 | 0 |
CHI | 17 | 34 | 48 | 1 |
MIN | 0 | 0 | 1 | 99 |
NFC SOUTH |
Team | 1st | 2nd | 3rd | 4th |
NO | 90 | 10 | 0 | 0 |
CAR | 10 | 90 | 0 | 0 |
TB | 0 | 0 | 62 | 38 |
ATL | 0 | 0 | 38 | 62 |
NFC WEST |
Team | 1st | 2nd | 3rd | 4th |
SEA | 98 | 2 | 0 | 0 |
SF | 2 | 69 | 28 | 1 |
ARI | 0 | 28 | 68 | 3 |
STL | 0 | 0 | 4 | 96 |
AFC Percent Probability Playoff Seeding |
Team | 1st | 2nd | 3rd | 4th | 5th | 6th | Total |
DEN | 72 | 14 | 3 | 0 | 10 | 0 | 100 |
NE | 14 | 33 | 33 | 18 | 0 | 1 | 99 |
CIN | 5 | 35 | 37 | 16 | 0 | 3 | 96 |
KC | 8 | 2 | 0 | 0 | 85 | 4 | 99 |
IND | 2 | 16 | 23 | 57 | 0 | 0 | 98 |
MIA | 0 | 0 | 1 | 0 | 1 | 22 | 25 |
PIT | 0 | 0 | 1 | 4 | 1 | 19 | 25 |
BAL | 0 | 0 | 1 | 1 | 0 | 8 | 10 |
TEN | 0 | 0 | 0 | 2 | 0 | 11 | 14 |
NYJ | 0 | 0 | 0 | 0 | 0 | 9 | 9 |
SD | 0 | 0 | 0 | 0 | 2 | 15 | 17 |
BUF | 0 | 0 | 0 | 1 | 0 | 2 | 3 |
CLE | 0 | 0 | 0 | 0 | 0 | 4 | 5 |
HOU | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
OAK | 0 | 0 | 0 | 0 | 0 | 1 | 1 |
JAC | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
NFC Percent Probability Playoff Seeding |
Team | 1st | 2nd | 3rd | 4th | 5th | 6th | Total |
SEA | 65 | 32 | 0 | 0 | 2 | 0 | 100 |
NO | 34 | 55 | 1 | 0 | 9 | 1 | 100 |
CAR | 1 | 9 | 0 | 0 | 51 | 25 | 86 |
PHI | 0 | 1 | 41 | 23 | 1 | 4 | 70 |
DET | 0 | 0 | 23 | 24 | 0 | 2 | 49 |
SF | 0 | 2 | 0 | 0 | 20 | 45 | 67 |
GB | 0 | 0 | 15 | 21 | 0 | 1 | 37 |
DAL | 0 | 0 | 10 | 24 | 0 | 2 | 37 |
CHI | 0 | 0 | 9 | 8 | 0 | 1 | 18 |
ARI | 0 | 0 | 0 | 0 | 17 | 19 | 36 |
NYG | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
WAS | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
STL | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
TB | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
MIN | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
ATL | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
I fail to see how Philly can be favored so decisively for the NFC East (even before Dallas won today). In order for then to not need to win in Dallas in week 17 they need to be 2 games ahead going in, which seems unlikely. Even accounting for rating Philly better than Dallas there is no way that week 17 match is anything like a 66/34 proposition in their favor.
I bet this model doesn't take into account the possibility that in the last week or two of the season, some teams may have their playoff seeds locked up and therefore won't play at 100% (because of resting players they wouldn't otherwise). For instance, if the Seahawks beat the Saints and then have #1 seed wrapped up going into Week 16, would they still be just as strong against the Cardinals? I bet the model assumes they will.
"there is no way that week 17 match is anything like a 66/34 proposition in their favor."
The ANS win probability model disagrees with you: http://www.advancednflstats.com/2013/11/team-efficiency-rankings-week-12.html
It thinks the Eagles are the 4th best team in the league (GWP = 0.66) where the Cowboys are the 22nd (GWP = 0.43). Look at the efficiency numbers in the same link. The Eagles have massively better pass Y/A and run success rate numbers, and everything else is comparable.
J.D., you are right. The model does not account for "throwaway games." The only inputs into the equations are team efficiency numbers - nothing about current week number, playoff seeding, etc. is considered.
Rob,
I disagree with your argument here. Using the model for a reference here is not very convincing. The model assumes a .66 win probability given any random opponent. If you were referring numbers in terms of DAL vs PHI you might have something to say ... I am totally skeptical of the 65/35 argument for PHI vs DAL, this makes the model lose credibility in my opinion.
Toby - The model says the Eagles have a 66% probability against an average team (on a neutral field) and Dallas is a below average team. If you go through the math, the model says that the Eagles would have more than a 70% chance against Dallas on a neutral field. Because the game is in Dallas, the odds are more like 64%.
Pointing to a random game and saying "I am skeptical about the model's output" isn't much of an argument. The Eagles have better efficiency numbers in several important categories. Those efficiency numbers are the basis of the model. If you want to argue against the model's probabilities for that particular game, you are probably going to have to point out systematic issues with Brian's logistic regression methodology.