| Rank | Team | Agg Rank | POFF Prob | Change |
| 1 | Clemson | 2 | 85.47% | 2.1% |
| 2 | Alabama | 1 | 79.01% | 12.1% |
| 3 | Notre Dame | 9 | 56.28% | 3.9% |
| 4 | Ohio State | 3 | 48.78% | -2.5% |
| 5 | Oklahoma | 6 | 22.61% | 11.7% |
| 6 | Georgia | 4 | 19.14% | -32.6% |
| 7 | Michigan | 5 | 17.97% | 10.5% |
| 8 | North Carolina State | 23 | 17.15% | 7.7% |
| 9 | Florida | 11 | 12.90% | 7.4% |
| 10 | LSU | 10 | 12.12% | 10.1% |
| 11 | Iowa | 13 | 3.86% | 2.4% |
| 12 | Texas | 19 | 3.39% | 1.5% |
| 13 | Washington | 8 | 3.05% | -11.5% |
| 14 | Kentucky | 24 | 2.97% | 1.0% |
| 15 | West Virginia | 18 | 2.71% | -8.6% |
| 16 | Penn State | 7 | 2.43% | -18.0% |
| 17 | Texas A&M | 15 | 2.40% | 1.7% |
| 18 | Duke | 28 | 2.23% | 1.7% |
| 19 | Utah | 17 | 1.06% | 0.8% |
| 20 | Washington State | 30 | 1.01% | 0.2% |
| 21 | Central Florida | 25 | 1.01% | 0.6% |
| 22 | Oregon | 33 | 0.40% | 0.3% |
| 23 | Stanford | 27 | 0.39% | 0.2% |
| 24 | Mississippi State | 12 | 0.36% | 0.2% |
| 25 | Miami (FL) | 16 | 0.32% | -2.6% |
| 26 | Texas Tech | 29 | 0.24% | 0.2% |
| 27 | Colorado | 51 | 0.15% | -0.1% |
| 28 | Michigan State | 22 | 0.10% | 0.1% |
| 29 | Syracuse | 44 | 0.10% | 0.1% |
| 30 | Cincinnati | 48 | 0.09% | 0.1% |
| 31 | USC | 38 | 0.08% | 0.1% |
| 32 | Wisconsin | 14 | 0.08% | -0.3% |
| 33 | Boston College | 41 | 0.07% | 0.1% |
| 34 | Virginia Tech | 42 | 0.06% | 0.0% |
| 35 | South Florida | 59 | 0.01% | 0.0% |
| 36 | Virginia | 53 | 0.01% | 0.0% |
| 37 | Maryland | 47 | 0.00% | 0.0% |
1. Our second straight week of goofy upsets clarifies the top four, which you could have probably guessed without clicking through to my blog. Ohio State is a little behind everyone else (largely because of the eerie similarities to 2015 Ohio State), but the top four shouldn't engender much controversy. The thing you might not have guessed is that an idle Oklahoma team was one of the biggest risers of the week, climbing all the way to #5. I did not explicitly give them a bonus for firing Mike Stoops, but I wouldn't be surprised if my computer did that on its own. The Sooners tower above fellow one-loss Big 12 teams Texas and West Virginia not because of an SOS discrepancy (they're fine - all three have roughly average schedule strength for power conference teams), but because they are better and are thus projected to do better over the rest of the season (23% chance of winning out vs. 2% for each of the other two teams). Clearly, my computer is doing this to torture me, because it would be the most A's thing ever for Oklahoma to steal the four-seed, and then have Kyler Murray be literally dismembered by Alabama.
2. Since we're halfway through the season, I wanted to chat for a second about my favorite topic: Chaos. It's still entirely possible that the final Playoff rankings are neat and orderly, but the events of the last two weeks have drawn us much closer to some more chaotic possibilities. And while there have been a few minor selection controversies in the first four years of the CFP, we have yet to see the two specific things that could lead to anarchy:
A conference title game upset - Prior to 2014, major conference title games regularly produced dramatic upsets. 2013 saw Michigan State knock Ohio State out of the top two. Oklahoma lost a stunner to Kansas State in 2003. And the bizarre seasons of 2001 and 2007 were shaped in part by final-week shenanigans. But since the advent of the Playoff, the committee has never had to deal with a top team losing in the final week. There are no obvious candidates for this yet, as we're too far away to predict anything accurately (see #5 below). But if everything holds to form over the next six weeks, this is still something to keep in the back of your mind.
A two-loss team with a very strong strength of schedule - As you might know, my model uses data going to back to the advent of the BCS in 1998. In that timespan, nine teams finished in the top four of the BCS standings with two losses. Excluding the three teams that did so in 2007 (what a year), the common thread between these teams was a very strong SOS. 1999 Alabama (.630), 2002 USC (.627), 2005 Ohio State (.602), and 2001 Colorado (.597) would have all have very strong cases to be included in the Playoff committee's top four, had it existed at the time. While there have been teams with such daunting schedules in the past four years, none of them fell into the two-loss bucket. The ten best SOS among two-loss teams in the Playoff era are as follows:
| Year | Team | SOS |
| 2017 | Ohio State | 0.582 |
| 2015 | Northwestern | 0.565 |
| 2017 | Miami (FL) | 0.563 |
| 2016 | Penn State | 0.561 |
| 2016 | Michigan | 0.559 |
| 2017 | Penn State | 0.558 |
| 2015 | Stanford | 0.555 |
| 2014 | Mississippi State | 0.553 |
| 2015 | Notre Dame | 0.551 |
| 2017 | USC | 0.550 |
2017 OSU is appropriately at the top as they were probably the closest two-loss team to inclusion in that span, just losing out to eventual champion Alabama. But no one else even came close to the stalwart two-loss teams of the BCS era, so there hasn't been much controversy to speak of. In completely unrelated news, this year's Texas A&M team has a projected SOS of .604 and a 6% chance of winning out.
3. A bloody week from top to bottom eliminates ten more teams: South Carolina, Auburn, Oklahoma State, Missouri, TCU, Indiana, California, Minnesota, Baylor, and Illinois. Auburn is the first pre-season contender to bite the dust. Their case was built on their decent odds of weathering a strong schedule (33% chance of 2 or fewer losses against a projected .591 SOS), but they just didn't have the offense to do it. Maybe we'll finally predict them correctly next year.
4. The conference odds finally got a shake-up this week. The Pac 12 is basically dead. The Big 12 takes another significant hit. The benefactor of all of this is...the ACC? Sure, why not?
| Conference | Exp Playoff Teams |
| SEC | 1.289 |
| ACC | 1.054 |
| B10 | 0.732 |
| IND | 0.563 |
| B12 | 0.289 |
| P12 | 0.061 |
| AMER | 0.011 |
5. The conference title grid is low-key bonkers, with a number of divisions still completely up for grabs. In the Big Ten West, Northwestern is a strong third with 16% after escaping against the Huskers. West Virginia (26%) and Texas Tech (22%) still have chances to prevent a Red River Rematch. Stanford (13.4%), Oregon (13.2%), and Wazzu (12.8%) are all comically close in their bid to usurp the Huskies. Virginia (10%!) and Duke (8%) are still alive because the Coastal is trash again. And Kentucky still has a 17% chance of earning the honor of getting destroyed by Bama in Atlanta. Because of the sheer number of weird teams with decent odds of making a title game, we're probably going to get at least one of them in. Cool.
| Conference | Favorite | Perc | Runner-Up | Perc | |
| ACCA | Clemson | 87.5% | North Carolina State | 11.5% | |
| ACCC | Virginia Tech | 47.7% | Miami (FL) | 32.2% | |
| AMEE | Central Florida | 64.8% | Cincinnati | 19.8% | |
| AMEW | Houston | 78.0% | Tulane | 12.1% | |
| B10E | Ohio State | 62.8% | Michigan | 33.8% | |
| B10W | Wisconsin | 46.2% | Iowa | 33.1% | |
| B12 | Texas | 69.7% | Oklahoma | 69.2% | |
| CUSAE | Florida International | 51.1% | Marshall | 23.0% | |
| CUSAW | UAB | 51.0% | North Texas | 32.7% | |
| MACE | Buffalo | 66.6% | Miami (OH) | 21.0% | |
| MACW | Northern Illinois | 44.2% | Western Michigan | 40.4% | |
| MWCW | Fresno State | 82.4% | San Diego State | 16.5% | |
| MWCM | Utah State | 74.2% | Boise State | 25.4% | |
| P12N | Washington | 60.7% | Stanford | 13.4% | |
| P12S | USC | 52.6% | Utah | 30.8% | |
| SECE | Georgia | 55.9% | Florida | 26.6% | |
| SECW | Alabama | 90.9% | LSU | 7.8% | |
| SUNE | Appalachian State | 81.3% | Georgia Southern | 9.9% | |
| SUNW | Arkansas State | 59.1% | Louisiana-Lafayette | 23.6% |
6. One thing this series of posts largely overlooks is quality Group of Five teams that aren't undefeated. Sure, you might see them in the conference title grid, but you probably don't get a good idea of just how good they are. In light of that here's a list of where the top Group of Five teams are ranked in the computer aggregate (And yes, I am mostly just doing this to sing the praises of Appy State):
| Team | Rating |
| Fresno State | 21 |
| Central Florida | 25 |
| Appalachian State | 26 |
| Utah State | 34 |
| Boise State | 36 |
Week 8 Preview
| Home | Away | Home Win Prob | Playoff Teams Lost |
| Clemson | North Carolina State | 86.7% | 0.088 |
| LSU | Mississippi State | 59.5% | 0.027 |
| TCU | Oklahoma | 22.7% | 0.026 |
| Michigan State | Michigan | 27.6% | 0.026 |
| Purdue | Ohio State | 15.6% | 0.025 |
| Tennessee | Alabama | 4.6% | 0.012 |
| Washington | Colorado | 86.0% | 0.005 |
| Iowa | Maryland | 79.0% | 0.004 |
| Utah | USC | 69.8% | 0.004 |
| Washington State | Oregon | 57.7% | 0.003 |
Week 8 is not deep at all, but there's enough excitement at the top to earn your eyeballs. A Wolfpack win in Death Valley would give them an unprecedented leap up the Playoff Odds standings, so that's worth watching for that alone. TCU hopes to salvage their season (and Ohio State's SOS) at home against the rested Sooners. Purdue just needs to roll a 6 to upset the Buckeyes. And a couple of Pac-12 games will go a long way in deciding the division champions.
No comments:
Post a Comment