Just for fun, I revisited Football Outsiders’ advanced statistical analytics one more time before the College Football Playoff. Some interesting things happened in the F/+ Ratings as a result of FBS bowl Season, but they didn’t really change the CFP landscape much. The Top 4 stayed intact with ‘Bama remaining #1, Clemson #2, and tOSU and OU swapping spots at #3 and #4. There were several minor changes of one to four spots up and down throughout the Top 25, but only a couple jumps/plummets of five or more places. What the bowl games gave us more than anything was an opportunity to see the effects of OOC play, and a chance to validate the CFP committee’s rankings versus the F/+ ratings. The result unquestionably validated F/+ over the eyeball test.

Observations from Bowl Season F/+ data

1) And Then There Were Two.  ‘Bama’s throttling of Sparty through the air pushed the Tide’s F/+ to 70%; the last two teams to see that were Clemson in Week 8 this season and tOSU after the NCG last year. Clemson at #2 improved its F/+ by 1.9% after beating OU, and now sits at 60.9%. The Buckeyes improved 3.6% to 54.7% at #3, and Oklahoma’s F/+ slipped for the second straight game to 50% at #4. No other team in FBS broke 50% after the bowl games, which tells me ‘Bama, Clemson, tOSU and OU were the four best teams in FBS. Interestingly, in terms of S&P+, these four teams are the most balanced in overall Offense and Defense S&P+ ranking in the Top 25 in both categories. Only Ole Miss and FSU achieved this as well. The fact that ‘Bama and Clemson are clearly ahead of the Buckeyes leaves me confident they deserve to be in the NCG.

2) A Tale of Two Seasons. It is not insignificant that from Week 6 (Maryland) to the end of the bowl games, the Buckeyes’ F/+ improved from 34.2% to 54.7%: a 20-point increase. Stated another way, beginning with the PSU “Blackout” game, tOSU made the largest increase in F/+ percentage of any team in the Top 10. Only two other teams in the final F/+ Top 25 saw larger improvements over the same period: BGSU (21.9%) and Oregon (21.7%). Maybe it had something to do with the Buckeyes playing #not our rival, perhaps it was the black uniforms, or maybe it was a personnel change on offense. Draw your own conclusions.

3) Pride and Prejudice. (DISCLAIMER: I hate the SEC. I really do.) The SEC is a very good conference in football, especially across its middle teams. The top three teams in the B1G and SEC are outstanding: average F/+ for top 3 SEC is 52.63%, for the B1G it’s 45.23%. The bottom four teams in both conferences are terrible: average F/+ for last 4 SEC is -10.0%; for the B1G it’s -12.7%. But the teams in the middle of the SEC are superior to their B1G counterparts without question. The average F/+ for 7 SEC team in the middle of the conference (Ark, Miss St., Tenn, UF, UGA, Auburn, A&M) is 27.24%. For the B1G teams (Wisc, NU, Iowa, PSU, Northwestern, Minn, IU) the average F/+ is 13.63%. The competition across the middle of the SEC is tougher. It’s nearly impossible to argue the contrary.

4) Fahrenheit 451. (Disclaimer: See 3 above) Whatever your personal opinion is on relative importance of Power Five conference strength, the reality is that perceptions matter. I refuse to give in to conspiracy theories about media bias and the like. I do, however, dislike the human polls because of the cognitive dissonance, capriciousness and gross bias of some voters. If you disagree, think about this: 11 of the AP’s Preseason Top 25 finished outside the Top 25 (nearly 50%), and six (25%) received zero votes in Week 15 meaning they didn’t even crack the Top 40. Of those 11 teams, five were from the SEC, four were PAC 12, one ACC and one Mountain West. As a left-brain, analytical thinker, I prefer to look at the statistical data regarding conference strength.

Keep in mind, F/+ combines FEI and S&P+ data. It looks at every snap, down-and-distance, efficiency, etc., and it throws out garbage time when the game is no longer in question as well as end of half clock kills. As a conference, the SEC in 2015 was clearly the best conference in advanced stats. The worst teams in F/+ from each of the Power Five conferences were: #127 Kansas (0-12), # 107 Oregon State (2-10), #101 Rutgers (4-8), #92 Wake Forest (3-9), and #91 Kentucky (5-7).

5) CFP Top 25. My bowl picks went 7-4 (OU, ‘Bama, Stanford, tOSU, FSU, UNC, Oregon, Ole Miss, Tenn, TTUN, Navy). 33-17 for the season ain’t too shabby. Three of my four L’s this week were upset picks against F/+: OU, UNC, and Oregon were predicted losses according to F/+. The only true upset (by F/+) in the Top 25 bowl game matchups was Tom Herman’s Houston Cougars. IMHO, Houston’s weak competition early in the season artificially depressed their F/+ numbers. That, and Ward is a baller.

6) National Championship Game. This should be a great game.  #1 ‘Bama vs. #2 Clemson is exactly how they stand in F/+, S&P+ and FEI. I prefer the individual breakdowns in S&P+, because it gives more detail. According to that, ‘Bama is about eight points better on defense (6.8 vs. 14.9), and Clemson is just under five points better on offense (41.6 to 36.8). That should make the Tide about a three-point favorite in this game. What makes this matchup interesting, is that Clemson’s offense is nearly a statistical match to Ole Miss in S&P+ across the major categories. Add to that, Clemson has a better defense than Ole Miss (which forced five ‘Bama turnovers) in terms of S&P+ against the rush but more specifically against the pass. I’m not a fan of either team, but I like Swinney a lot more than I do Saban. And I have no interest in hearing the endless claims that Saban in the best FBS coach in history, because I find it far too early to tell and a patently flawed opinion. So, I’m going with advanced analytics my opinion/gut and picking Clemson.  Go Tigers!

Previous weeks:
Week 1  Week 2  Week 3  Week 4  Week 5  Week 6  Week 7  Week 8  Week 9  Week 10  Week 11  Week 12  Week 13  Week 14