2015 College Football Preseason Analysis

(Originally posted to various internet sources in July of 2015)

Hello. My name is Paul, and I have a problem: I love college sports and I love spreadsheets. When I put these two loves together, strange, time-consuming things happen. Back in the spring, I made a couple of Fanposts detailing the Excel database that I created to track NCAA basketball tournament history and some of the data mining that I have performed on it. But basketball is only the tip of the iceberg. I actually first flirted with sports and spreadsheets long before the basketball database every existed. My first love, in fact, is the College Football Spreadsheet, and I would like to share that love with Spartan Nation today, especially those that are #teammath.

Many years ago now, I developed a spreadsheet and algorithm that allows me to generate a power ranking, predict point spreads, and probabilities of victory for every game in the Division 1 college football season. During the long dark summer of no college sports, I have made it a tradition to rebuild and improve the spreadsheet including each team’s schedule for the current year. For the last 5 years or so, I have used the spreadsheet to simulate the results of the upcoming season using the preseason magazine rankings as a guide to set power rankings. In general, the preseason rankings are not a terribly accurate predictor of which teams will actually be good or not, but it is the best thing that we have to go on. If nothing else, this analysis has helped me to understand the impact each team’s schedule might have on the upcoming season. For this year’s analysis, I have taken things a bit farther. This year I actually conducted over 10 different simulations, all using different levels of parity between teams (i.e. how much better, mathematically, team #20 is compared to team #21). One of the simulations I have deemed my "base" or "most trustworthy" simulation because it yields the distribution of upsets (about 20%) and distribution of wins and losses that seems to best match the distribution of past seasons. I use this base simulation for all my actual predictions on each team's individual record. However, I will also use the trends in the other simulations to suggest other likely scenarios, including the results of the simulation that I believe give the most accurate point spreads and victory probability, based on correlations to previous season's data. I will refer to this simulation as my "probabilistic simulation" (This particular simulation is too conservative to use to predict actual outcomes. There are too few upsets.) For me, the results are always surprising and interesting, and I hope that you will agree. (If you have interested in more detail about the methodology I use for my simulations, I have posted a companion article that outlines my methodology in more detail. Otherwise, let’s jump right in.)

As an overview, here is a table of selected Division 1 teams showing three basic results from my probabilistic simulation: the consensus preseason ranking based on the average of four preseason publications, the probability that each team will finish the regular season with 0, 1, or 2 losses, and the expected value of total win and conference wins for each team (the sum of the probabilities of victory for each conference game and total games, which should correlate to the over/under for each team). A quick glance reveals that seven teams have an expected win total great than 10, and 5 teams win greater than a 10% chance to go undefeated. I will refer to the data in this table throughout the remainder of this discussion.

I will start, as always with the only conference that matters: the Big Ten. Midwest pride was recaptured back in January when the 4-seeded Ohio State Buckeyes dumped the Crimson Tide and Ducks using their 3rd string QB to claim the first National Title decided by a playoff. As we approach kick-off for the 2015-16 season, the Buckeyes may not know who their QB will be, but they do know that they will be kicking off as the consensus #1 team in the country. What they may not know is that they also have one of the more favorable schedules in the country. Based on my projected point spreads, their toughest games of the season will be the opening weekend road game at Virginia Tech (with a consensus ranking of 25) and the highly anticipated home game against Michigan State (consensus rank of 9). My base simulation suggests OSU wins both games. The Buckeyes are certainly likely to win all of their home games, and the only other road game that *might* present a challenge is the finale in Ann Arbor. Ohio State's conference road schedule is very favorable with match-ups at Indiana (ranked 70th), Rutgers (64th), Illinois (63rd), and Michigan (32). I suppose it possible that Coach Harbaugh could part the Scarlet and Gray Sea or rain fire and brimstone from the sky... or something and somehow pull the upset, but I doubt anyone outside of Ann Arbor would bet on that. The probabilistic simulation gives OSU a 70% chance to beat both MSU and VA Tech, a 75% chance to win at Michigan, and over a 90% chance to beat everyone else on their schedule. This manageable schedule is the reason that the Buckeyes expected value for total victories is the highest in the nation at almost 11, and their odds on finishing undefeated are the 2nd highest at 22%. In one of my simulations, OSU drops the road game at VA Tech, but in all the rest they finish a perfect 12-0.

As for the rest of the Big Ten East, MSU (9th) is obviously picked to have a strong team, while Michigan (32) and Penn State (36) are predicted to have OK teams, and Rutgers (64), Maryland (67), Indiana (70) and predicted to bring up the rear of the Division. MSU's schedule is tricky this year, but I project the home field advantage of Spartan Stadium to be enough to get by Oregon (5) in Week 2 and Penn State (36) in the regular season finale. Road games at Michigan (32) and Nebraska (34) could also be tough, but in all but one of my simulations, the Spartans are up to the task and are projected to finish at 11-1. The Spartans also have the 7th highest expected value for wins at just a hair over 10 and a 60% chance to win at least 10 games. Penn State (36) is expected to lose both at OSU (1) and MSU (9), but is expected to beat Michigan (32) in Happy Valley and handle the manageable cross-over games at Northwestern (60) and vs. Illinois (63) to finish Big Ten play at 6-2. An early road trip to Temple (56) however, I actually project them to drop, which seems to be par for the course for Big Ten teams in September. Michigan (32) is also projected to lose on the road in Week 1 to a slightly better Utah (30) squad in additional to their projected losses at Minnesota (46), at Penn State (36), vs. MSU (9), and vs. OSU (1) to finish at 4-4 in conference and 7-5 overall. Probabilistically, I have UofM’s expected value for wins at just barely over 8 overall and barely under 5 in conference, so they have a reasonable chance to do a little better than my base simulation suggests, but I still have them at only a 14% chance to win 10 games or more. As for the bottom of the East, I project that Indiana (70), Rutgers (64), and Maryland (67) all finish below 500 and out of Bowl contention. There is one nightmare scenario for MSU, however, in the low parity simulation where Michigan has just enough home field advantage to beat both MSU and OSU in the Big House and MSU is not able to get past Nebraska on the road. In this case MSU would finish in 3rd place in the Division, behind Michigan. Unless Harbaugh has the ability to raise Bo from the dead, I doubt we will see this scenario.

Turning our focus to the Big Ten West, Wisconsin (ranked 23) is projected to be the cream of the crop, but Nebraska (34) is not far behind and Minnesota (46) seems dangerous, at least at home. Wisconsin does have to travel to Lincoln, where I project them to lose to the Huskers, but I have them also just squeaking by the Gophers in Minneapolis to finish 7-1 in league play and 10-2 overall (with a loss to Alabama in Week 1). Despite Nebraska's projected win over Wisconsin, they are project to lose to both MSU (9) and on the road at Minnesota (46) to finish at 6-2 in league play (and 9-3 overall due to a projected loss at Miami (37) in Week 3). In theory, I like Minnesota to make some noise this year, but home games against TCU (3) and Wisconsin (23) and tough road games at Ohio State (1), Iowa (52), and even Northwestern (60) could add up to a 7-5 season. The probabilistic simulation tells a similar story as Wisconsin’s expected value for total wins is 9.9, while Nebraska’s is 8.5, and Minnesota’s is only 6.7. Wisconsin’s value is particularly high (6th amongst the Power 5 teams) considering their preseason ranking of only 23. I think this speaks to the subtle softness of their schedule outside of the games against Alabama, Nebraska, and Minnesota. In their remaining 9 games, I project the probability of victory to be over 85% in all of them and over 95% in six of them. I guess this makes some sense knowing that the Big Ten’s scheduling "computer" is code-named "Barry"… Alvarez. I project Iowa (52) to finish at 4-4 and 8-4 overall and Northwestern (60) to barely make a Bowl Game at 4-4 and 6-6 overall (with non-conference losses vs Stanford (20) and at Duke (48) both during which more homework will be completed in the stands than at all other Division 1 games combined). There are a couple of high parity simulations where Nebraska (34) comes out on top in the Division (by virtue of beating MSU (9) or having Wisconsin (23) lose at Minnesota (46)) but in general either team would be expected to get steamrolled by OSU in the Big Ten Championship game.

Prior to OSU claiming the National Title this past winter, the trophy resided not in SEC country, but in the less (slightly) arrogant confines of the ACC. The ACC is intriguing this year, as no team has a preseason rank in the top 10, yet four teams are in the Top 25: Florida State (12), Clemson (14), Georgia Tech (21), and Virginia Tech (25). Six additional teams round out the Top 50: Miami (37), Louisville (38), NC State (39), UNC (40), Pittsburgh (43), and Duke (48). The team, by far, with the highest schedule upside is Clemson (14). They play three top 25 teams: Notre Dame (8), FSU (12) and GA Tech (21), but they play all three of them at home, where I project them to win in my base simulation. As for the Tiger's road games, they are tough, but all potentially manageable: at Miami (37), Louisville (38), NC State (39), and even at South Carolina (44). Based on this amazing schedule, I actually project Clemson to be the surprise team of year, winning the Atlantic Division and going 12-0 overall. Of course, the tricky part is that I have them favored by 5 points or less in a whopping 7 out of 12 games, so it will certainly be easy for the Tigers to stumble, in which case Florida State (12) is likely to come out on top. The probabilistic simulation bears this out, as Florida State actually has a slightly higher expected value for total wins (10.1 vs. 9.6), but the difference is overall fairly small. The true strength of Clemson’s home field advantage on that particular day might be the difference.

As for the Coastal Division, there is a fair amount of parity, but Georgia Tech (21) appears to have the upper hand by virtual of hosting their biggest competition, Virginia Tech (25). I project them to lose at Clemson (14), at Miami (37) and at Norte Dame (8) to finish 6-2 in conference play and 9-3 overall, which would be good enough to tie VA Tech (who I also project to lose at Miami) and earn a ticket to the Conference Championship Game via the head-to-head tiebreaker. Similar to Clemson’s situation, Georgia Tech does seem to have a slim margin for error. Their expected value for total wins is only 7.9, which is well below average for a team in the Top 25, while VA Tech’s expected value is almost a full game higher at 8.8. Georgia Tech does project to lose again to Clemson on a neutral field (as would VA Tech), but the Yellow jackets would at least be comforted by their upset win in Atlanta against Georgia (7) the week before as well as an earlier season win over FSU (12). That does seem like a tough schedule for the Yellowjackets. Some coaches might even seem to be paralyzed by fear with that kind of schedule. Fortunately, I heard that Coach Harbaugh visited Atlanta sometime this summer, and after that the paralysis seemed to suddenly disappear.

Last year, TCU and Baylor proved to the nation that Texas can produce some quality football programs that do not hail from Austin or College Station. This year, both Big 12teams are projected to be in the Top 5 with TCU ranked 3rd in the consensus rankings and Baylor ranked 4th. Oklahoma (16) and Oklahoma State (24) round out the top of the conference with Texas (31) and West Virginia (35) not far behind. The Big 12's ten teams use a full road robin 9-game conference schedule, which makes the impact of the schedule a bit lower than it is in other leagues. TCU appears to get the nod here, thanks in large part to a home game with Baylor. MSU saw first-hand in the Cotton Bowl the impact of the blinding speed with which Baylor runs their offense, but I heard that Coach Harbaugh was able to cure this blindness during his trip to Dallas this summer. As a result (perhaps) TCU wins the league in 11 of my 13 simulations. However, TCU does go on the road to play both Oklahoma schools. My base simulation has the Horned Frogs dropping the game in Norman, but beating OK State, which would be enough to clinch the conference at 8-1, (11-1 overall). Baylor is projected to lose both at TCU and Oklahoma State, but is projected to win the rest of their games to finish 10-2. You might think that Baylor is most likely team to win the conference if TCU stumbles (and they likely are), but my spreadsheet actually had Oklahoma State as the Conference Champs in the two highest parity simulations. In these scenarios, there is a 4-way tie at 7-2 for first place among TCU, Baylor, OK, and OK State. The Cowboys actually draw the other three teams at home, and would win the tiebreaker in this scenario even though they would have losses at Texas (31) and at West Virginia (34). That being said, Baylor actually has a slightly higher expected value for wins compared to TCU (10.3 vs. 10.2) but this is due entirely to the Bear’s Charmin-like non-conference schedule as TCU’s expected value for conference wins is slightly higher at 7.34 compared to 7.29 for Baylor. Of all the Power Five teams, Baylor and TCU also rank 2nd and 3rd for the highest probability of going undefeated at 14% and 10.4% respectively.

In the Big Ten, ACC, and Big 12, my simulations gave pretty consistent results in most of the cases. However, the same cannot be said of the Pac-12. The North Division appears to be a bit top heavy with both Oregon (5) and Stanford (20) in the Top 20, but no other team in the Top 40. (Cal is the next best team at 41.) Despite Oregon being projected to be the best team with an expected value of wins at 9.8, the majority of my simulations, including the base simulation, suggest that Stanford (20) will win the Division. The Cardinal are not projected to beat USC (6) on the road in a mid-September tilt, but they have the benefit of facing Oregon (5), UCLA (15), and Arizona (29) all in Palo Alto and avoiding Arizona State (19) all together. Stanford is even projected to beat Notre Dame (8) at home at the close of the regular season to finish 8-1 in conference play and 11-1 overall. By contrast, Oregon (5) is expected to lose tough road games at MSU (9) and Arizona State (19) in addition to the loss to Stanford. In the lower parity simulations, Stanford's home field advantage is no longer strong enough for the Cardinal to overcome the Ducks, which suggests the November 14 match-up of Oregon at Stanford is essentially the only game that matters in the Pac 12 North. Stanford does seem to have a much higher margin for error, however, as their expected value for total wins is only 8.5.

The story in the stronger Pac-12 South is eerily similar. USC (6) is projected to be the best team in the Division, with UCLA (15), Arizona State (19), Arizona (29), and Utah (30) not far behind. The Trojans problem, however, is that they have a fairly brutal schedule, including projected losses at ASU (19), at Notre Dame (8), and at Oregon (5) which my base simulation suggests will drop them to 9-3. UCLA may have even a worse draw with road games at Arizona (29), Stanford (20), Utah (30), and USC (6) (all projected losses). With those tough schedules, the LA schools probably feel like they are a bit underwater. I seem to remember a shirtless man wearing khakis with the ability to walk on water being spotted in the Bay Area in the recent past. I wonder if he could help? Regardless, Arizona State seems to be the team with the upper hand in the South. The Sun Devils are projected to lose at UCLA and Utah (and at Texas A&M to start the season), but an early season home win against USC would give ASU the tiebreaker as both teams are project to finish at 7-2. In this scenario, the 3-loss ASU team would beat the 1-loss Stanford team and claim the Pac-12 crown. In the low-parity simulations, USC is able to overcome the Arizona State home field advantage and claim the crown for themselves. This is also reflected in the expected values for total wins where USC leads the pack at 9.2 while UCLA’s tally is 9.1 and ASU’s is only 8.3.

It seems like I am missing a power conference. Let's see, there is one conference left that is 1-6 in their last 7 BSC / New Years 6 Bowl Games. They probably aren't important, but let's review them anyway. I believe that they are called the SEC. (FYI, It is usually spelled with a dollar sign, in case this more traditional spelling is unfamiliar). Despite the eye-openingly bad performance of the SEC West in Bowl season last year, the preseason rankings have the Division right back in the hype zone with ALL SEVEN TEAMS ranked in the Top 30. Alabama has the highest overall ranking at #2 followed by LSU (10), Auburn (11), Mississippi (13), Arkansas (17), Texas A&M (22), and Mississippi State (28). Based on these rankings, the spreadsheet predicts the SEC West to be a major bloodbath. Most of my simulations (6) have Alabama coming out on top, a few have Auburn winning in a crazy four team tie breaker situation, but the third outcome is perhaps the most interesting and surprising. In 5 of the 13 simulations, including my base, most trustworthy simulation, Texas A&M actually wins the Division, despite being objectively the 2nd worst team. As you can probably guess by now, the main reason is that the Aggies have a more favorable schedule. They draw Alabama, Auburn, and Miss State at home, and they are somehow able to play Arkansas at a friendly neutral site in Arlington, TX. I have them winning all 4 of those games as well as a home game against mediocre South Carolina (44) and a road game at conference doormat Vanderbilt (74). Even with probable loses at LSU and Mississippi, that puts the Aggies at 6-2. I have no other SEC West team with less than 3 losses in this scenario due to the high number of tough road games. Alabama in particular is hurt by a scary road game at SEC East favorite Georgia (7), but the nexus of this somewhat crazy prediction is when Alabama travel to Texas A&M on October 17th. To me, this looks like the hidden key match-up for the entire conference. The probabilistic simulation still favors Alabama with a 54% chance of finishing with 2 or less losses and with an expected value for wins at 9.1 (neck and neck with LSU at 9.0, and ahead of Auburn at 8.5 and A&M at 7.8) so similar to teams like Clemson and Arizona State, Texas A&M has a slim margin for error. If Alabama were to survive the SEC West and make the plays, I am sure that the faithful would be breaking open a few bottles of wine to celebrate. I heard that Saban recently received a shipment as a peace offering from somewhere up north. Oddly, it seemed to all be in sealed water bottles with block "M"s on the side.

Fortunately, the SEC East appears much less complicated. Georgia (7) is predicted to be the best team in the Division, with Tennessee (18), and Missouri (26) within shouting distance. The Bulldogs do have tough road games at Tennessee, Auburn, and Georgia Tech (21) which could leave them at 6-2 (9-3 overall), but a home win against Missouri (coupled with a Missouri loss at Arkansas) looks to be enough to sew up the Division crown. Tennessee could cause problems, but road losses at Florida (33), Alabama (2), and Missouri (26) should throw them out of contention. In this scenario, Georgia would be favored to beat Texas A&M in the SEC title game, but they would finish with 3 losses.

As a final note on the power teams, I would be remiss if I did not mention our friends down in South Bend. For the first time in perhaps forever, The Fighting Irish of Notre Dame (8) have zero Big Ten teams on their schedule, but they have none the less put together an intriguing schedule. I have them earning victories in their 3 toughest home games vs Texas (31), Georgia Tech (21), and even USC (6), but I do not project them to beat Clemson (14) or Stanford (20) on the road, which would leave the Irish sitting at 10-2 overall, which matches well with their expected value for wins coming in at 9.7.

With the "Power 5" plus Notre Dame now covered, it seems only fair to take a quick spin through the "Group of 5" Conferences. Keep in mind that the highest rated team from the Group of 5 does earn a spot in one of the New Year’s Six Bowl Games.

In the American Athletic conference, only one team, Cincinnati (47) enters the season with a consensus ranking in the Top 50. With six other teams in the top 75 or so, the does seem to be some parity in this league such that no team is likely to finish with fewer than 2 loses. The majority of my simulations have Cincinnati winning the East division, although in a few cases Temple’s (56) somewhat softer schedule boosts them to the top of the standings, despite a likely loss at Cinci on September 12th. On the West side of the conference, the best team on paper looks to be Memphis (54), but Houston (62) draws both the Tigers and Bearcats at home and is projected to win the Division in all of my simulations. Cincinnati would be favored to win the Conference Championship game, but with a project record of 10-3 (including a home win against Miami and a road loss at BYU) is not likely to put them in contention for a New Year’s Six Bowl.

It seems like Conference USA used to be a fairly decent conference, but when 9 of the 13 teams are ranked outside of the Top 90…. I don’t think that moniker applies. The highest ranked teams are Marshall (57) and Western Kentucky (66) in the East and Louisiana Tech (72) in the West. All of my simulations have LA Tech winning the West and the East will almost certainly come down to the final weekend when Marshall travels to Western Kentucky for a winner-take-all showdown. Most of my simulations have WKU winning that contest and the C-USA title game over LA Tech the following week. In this scenario, WKU will still end the season with 3 losses as they have non-conference games at LSU (10), Vanderbilt (74) and Indiana (70). But, Marshall is not quite so adventurous. Their non-conference slate includes Norfolk State, road games against bad MAC teams Ohio (90), and Kent State (109), and a winnable home against Big Ten bottom feeder Purdue (80). If Marshall beats WKU, they could very well end up at 13-0 and in the conversation for the Peach Bowl even though they would have ZERO wins against the nation’s Top 60. In fact, the probabilistic simulation suggests that Marshall is the 3rd most likely team to go undefeated with a 15.6% chance.

Speaking of the MAC, the always entertaining Midwest conference looks to be a three-team race between Bowling Green (75) in the East and Northern Illinois (68) and Toledo (73) in the West. I would like to say that Western Michigan (79) also has a shot, but they face both NIU and Toledo on the road. All of my simulations have BGSU winning the East, but the West is a bit tricky. NIU is projected to lose on the road to Toledo, but Toledo’s schedule is a bit tougher with road games at BGSU (75) and Ball State (93). Most of my simulations have Toledo winning the Division, but my base simulation has Northern Illinois finishing on top and winning the Title Game, so I will go with that. The probabilistic simulation also favors NIU over Toledo by almost a full game of expected value (8.8 vs. 8.1) In any event, no MAC team is projected to finish with fewer than 3 losses, so I do not see any of them making a push for a New Year’s Six Bowl.

If I were a betting man, I would wager that the Group of Five representative will instead come from the Mountain West, which makes sense as Boise State (27) is the only team from the Group of Five other than Cincinnati (47) ranked in the Top 50. There appears to be enough separation between the Broncos and the rest of the league that I project them to run the table in conference to finish 8-0.. However, the road game at Utah State (53) on October 17th could be a challenge. In Boise were to lose this match-up, Utah State has the schedule to actually win the Division and ultimately the conference (which does happen in a few of my simulations). In either case, San Diego State (69) has the clear upper hand in the West Division as they are projected to be the strongest team with a very manageable road schedule. I have them winning the Division by at least two games in every simulation that I ran. However, the Aztecs are projected to lose to either Boise or Utah St in the conference title game. If Boise State does win the conference, they could even run the table as their non-conference schedule includes winnable games against vs. Washington (50), Idaho St (non-D1), at Virginia (55), and at BYU (45). On paper, BYU is the best team that Boise will face. I do project them to lose that game, but 13-0 is a serious possibility for the Broncos. The probabilistic simulation agrees and gives Boise a 23% chance of finishing undefeated which is the highest in the nation. Boise has tasted a lot of success lately, and seems hungry for even more. Fortunately for them, I hear that coach Harbaugh is planning to swing by Boise this fall with a "very large" seafood and bread meal. I am not sure where he plans to get it from, though. (That is the last one, I promise.)

The last remaining conference is truly the bottom of the barrel: the Sun Belt Conference. Their best team is projected to be our old friends Appalachian State (78) and over half of the conference is in the bottom 10% of all Division 1 teams. App State has as easy conference road schedule and Georgia Southern (84) appears to be the only team that could threaten them for the conference title, but they play at App State this year. App State even has some winnable non-conference games against Howard (non-D1), at Old Dominion (107), and vs. Wyoming (108), but they are likely to get obliterated at Clemson (14) in week 2. 11-1 and a token Top 25 ranking is possible, but I don’t see them challenging for a major bowl game.

So, assuming this all comes to pass, how would the play off and other New Year’s Six Bowl games shake out? In my base scenario, there would be two undefeated teams at the end of the season: ACC Champs Clemson (14) and Big Ten Champs Ohio State (1) both at 13-0. There would also be two Power Five conference teams with only one loss: Big 12 Champ TCU (3) and Michigan State (9) both at 11-1. I think that Clemson, OSU, and TCU would all be shoe-ins for the playoff, but the forth team gets tricky.

MSU would have a very solid case, as we would have a win over Oregon (5) and our only loss would come on the road to the team that would in this case be the wire-to-wire #1 team in the country. But, there would still be several other teams that might challenge for the 4th spot. Boise State (12-1), Marshall (11-1), and Appalachian State (11-1) would all also have one loss, but none of those teams would have the resume that MSU would. Georgia would be the official SEC Champ at 10-3, but if (as I project) they do lose at Georgia Tech, I think that it would be hard to justify putting a 3-loss team (even if they are the SEC Champ) in over a 1-loss team with MSU’s resume. Missouri would be at 10-2 and the rest of the SEC West would have 3 losses or more. I don’t see any of those non-champs making it either. As for the Pac-12, Arizona State would be the official champ also at 10-3, but I think that same logic would hold. Stanford would be 11-2, but would be fresh off a loss to ASU in the Pac-12 Title Game. No other Pac-12 team would have less than 3 losses either. Again, it is hard to imagine those teams leapfrogging MSU. In the Big 12, Baylor is projected to finish 10-2, and if they can win on the road at OK State, 11-1 is also possible. But, Baylor has a dreadful non-conference schedule, and I seem to remember them losing to MSU last year at the Cotton Bowl (which shouldn’t matter… yet probably does matter). A second team from the ACC does not seem viable either, so that only leaves a 10-2 Notre Dame team as the only possible competition. This could get tricky, but Notre Dame’s final week loss at Stanford would likely be enough to give MSU the edge. If ND were to beat Stanford and finish 11-1… well I think MSU would wind up on the outside looking in.

So, with the Final Four set, it is now a question of seeding and placing the four teams into the Cotton Bowl and Orange Bowl. OSU would be the obvious #1, and I think MSU would be the obvious #4, but the selection of TCU or Clemson as #2 and #3 is a toss-up. Furthermore, would the committee really want to have an all-Big Ten Semifinal rematch? In addition, the favored team in not supposed to play in a location that is geographically a disadvantage. The Cotton Bowl is clearly home turf for TCU, and the Orange Bowl is technically an ACC Bowl, so I am not sure how that might work. If it were up to me, I would match-up OSU vs. Clemson in the Orange Bowl and TCU vs. MSU in the Cotton Bowl. In order to do this, TCU would have to get the 2-seed and MSU would get the 3-seed. This might sound odd, but this is actually the order in which the teams are ranked in the preseason and I think that it would be the best way to avoid rematches and minimize unfair home field advantages. In this scenario, I should point out, OSU would be favored to repeat as National Champs with a win over TCU.

As for the remaining New Year’s Bowls, I believe the Rose Bowl and Sugar Bowl are obligated to select teams from the Pac-12, Big Ten, SEC, and Big 12 respectively. So, this would suggest a Rose Bowl match-up of Arizona State vs. Wisconsin (being the next highest ranked Big Ten team, most likely) and a Sugar Bowl match-up of Georgia vs. Baylor. Those sound like fun match-ups to me. That leaves four slots remaining for the Peach and Fiesta Bowl, with one slot reserved for the top ranked team from the Group of Five, which I project to be Boise State… again. Several teams would seem to be in line for the remaining three slots, including Notre Dame at 10-2, Stanford at 11-2, Florida State at 10-2, and various teams from the SEC West (Texas A&M, Alabama, LSU, and Auburn being the most likely all with 3 losses). Of those teams, I think Notre Dame, Stanford, and Alabama would be the most likely to land at the top of the select Committee’s rankings. If nothing else, Bama is going to start higher in the rankings and will therefore likely not fall that far. For geographic reasons, an Alabama – Notre Dame match-up in the Peach Bowl and a Stanford – Boise State match up in the Fiesta Bowl makes the most sense.

In summary, this is how I see the season shaking out, based on my base simulation:

National Title: Ohio State (1) over TCU (3)

Orange Bowl: Ohio State (1) over Clemson (14)

Cotton Bowl: TCU (3) over Michigan State (9)

Rose Bowl: Arizona State (19) over Wisconsin (23)

Sugar Bowl: Baylor (4) over Georgia (7)

Peach Bowl: Alabama (2) over Notre Dame (8)

Fiesta Bowl: Stanford (20) over Boise State (27)

That all being said, this scenario hinges on some very specific game outcomes (albeit, mathematically consistent outcomes). In my heart of hearts, I just don't see 2 Big Ten teams making the playoffs. I think it is likely that one team does emerge from the SEC West with 1-2 losses and they would most certainly wind up in the play-offs instead of MSU, likely matched up against TCU in the Cotton Bowl. That would drop MSU back to the Rose Bowl in an winnable game against the Sun Devils. That is no small consolation prize.

As a final note, based on my analysis, here are the games that I believe will have the biggest overall impact on this upcoming year's college football season:

9/5 Arizona State (19) vs. Texas A&M (22) = Early match-up between two of my teams that could surprise

9/12 Oregon (5) at Michigan State (9) = could be a playoff elimination game

9/12 Boise State (27) at BYU (45) = could be only hurdle between Boise, 13-0, and playoff berth?

9/26 USC (6) at Arizona State (19) = Winner looks likely to claim Pac-12 South crown

10/3 Alabama (2) at Georgia (7) = Battle of two best SEC teams. Somebody has to lose

10/3 Notre Dame (8) at Clemson (14) = winner has inside track to playoffs

10/3 Georgia (7) at Tennessee (18) = Could be one too many losses for the Bulldogs

10/17 Michigan State (9) at Michigan (9) = a loss to the Wolverines could knock Spartans out of playoffs

10/17 Alabama (2) at Texas A&M (22) = One of two games that could send SEC West into chaos

11/7 Florida State (12) at Clemson (14) = winner take all for ACC title and possible playoff birth

11/7 Auburn (11) at Texas A&M (22) = Second of two games that could send SEC West into chaos.

11/14 Oregon (5) at Stanford (20) = Winner looks likely to claim Pac-12 North crown

11/21 Michigan State (9) at Ohio State (1) = winner has inside track to Big Ten title and playoffs

11/28 Georgia (7) at Georgia Tech (21) = could spell one too many losses for the SEC to make playoffs

11/28 Baylor (4) at TCU (3) = winner has inside track to Big 12 title and playoffs

11/28 Notre Dame (8) at Stanford (20) = a win by the Irish here could put them at 11-1

11/28 Alabama (2) at Auburn (11) = possible SEC West winner take all, including play-off birth

That is about it for now. Who is ready for kickoff? I know I am...