And the Committee has spoken!

Originally, I wanted to talk about the surprise teams of the 2017 college football year: Iowa State, Boston College, Arizona, even Austin Peay and Elon in the FCS. But the elephant in the room was too large.

The first College Football Playoff rankings came out Tuesday night (Wednesday, Australian time). Wisely, the 13-member committee of experts (how sarcastically that term is used depends on your opinion of the process; I think it’s as good as we can expect) do not issue any preliminary rankings until a few weeks before the actual need for a top four and a top twelve, so as not to waste our time or theirs. As expected, their top four teams as of week 9 are (drum roll, please)…

…the Georgia Bulldogs, Alabama Crimson Tide, Notre Dame Fighting Irish, and the Clemson Tigers. (You could hardly ask for a better, more varied set of mascots than that, could you? Maybe if we had the TCU Horned Frogs in there, or the Ohio State Buckeyes…)

In the ELO-Following Football rating system, there are five teams who stand head-and-shoulders above the others, and they particularly show up when you remove the pre-season ranking bias that we and every other system incorporates in September at least. [I’ve found that maintaining the pre-season portion of the rating improves prediction results over the adjusted scores, which is why I only use the adjusted score for demonstrations like these.] And one of those teams named today isn't among them, despite their on-field credentials. When you make that adjustment in our numbers, here are the top five teams:

Alabama (adjusted score: 0.) ELO-FF rating: 5, 1st place

Georgia (adj score: 1.)          ELO-FF rating: 11, 4th place

Ohio State (adj score; 1.) ELO-FF rating: 7, 2nd place

Penn State (adj score: 2.) ELO-FF rating: 9, 3rd place

Notre Dame (adj score: 2.)          ELO-FF rating: 14, tie for 6th

Then, there’s a significant gap to the next team, Oklahoma State (adjusted score of 5, fifth place ELO-FF rating of 12), and an even bigger gap to seventh place and beyond (with adjusted scores of 9 to 13) – Iowa State, Mississippi State, TCU, Auburn, Central Florida, Virginia Tech, Stanford, THEN Clemson, Wisconsin, and the University of Washington. (Strictly speaking, the next lowest score belongs to defending FCS champion James Madison at 14, with North Dakota State close behind them at 16!)

If you’re serious about picking out the four best teams for the playoff at the end of the year, that’s as far down as we need to go when we’re two-thirds of the way through the season. (Seriously: do you think Arizona or Texas are getting in? Those are teams #17 and #18 on our list right now.)

As for Clemson, undoubtedly the reasoning the committee used was that their one loss, as bad as it was (27-24 to lowly Syracuse), came on a short week, in a road game, when their starting quarterback went down early in that game. But the ELO-FF system ignores that for the simple reason that if you're evaluating teams strictly on their on-field performance, as they claim to be (for example, the committee famously does NOT account for games yet to be played - see Baylor/TCU three weeks ago). Clemson played that game, and they lost. The teams the one-loss crew of ours failed against - Notre Dame lost to Georgia by one; Penn State to Ohio State by one, and Ohio State to 7-1 Oklahoma - all fall in the CFP top six. (OU was 5th, OSU 6th, and Penn St 7th). Syracuse, not surprisingly, is not ranked by the committee, despite their "signature win".

We still believe that those five teams that have separated themselves from the peloton – two SEC schools, two from the Big Ten that just played a phenomenal, 39-38 thriller won by OSU in the final minutes, and the independent powerhouse Notre Dame – are basically guaranteed to have first say in the make-up of the CFP foursome. The winner of the SEC championship game on December 2 – and if they’re both still undefeated going in, very possibly the loser as well – will be there. Ohio State now has a clear run to their finals spot if they beat Michigan in a few weeks, and then win the conference title game (most likely against Wisconsin). Both Penn State and Wisconsin will have something to say if they hiccup, but it’s now the Buckeyes’ spot to lose. And a 12-1 Power Five conference champ WILL be in the finals.

As for Notre Dame, their independent status is advantageous only when it allows them to schedule well – and this year, it has. Besides the one-point loss to Georgia in one of the games-of-the-year, they’ve wiped up ACC Atlantic leader North Carolina State by 21, resurgent Boston College by 29, and Pac-12 South leader USC by 35 points! Their scheduled “lightweights” aren’t that light, either: Miami-Ohio (top tier team in the MAC), equally resurgent Michigan State, and two teams expected to be better than they have been, Temple and North Carolina. If they finish 11 and 1, they’ll also have beaten a strong Naval Academy club, a Wake Forest team we have in our top 25, #18 Stanford and #9 Miami of Florida. No other team has that kind of “Murderer’s Row” to battle through, so if the Irish succeed, they’ve written their own ticket to the CFP.

Do any other teams have a chance? Of course.

Even if these five teams hold up their end of the bargain, you’ve only got two conference champs in the bunch. It would be easy for the committee to “reserve” a spot beyond Notre Dame and the SEC and Big Ten champs (and off course, if more losses creep into their resumes, more spots may open up): Justify Penn State’s absence as having been the victim of their unofficial “elimination game” this past weekend, and remove the loser of the Alabama-Georgia SEC title game the same way.

So who’s most likely to fill that spot? Well, Clemson will evidently get first ups, according to the CFP committee.

The ACC champ will certainly have an argument, especially if it IS defending champion Clemson. But even an undefeated (or even a one-loss) Miami of Florida, or possibly a one-loss Va-Tech, will get serious consideration from the committee - as we said before, a 12-1 Power Five champ will probably get into the final four.

Perhaps the carnage in the Big Twelve will ease up these last four weeks, and one of TCU, Iowa State, Oklahoma or Ok-State will end up with just the one conference loss, a victory in the nonsensical title game and thus a legitimate argument for a spot at the finals table. But even with Oklahoma at five right now, it's hard to imagine them going through OSU, TCU, and then one of those three top teams again unscathed.

And in the land of make-believe, there’s also the possibility that the CFP will forget that the one-loss Washington Huskies were the victims of perhaps the ugliest of the losses any of that category of teams suffered, or that carnage will ensue and either the two-loss Stanford Cardinals, the two-loss Southern Cal Trojans, or the remaining AAC-unbeaten, Central Florida, will emerge as a viable option when December 3rd comes around. More likely, those teams will be hoping for a seat at the New Year’s Six banquet, rather than a backstage pass to the playoffs.

But if I were a betting man (and I’m not), the top four on the initial CFP ranking is also the way I’d bet for the top four to look in five weeks, when it matters.

Let’s take a look at this week’s top games:

There are two huge games in the ACC Saturday: Clemson travels to North Carolina State to try and re-take control of the Atlantic division, while Miami of Florida hosts Virginia Tech and hopes to hold off the Hokies’ attempt to do the same in the Coastal division. While Vegas has Clemson as a 7.5 point favorite and Miami as a three-point underdog, we see Clemson as just a three-point favorite, and rate the Miami/VT game as a toss-up. (Georgia Tech at Virginia may also be a classic, although we favor the “ramblin’ wreck” by 12.)

Nothing in the Big Ten rivals the Ohio St/Penn St heavyweight clash from last Saturday night, but Penn State does have to host Michigan State, and despite being touchdown favorites (we have it more like two scores), PSU could be in danger of a letdown game against a good Spartans team. The other interesting games are Ohio State waving to the children’s hospital patients at Iowa (and winning by about 15), and Northwestern playing at Nebraska (nobody knows what to make of Nebraska this season: the spreads range from one point Nebraska’s way to four points the Wildcats’ way. We like Northwestern by two).

Bedlam reigns in the Big Twelve this weekend – Oklahoma plays at Oklahoma State. For once, the Sooners are underdogs: the home team is favored by three in Vegas, and by us by seven points. But each is 4-1 and has company at the top from TCU and surprising Iowa State, which has beaten both TCU (14-7 this weekend) and OU (38-31 earlier this season). They have a tough game at West Virginia (we see WVU by two), while TCU hosts Texas (the Frogs are a five-point favorite everywhere).

Stanford goes to Washington State, and the loser’s third loss eliminates them from title consideration completely (we have Stanford by five; Vegas has WSU by two!). In the Pac-12 South, Arizona has to go to the Coliseum and must play as a touchdown underdog to USC in their climb to conquer the division.

Alabama has the first of its big tests in the SEC, hosting LSU this Saturday as twenty-point favorites, while Georgia hosts South Carolina at the same spread. The most interesting game may be watching how Florida reacts to the sudden ousting of its coach following a bizarre disagreement between he and the administration regarding death threats. Florida plays at Missouri Saturday, and while Vegas has them as field goal underdogs, it’s hard to fathom how to predict the way interim coach Randy Shannon and the Gators team reacts this weekend.

In the Group of Five, we’ll be keeping an eye on the Conference-USA matchup between two rising teams in the east Friday night. Marshall goes to Florida Atlantic as a touchdown underdog. In the west, North Texas’ run to the conference title game will have to go through Louisiana Tech, which is between a three and ten-point favorite (we have it at ten) at home.

As for the FCS, North Dakota State faces perhaps its toughest challenge of the home-and-away season, travelling to South Dakota State as just three-point favorites. They’ll finish out the season against the two teams directly beneath them in the Missouri Valley Conference, South Dakota and then Illinois State. Their major rival nationally, 2016 national champ James Madison, has a fairly easy game at Rhode Island this weekend (they’re favored by 26), but then finishes their season against traditionally tough Richmond and 2017 upstart Elon as they seek the Colonial Athletic Association crown.

We went 6-3 in the games we picked in this space last week, and 76-33 overall for the week. For the year, we’re batting 76.5% in our overall picks, and compared to the Vegas lines and the Sagarin system predictions over all FBS games this season (Vegas traditionally doesn’t publish lines for FCS games), we have enjoyed a margin of accuracy over the two of 175 to 164 for Vegas and 152 for Sagarin. (Past performance does not guarantee future accuracy, however, and we don't recommend using anything written here for monetary purposes.)

Week 8 in College Football

[originally published in The Roar, Oct 24 2017]

Welcome to our “Following Football” weekly review of American collegiate football! First of all, let’s take a look at the predictions I made last week in this space. Of the 20 significant games we picked, 15 went to the team our “ELO-Following Football” rating system forecast, with Louisville’s last second victory over Florida State being our highest profile miss. Fresno State has become a force in the Mountain West a year after not winning a game, waxing league favorite San Diego State 27-3 Saturday night (we had SDSU by at least a touchdown). South Dakota was upset by Illinois State, and we went 0-2 in the Ivy League, missing on Princeton’s 52-17 beatdown of Harvard and Columbia going 6-0 for the first time in decades by upsetting undefeated Dartmouth 22-17.

In the other 15 games, we had varying degrees of accuracy, as is always the case with punting. Some games we nailed the team but underestimated the margin drastically, such as Notre Dame’s 49-14 domination of Southern Cal or Florida Atlantic going 69-31 on upstart North Texas. But there were games where we hit the nail on the head - in nine of the games, in fact, we came within one score of predicting the exact margin of victory.

In a three way comparison of the consensus Las Vegas oddsmakers, the leading computer predictor (Sagarin), and our ELO-FF rating system, we not only outperformed the other two this weekend (24 where we were closest, versus 19 for Vegas and 16 for Sagarin) but have done so over the course of the season as well (154-150-130).

So let’s look at the AP top 20 and how the ELO-FF system sees the teams moving forward:

1. Alabama (rating of 5 - 1st) 8-0, 5-0 SEC. The current consensus best team in football, and everyone’s favorite for the national title as of this moment. 45-7 winners over Tennessee this weekend.
2. Penn State (rating of 10 - 3rd) 7-0, 4-0 Big Ten. We are equally impressed with the Nittany Lions, but their division rivals are playing as well as anyone in the country right now.
3. Georgia (rating of 14 - 5th) 7-0, 4-0 SEC. Doing their best to squash the notion of Alabama and the 13 dwarves in the SEC. Will get the chance December 2.
4. TCU (rating of 17 - 13th) 7-0, 4-0 Big Twelve. We have the two Oklahoma schools rated ahead of the Horned Frogs, but we have to admit TCU looked awfully good Saturday, holding Kansas to just 21 yards in total offense, the lowest total in two decades in the FBS.
5. Wisconsin (rating of 16 - 11th) 7-0, 4-0 BIg Ten. And this isn’t even the rival we had in mind. UW has a schedule where they really should go 12-0 on their way to the conference title game.
6. Ohio State (rating of 6 - 2nd) 6-1, 4-0 Big Ten. It’s the Buckeyes who suffered a bad loss to #10 Oklahoma and have looked unbeatable ever since. We have them as the closest challenger to Alabama and a favorite against Penn State next Saturday.
7. Clemson (rating of 14 - 5th) 6-1, 4-1 ACC. Still the favorite in the conference, and unless there are four undefeated teams left in December, winning the ACC should get them back to the National Playoff.
8. Miami-FL (rating of 19 - 15th) 6-0, 4-0 ACC. An example of an undefeated record hiding blemishes that playing Clemson in a conference title game will uncover down the road. But human polls always prefer gaudy records over actual strength of schedule evaluations, which is why Ohio State is only 6th.
9. Notre Dame (rating of 15 - 9th) 6-1 independent. Demolished Southern Cal this weekend 49-14 and has gotten stronger (and dropped its rating number) every week. Pretty good for a team that won four games last year and thought about firing its coach.
10. Oklahoma (rating of 16 - 11th) 3-1, 6-1 Big Twelve.
11. Oklahoma State (rating of 12 - 4th) 3-1, 6-1 Big Twelve. The loser of their November 4th game is out of luck; the winner can still play for all the marbles with a bit of luck. OU plays TCU on 11 November, so they control their destiny; OSU has one of the top ten strength of schedule records and can still be top two and in the conference title game at 8-1 in conference.
12. Washington (rating of 15 - 9th) 6-1, 3-1 Pac-12. The Pac-12 will probably be left out of the College Football Playoff, with only the two Washington schools under two losses in their cannibalistic conference. 11 of the 12 teams in the conference have sub-30 ratings, placing them in the top 50 teams in the country.
13. Virginia Tech (rating 17 - 13th) 6-1, 2-1 ACC.
14. North Carolina State (rating 22 - 21st) 6-1, 4-0 ACC.
15. Washington State (rating 21 - 19th) 7-1, 4-1 Pac-12.
16. Michigan State (rating 29 - 43rd) 6-1, 4-0 Big Ten.
17. South Florida and Central Florida (both rating 22 - 21st) Both undefeated, both 4-0 in the AAC.
18. The two directional Floridas are far and away the leaders in the Group of Five race for the one New Year’s Six berth, both in the human polls and in the ELO-FF ratings - the closest club in the ratings is Boise State at 28, and Memphis at #24 is the only other AP top 25 club from outside the Power Five conferences.
19. Auburn (rating 14 - 5th) 6-2, 4-1 SEC. The highest two-loss team, and while that will probably knock them out of the CFP, they still control their own destiny in conference when they go to Alabama on November 25.
20. Stanford (rating 14 - 5th) 5-2, 4-1 Pac-12. The highest rated and often most impressive team out west, their rout against USC will probably keep them out of most of their desired goals.

Teams in the ELO-FF top 20 that aren’t in the AP list include Florida State (2-4, 2-3 ACC, rating of 19 - 15th); Southern Cal (6-2, 4-1 Pac-12, rating of 20 - 17th); LSU (6-2, 3-1 SEC, rating of 20 - 17th); and Mississippi State (5-2, 2-2 SEC, rating of 21 - 19th).

On the bottom end of the FBS, starting from the highest rating on the ELO-FF scale:

130. Texas State (rating of 53, 1-6, 0-3 Sun Belt).

126 (Tie). Charlotte (rating of 52, 1-7, 1-3 C-USA); Rice (52, 1-6, 1-2 C-USA); UTEP (52, 0-7, 0-3 C-USA), and San Jose State (52, 1-7, 0-4 Mountain West).

125. Kent State (rating of 51, 2-6, 1-3 MAC).

122 (Tie) Ball State (rating of 50, 2-5, 0-3 MAC); Georgia Southern (50, 0-6, 0-3 Sun Belt), and Coastal Carolina (50, 1-6, 0-4 Sun Belt).

121. Bowling Green (rating of 49, 1-7, 1-3 MAC).

Behind those ten schools, sit five more with a rating of 47, including Kansas, the lowest-rated Power Five school by six full rating points.

Meanwhile, in the FCS, the ratings tell a very simple story. Two schools, the last two FCS national champions, are tied with a phenomenally low rating of 24 - defender James Madison (7-0, 4-0 Colonial) and five-time champ North Dakota State (7-0, 4-0 Missouri Valley). The next closest team to those two is Ohio Valley leader Jacksonville State (6-1, 4-0) with a rating of 33. Other teams in the low to mid 30s in ratings include six Missouri Valley clubs (the best are South Dakota State and Youngstown State, both at 34) and three in the Southland (led by Central Arkansas at 35). The Big Sky, Southern conference, MEAC and Colonial contribute ten more clubs with ratings of 38 or 39.

The highest ratings in the FCS are Mississippi Valley State (70), Davidson (69), Arkansas-Pine Bluff (68), and then Stetson and Morehead State (both 67), which played in Pioneer League action Saturday; Morehead came from behind to win at home by three, 29-26.

⦽

What about Week 9? What does our system say about key games, and how does that compare to the “professional” oddsmakers?

The biggest game of the week, maybe the season so far, would be in the Big Ten: Penn State @ Ohio State. Vegas has OSU as a six point favorite; we think it should be more like nine.

In the ACC, Clemson hosts Georgia Tech, and is either a 12 (ELO-FF) or 14 (Vegas) point favorite, depending on your point of view.

The most relevant game in the Big Twelve this week is Oklahoma State travelling to West Virginia. We agree on the 7 point line in OSU’s favor there. We disagree, though, on the line for the “World’s Biggest Cocktail Party”, the annual neutral-site game between SEC powers Florida and Georgia. As dominant as Georgia is, we don’t see more than a nine-point spread in their favor; Vegas gives them 14.

The leading powers in each AAC division square off Saturday when Houston goes to South Florida. We have a 12-point gap in USF’s favor; the oddsmakers give them 10.5 instead.

In Conference-USA, the East may be decided Saturday when Florida Atlantic takes its unbeaten conference record to defending champion Western Kentucky, whom we favor by one. Vegas disagrees vehemently, picking FAU by 7.5 points.

There are two critical games in the FCS Southland conference this Saturday. McNeese State travels to conference leader Central Arkansas, whom we have as a four-point favorite (Vegas only publishes odds on FBS games), and pre-season favorite Sam Houston State has the unusual feeling of being a one-point underdog at Southeast Louisiana Saturday night.

Finally, Columbia (fresh off an upset at Dartmouth) has to defend its undefeated record by going to traditional Ivy League powerhouse Yale, which we have as an eight-point preference.