My Top 25
1 Ohio St. 3
2 Florida St. 1
3 Oregon 4
4 Alabama 2
5 Boise St. 6
6 TCU 5
7 UCLA 11
8 Mich. St. 13
9 Marshall 12
10 Ga. Tech 14
11 Missouri 16
12 Baylor 7
13 Georgia 17
14 Wisconsin 15
15 Ole Miss 8
16 Miss. St. 9
17 Arizona 10
18 Clemson 21
19 Arizona St. 23
20 USC —
21 Auburn 18
22 Colo. St. 20
23 Kansas St. 19
24 N. Illinois 24
25 Memphis —
Out of top 25: (22) Nebraska, (25) Louisville
I guess I’ll start with some good news. I like how the top four stayed in place. If after the Alabama game anyone still believed Ohio St. didn’t belong in the top four, I’m sure they do now. Other than that, pretty much everyone agreed with the other three teams.
Why does the fact that teams won these games mean that suddenly some of them aren’t in the top four? It doesn’t. If we made 5-8 play each other and then the winners of those games play each other, that would leave three teams in that group with one more loss than they had before the bowls as well.
I am also pleased with my rankings before the bowls. Eleven of my top 17 (including Oregon) won their bowl games. All of the top 19 either won their bowl game or lost to another top-19 team. Colorado St. and Nebraska were the only top-25 teams to lose bowl games to other teams.
Obviously, some people will criticize where TCU is, but only one team they played all year won a bowl game, and the Horned Frogs don’t get a lot of points from Oklahoma St.’s win over Washington. The margin of victory over Ole Miss doesn’t help in my ratings, but like most fans, I would have liked to have seen TCU play one of the top four. (I still don’t want another round of playoffs though, at least not without a major overhaul.)
I know Boise St. lost to Ole Miss, but losing to Baylor doesn’t hurt that much less. Boise St. also played an extra game. The Broncos were one of four 10-win teams in the Mountain West (all in the same division), while only two teams in the Big XII won 10 or more.
This might be boring, but it might also be a way of explaining my rankings, so bear with me. On 11/23, TCU led Boise St. by about 0.12 (which was the approximate distance from Ohio St. to Oregon before the bowls). Boise St. won three games between 11/23 and the bowls (Wyoming, Utah St., and Fresno St.) while TCU only won two (Texas and Iowa St.). This cut TCU’s lead to 0.006.
All things being equal, the TCU lead should have increased to 0.017 because Ole Miss counts for more points than Arizona does. So the 1-6 performance in bowl games by TCU opponents cost the Horned Frogs 0.021 by comparison since they ended up 0.004 behind the Broncos. Only three of Boise St.’s, but of the 13 games they played, 10 were against teams that went to bowl games (they played Fresno St. twice, Ole Miss, Colorado St., ULL, Air Force, Nevada, BYU, San Diego St., and Utah St.).
For a non-mathematical argument, a lot of people thought I had Boise St. way too high and they shouldn’t have been ahead of Arizona. I think they corrected that perception.
This is the fifth time since 2008 (when I started my current system) that Boise St. has finished in the top 6. By comparison, this is only Alabama’s fourth time in that time frame. Florida, Ohio St., and Oregon have finished in that group three times apiece. Not apropos of anything, but every time Ohio St. has been in that group, so has Oregon.
I would have liked to have seen Marshall play a P5 team, but they did about as well against then-11-2 Northern Illinois as Arkansas had back in September.
I do think Georgia Tech would have likely beaten them, but remember the point of my formula is so that teams with few losses end up near the top. Georgia Tech obviously had a much better schedule, but you have to have a much better schedule to account for two extra losses.
I’m sure most of you remember when LSU had two losses and made the BCS Championship ahead of one-loss Kansas. That was a huge difference in schedule, but if LSU had three losses or Kansas had no losses, LSU would not have gone ahead of Kansas. I first started working on my current formula after the 2007 season, so that’s probably something I considered when I decided how much winning percentage counts and how much strength of schedule counts. I also think LSU would have beaten Kansas even if the Jayhawks had gone undefeated in 2007, but that doesn’t mean LSU would have deserved to play in the game ahead of them.
Anyway, Marshall is in the CUSA, which went 4-1 in the bowls. Compare that to the 4-7 mark of the ACC. Also, bad losses hurt more in my formula than good wins help. The one team that beat Marshall (Western Kentucky) won its bowl game, but all three teams that beat Georgia Tech lost theirs.
The top two teams of the SEC East (Missouri and Georgia) got into the top 15, finally passing up Ole Miss and Mississippi St., but they had too far to go to make it into the top 10. I think 19 of the top 21 being in P5 conferences is about right. I don’t mind giving some credit to the better teams of the other conferences at the bottom of the top 25.
USC made a pretty decent jump forward after they beat Nebraska and several teams in the group immediately ahead of them lost.
This is the completion of my 20th year ranking teams (although it was purely subjective before 2004), and this is the very first time I am ranking Memphis. Memphis is the 101st team I have ranked.
Note on conferences
This is slightly different analysis from my “conference report” series. In this blog, I’m commenting upon the rankings of the various teams rather than strictly looking at games between conferences. I do a summary of this with every ranking (except for the one after the Army-Navy game). You can see my chart here.
The SEC finishes with only one team in the top 10, but six in the top 25, nine in the top 40, and twelve in the top 50 (the chart linked to above does not give any credit for the top 50, just the top 40). Even #13 Kentucky finished 68th, just four spots behind Virginia Tech (who, as I think most people know, beat Ohio St.). Although like most of the country, I was a little surprised by WHICH SEC teams won, I think the above shows how tough it was to go undefeated if 13/14 teams in the conference were good enough to beat who turned out to be the best team in the country.
The SEC has 14 teams, and only 8 of them got to play Vanderbilt. The SEC might not have the best top teams (as was thought for much of the season), but if you think any other conference schedule was tougher, you’re kidding yourself.
Kenneth Massey lists about 90 top-to-bottom computer rankings of teams on his comparison site. Every one of them has the SEC as the top conference (at least based on average ranking of the teams). I know people want to call me a homer, but until I’m not part of the overwhelming consensus (even if one or two come out that disagree, that’s still true), I won’t take that allegation seriously.
The only rankings he lists on there that put the SEC second are the two (subjective) polls, but only the top 25 of each one is considered. Since the SEC has six teams between #26 and #50 in my ratings compared to only two in that range for the Pac-12, that’s ignoring a lot of the SEC’s relative strength toward the middle and below the middle of the conference.