It is that time of year again! What time you ask? Fanson baby watch
2012? Well, yes, but it is also time
once again to break open the football spreadsheet. I introduced the spreadsheet last, year, but
just to refresh everyone’s memory, I starting putting the spreadsheet together
in grad school and I have been making steady improvements to it over the
years. I used to just track scores and
standings, but over time I have refined the algorithms to predict game outcome
/ point spreads and generate power rankings.
As the 2012 season approaches, I have once again inputted the schedule
information into the latest version of the spreadsheet and am taking an early
look at what to expect. If I take the
pre-season rankings for each team, convert that to an estimated power ranking,
and simulate the season by correcting for a teams performance at home or on the
road (roughly a 3-point advantage for the home team), I have been able to
predict the final record of all 124 FBS teams.
This year I even expanded the set of data used to generate the pre-season
rankings by taking the average of several pre-season publications (including
Athlon’s, Lindy’s, Phil Steele, ESPN, and USA Today). Keep in mind that these results are purely
based on math once the pre-season rankings are inputted. What I found was fairly interesting:
Let’s start with the Big Ten. Here are my projected
standings, which each team’s overall record in parenthesis:
Legends Division
1. Michigan
State
1. Nebraska
6-2 (10-2)
1. Michigan
6-2 (8-4)
4. Iowa 4-4
(8-4)
5. Northwestern 3-5 (7-5)
6. Minnesota
2-6 (6-6)
Leaders Division
1. Wisconsin 6-2 (10-2)
1. Ohio
State 6-2 (10-2)
3. Illinois 4-4 (7-5)
4. Penn State 3-5
(6-6)
5. Purdue 2-6
(5-7)
6. Indiana 0-8 (3-9)
I will start first with the Leaders Division, as that is
more straightforward. Wisconsin (ranked #15) and Ohio State Wisconsin is projected
to also beat OSU head-to-head (as they play in Madison this year), while Wisconsin is projected to lose at Nebraska and at Penn State
That being said, I still have my eyes on both Penn State Michigan
(8), Wisconsin
(15), and Penn State Nebraska
(19). They must travel to OSU (loss), Iowa (47), Illinois (54), and Minnesota (69). The spreadsheet only has Purdue beating Indiana (96) and Penn State West Lafayette . However, if the Boilers are a little better
than their preseason ranking of 50 suggests, several of these 6 losses could
all of a sudden turn into Ws, including a potential tie-breaking win against
the (over-rated?) Badgers. Similarly, PSU draws both OSU and Wisconsin at home and does not play MSU or
UofM. The spreadsheet has them losing
all of their road games (at Illinois, Iowa, Purdue, and OSU), but you have to
wonder if they could wind up squeezing a few more wins out of that schedule. Time will tell, I guess.
Moving on to the Legends Division, things are a bit more
complicated. The spreadsheet predicts a
3-way tie between the three projected Top 20 teams. MSU (14) is predicted to beat Nebraska (19) in East Lansing , but to lose
in Ann Arbor
(No!!!!!) and in Madison . The Wolverines (8) are projected to beat MSU
(No!!!!!), but to lose at OSU and at Nebraska . The Huskers are projected to loss at MSU,
beat UofM, but also lose at Ohio
State Alabama (3) in Dallas and at Notre Dame
(24)) will leave them at 8-4, but that does not appear to come into play. If I read the tie-breaker rules correctly,
the Legends Division representative will be the team out of the three with the
best record against all common Big Ten opponents. The only team that UofM, MSU, and Nebraska all play is
OSU… and MSU is the only team predicted to beat OSU. So, the spreadsheet currently likes an
MSU-Wisconsin rematch in Indianapolis
in early December, despite the fact the OSU and UofM are both ranked higher in
most preseason magazines.
So, who will win the Big Ten Championship game? If I take the average of all the preseason
magazine rankings, MSU and Wisconsin
actually finish in a dead heat. The spreadsheet
picked MSU, but that is only because “M” comes before “W” in the alphabet. 3 out of the 5 preseason polls have Wisconsin over MSU… but
I flipped a coin and MSU won, so I will go with the Spartans. Do Wolverine tears taste like Roses? Maybe this year they do!
How about the rest of the country? Well, let’s start with ESPN’s favorite
conference: the SEC . After winning the last 10,000 BCS titles in a row, it is hard not to imagine that
the SEC will be a major player
this year as well. However, the
spreadsheet thinks otherwise. The
reason? Parity. The SEC
is projected to have 9 teams in the Top 30 this year. That means a lot of very tough games, and a
high probability that they will simply beat up on each other. The situation looks particularly brutal in the
SEC West. Alabama (3) has the toughest draw, with 4
road games against Top 30 teams: Arkansas
(9), LSU (2), Missouri
(27), and Tennessee
(28). LSU (2) also has a tough draw with road games at Arkansas (9), Florida (21), and Auburn (26).
Of the SEC West’s projected
3 Top 10 teams, Arkansas
(9) seems to have the best draw by getting ‘Bama and LSU at home, but they also
have to travel to South Carolina
(10), Auburn
(26), and Texas A&M (31). As a result, the spreadsheet projects that no
team in the SEC will finish with
fewer than 3 loses. In fact, the SEC West is projected to have a four-way tie with Arkansas , LSU, Mississippi State Auburn all tied at 5-3. In this scenario, Auburn would actually win the tie-breaker due
to its projected home victories against LSU and Arkansas and despite road losses at Mississippi State SEC West is going to have to
win at least 2 tough road games to be in the National Title hunt, and right now
Arkansas and Auburn seem to have the edge in schedules.
On the other side the of the SEC ,
an edge in schedules may also pay dividends, in this case for South Carolina
(10), who I project to edge out Georgia (7) for the title and right to beat
Auburn in the SEC title game. The
Gamecocks are expected to drop road games at LSU (2) and Florida (21), but they do draw the Bulldogs
at home this year. Georgia must
also travel to Auburn (26) and Missouri
(27) and are thus projected to finish at 5-3, a game behind USC . So, if
the SEC breaks down as I have
outlined above, its 4th best team (USC )
will meet its 6th best team (Auburn )
in the title game, entirely due to the schedule.
Moving on to the PAC-12, the situation seems much more
straightforward, as the conference is only projected to have 3 top 30 teams: USC (1), Oregon
(4), and Stanford (20). Both USC and Oregon are projected to finish 8-1 in
conference play (losing at Stanford and at USC
respectively), winning their Divisions by two games, and meeting in the
conference title game where #1 ranked USC
would be expected to win.
In the Big 12, the conference appears to be more competitive
as 6 of the conferences 10 teams are ranked in the top 25 preseason: Oklahoma (5), West Virginia (12), Texas (13), OK State
(18), TCU (18), and Kansas
State Oklahoma
State Kansas
(78) and Baylor (43). I project a 3-way
tie between Oklahoma, Oklahoma State and West Virginia at 7-2, but OK State
would win the tie breaker due to its projected home wins against WVU and Texas
and TCU (projected to finish tied for 4th place). The key game of the season could be the Oklahoma – TCU game in Dallas to close the Big
12 season, which is currently projected to be a very close TCU win. A Sooner win would clinch them the conference
crown and could vault them into the National Title Game.
As for the ACC, it appears that schedules may also play a
key role here. Only Florida State Chapel Hill ,
which would result in the Tar Heels making an appearance in the ACC Title Game. In the Atlantic Division, a 7-1 Clemson squad
is projected to edge out a 6-2 FSU team, despite losing to the Seminoles
head-to-head. Clemson has to play FSU on the road on September 22nd,
but after that only has to travel to Wake
Forest Boston College South Carolina at home
and gets a games in Atlanta
against Auburn . Clemson has the schedule to finish 11-1 and
be solidly in the National Title hunt, especially after they win the ACC title
over UNC. Florida State
Elsewhere in the nation, Notre Dame (24) is expected to be a
borderline Top 25 team, but with road games at Michigan State Oklahoma
(5), and USC (1), they project to
finish at 9-3 and not be a factor in the national picture. Similarly, the Big East seems to have
descended further into chaos as the top ranked team is projected to be Louisville at only
#30. The Cardinals are projected to win
the conference with a 5-2 record due to a head-to-head wins over 5-2 Cincinnati and 5-2 South Florida . Louisville could finish
at 10-2 with non-conference wins over Kentucky
and UNC, but projected loses at Pitt and Rutgers
would likely keep them out of the national discussion. Nothing exciting predicted here.
However, they may be some interesting story lines this year
in the Mountain West, Conference USA , and the MAC . With
TCU moving on to the Big 12, Boise
State Nevada
(58), but Boise
is projected to go 8-0 and win the conference.
In fact, the only thing they may stop an undefeated season is the
Bronco’s stop in East Lansing
on opening weekend. MSU is projected to
win, but if they falter, Boise
could run the table and be hard to keep out of the BCS
title game, especially if MSU were to make the Rose Bowl, as predicted.
In Conference USA , the Knights of Central Florida
appear to be the cream of the crop, although the preseason rankings have
drastically varying opinions as to how good they will be. ESPN has them at 15, while Athlon’s has them
at 66. Their average ranking works out
to #35, which is much better than the second best team in the conference, Houston at #60. Due to this factor, UCF is projected to run
the table in C-USA and finish 11-1 overall.
They are picked to face and beat SMU
(71) in the title game after SMU
wins the head-to-head tie-breaker with Houston
after both teams finish 7-1. UCF will
face a tough test at Tulsa (61) near the end of the season, but they also get a
chance to play Missouri (27) at home (where I project a close win), and Ohio
State (11) in Columbus (projected loss).
If the Knights could somehow beat both OSU and Missouri … they could really throw a wrench
into the National picture.
In the MAC , the
top two teams in the conference are projected to be Ohio (65) and Western
Michigan (76). As luck
would have it, both teams also drew very nice schedules in which the toughest
road conference game that either team will face is when Ohio travels to Miami-Ohio (90). Due to this fact, both teams are projected to
sweep their divisions and finish 10-2 overall.
Due to its higher ranking, Ohio
is projected to win the Conference Championship game and finish with 11
wins! In the non-conference Ohio is expected to drop
road games at Penn
State Marshall , while Western Michigan is projected to lose at Illinois and Minnesota . I am not sure about you, but none of those
teams strike that much fear into me. I
think a 1-loss (or undefeated?) MAC
team could be a possibility this year.
For completeness, Florida International (75) is projected to
win the Sun Belt Conference while Louisiana Tech (62) is projected to win the
watered-down WAC . Both teams are projected to finish 8-4 and
not be a factor on the national stage.
So, if we add this all up, where does it leave us come
January, 2013? Based on the analysis
above, the automatic BSC bids go
to:
Clemson (ACC Champs) 11-1
That leaves four BCS
slots left with an additional three teams projected to finish with 11 wins (Oregon , Boise State Central Florida ). Oregon would seem to be a lock for an at
large bid and with Louisville’s only decent win coming against UNC, I seriously
doubt they finish in the BCS Top
16, which would require that the BCS
to take either UCF or Boise as an automatic bid. Boise
would have played the tougher schedule and would seem to be a lock to have the
higher BCS ranking than Louisville .
So, that leaves two addition slots for the remaining 2 or 3
loss teams. It seems quite unlikely that
the final two teams would come from non-BCS
conferences, so the options remaining would be: Oklahoma (10-2), Florida State
(10-2), Ohio State (10-2, but ineligible), West Virginia (10-2), Wisconsin
(10-3), and Nebraska (10-2) as well as the 3-loss teams from the SEC including LSU (9-3), Georgia (9-3), and
Arkansas (9-3). I suppose Notre Dame
might also get consideration at 9-3.
To figure out who the final two teams are, it is best to go
through the BSC selection
process. Based on all the analysis
above, if the season were to go this way, the National Title game would pit USC against Clemson as the only 2 BCS teams with only 1 loss. USC
would be expected to win that match-up and take the National Title. The Rose Bowl would get MSU and would almost
certainly snatch up a highly ranked Oregon
to keep the Big 10-Pac 12 alliance intact.
Since the Orange Bowl would lose Clemson to the Title game, it would get
to pick next, and I would guess that they would pick up a 10-2 Florida State
For the January 2013 Bowls, the Fiesta Bowl is next up to
pick. They would get Big 12
representative Oklahoma
State BCS
ranking would play a huge roll, as the Fiesta Bowl could go in several
directions. They could take Boise State Nebraska ),
or they could take a 3-loss SEC
team like LSU, Arkansas ,
or Georgia . Due to the old Big 12 connection, I suspect
the Fiesta Bowl would take Nebraska ,
so long as they were in the BCS
Top 14. Otherwise, I think they would
take one of the SEC teams… maybe Arkansas since they are
projected to have beaten LSU the last game of the season. Following this pick, the remaining picks fall
into place as the Sugar Bowl picks next and would match up South Carolina with
Boise State (to avoid picking up Louisville) and the Orange Bowl would get the
FSU-Louisville game (must see TV, indeed).
The big BCS losers are
Oklahoma and West Virginia, as they are projected to be better teams that
Nebraska, but the selection order in 2013 makes it hard for any Bowl to pick
them up, unless the Fiesta Bowl somehow agrees to take Boise State so that the
Sugar Bowl could take another Big 12 team (most likely West Virginia, since
they are projected to beat Oklahoma late in the season)
In summary I project the BCS
Bowl Match-ups to be:
Rose Bowl: Oregon (4) over Michigan State
Sugar Bowl: South
Carolina (10) over Boise State
Fiesta Bowl: Oklahoma
State Nebraska (19)
Orange Bowl: Florida State Louisville (30)
Keep in mind again that all these projections simply assume
that the preseason rankings are accurate and than home teams get a standard
advantage (about 3 points). Based only
on these factors, only 4 teams out of the projected top 10 will make BCS games.
Schedules matter. Will any of
this come to pass? It is hard to say. On any given Saturday almost
anything can happen. But, this analysis suggests an outcome that would
occur if all teams are exactly as good the pre-season rankings think that are,
a standard home field advantage applies, and all teams have an extremely
consistent level of play. I hope this has been entertaining!
Here are my projected key games that will impact the
National Title picture:
09/01 17 Clemson vs. 26 Auburn (in Atlanta ) (The winner of this battle teams
with good schedules gets an early leg up)
09/24 10 South
Carolina @ 17 Clemson (A win could vault Clemson into
the National Title game discussion)
10/06 6 Florida
State @ 33 NC State (a misstep here could knock the ‘Noles out of the ACC and
National Title hunt)
10/20 14 Michigan
State Michigan
10/27 8 Michigan
@ 19 Nebraska
10/27 14 Michigan
State Wisconsin
11/03 19 Nebraska
@ 14 Michigan
State
09/29 15 Wisconsin
@ 19 Nebraska
10/06 11 Ohio
State Nebraska
11/24 8 Michigan
@ 11 Ohio
State
09/29 11 Ohio
State Michigan State
12/01 5 Oklahoma
vs. 18 TCU (in Dallas) (A TCU win could knock the Sooners out of the National
Title hunt and even out of a BCS
Bowl)
09/08 35 UCF @ 11 Ohio
State (if the Buckeye’s choke, the Knight could run the table and create BCS Chaos)
09/01 65 Ohio
@ Penn State Ohio could run the table
and create BCS chaos)
09/01 25 Boise
State Michigan State (Boise could also run the
table with a win in East Lansing
12/08 Pac-12 Title Game: Oregon vs. USC ?
(Winner should play in National Title Game)
10/13 3 Alabama
@ 27 Missouri
10/20 3 Alabama
@ 28 Tennessee
09/23 9 Arkansas
@ 31 Texas
A&M
10/06 9 Arkansas
@ 26 Auburn
09/08 26 Auburn
@ 38 Mississippi
State
10/20 26 Auburn
@ 44 Vanderbilt
9/22 2 LSU @ 26 Auburn
10/06 2 LSU @ 21 Florida
09/15 3 Alabama
@ 9 Arkansas
11/03 3 Alabama
@ 2 LSU
11/24 2 LSU @ 9 Arkansas
Comments
Post a Comment