It may not seem like it this time of year, but I am actually
a bigger college basketball fan than I am a college football fan. A big reason
why goes all the way back to my childhood when I developed an absolute love of
March Madness. But, because I am kind of a nerd, I was originally almost drawn
to the bracket more than the actual games. Almost. I loooove brackets. More
recently, I have discovered a love of bracketology as well. Not in a Joe Lunardi
kind of way, but more in a DIY kind of way. A few years back, I stared to do my own
bracketology every March. I take it so seriously that I will specifically avoid
Joe Lunardi or anyone else’s work such that it does not “contaminate” my own
thinking.
So, three years ago, when the new Football Playoff System
was announced, complete with 4 additional games with would involve complex
selection criteria, I was pretty excited. When I realized that I could utilize my
spreadsheet and rankings to do further analysis on the season, leading up to the
ultimate goal of analyzing the “bracket” that is the NY6, I got even more
excited. I basically felt like this:
Ladies and Gentlemen, this is what I have been training and
preparing for.
So, for the last couple of days, I have been doing some
think and crunching some numbers and I think I have a pretty good handle of how
this weekend might go down. So, hang onto your hats and your slide-rules, and
let’s take one final trip around the country. The Spartans might not be playing
this weekend, but its Championship Week, baby. Let’s do this.
National Overview
Due to the relatively small number of games this week, I am
going to do things a little bit differently. Basically, I want to
explain the process that I went through to arrive at the results that I will
show below. What I first did earlier in the week was to try to rank the teams
that I predict to be NY6 contenders with a bit of a twist. I wanted to consider
each team that will be playing this week considering both situations where
either they win or lose. For example,
Clemson will either win or lose this weekend, so ranked them both as the ACC
Champ, 12-1 Clemson Tigers (which I would rank #1), and as the defeated, 11-2
Clemson Tigers (which I rank #10). When I considered this, I came up with the
following rankings:
1 Clemson (W)
2 Wisconsin (W)
3 Auburn (W)
4 Oklahoma (W)
5 Georgia (W)
6 Miami (W)
7 Ohio State (W)
8 Alabama
9 TCU (W)
10 Clemson (L)
11 Wisconsin (L)
12 Oklahoma (L)
13 Auburn (L)
14 Miami (L)
15 Georgia (L)
16 Ohio State (L)
17 Stanford (W)
18 USC (W)
19 Penn State
20 TCU (L)
21 Washington
Since Alabama, Penn State, and Washington are in the clubhouse,
they only appear once in the list. I also did not bother to rank the loser of
the Stanford / USC game, as I don’t expect the loser to remain in the NY6. Also,
based on the numbers, I expect any of the Championship losers will stay ahead
of Washington, so I project that either TCU or Penn State will draw the AAC
Champ in the final spot.
Now, we could endlessly debate this ranking. I put it
together using a combination of my power rankings and my best guess as to what
the committee will actually do. The toughest section is teams 7
to 12. Quite honestly, if I look at just the raw numbers, I am tempted to put
the losing Clemson and losing Wisconsin teams ahead of Alabama and the winning
TCU team. But, I am basically assuming that the committee is going to treat
each Championship game as a playoff elimination game and they are going to give
Alabama the benefit of the doubt. Using this assumption, the really hard
comparison is the one between Alabama, an Ohio State team that wins the B1G and
TCU, if they beat Oklahoma. My gut feeling is that the committee would bump OSU
above Bama, but TCU would get “Tide Blocked.”
Based on Tuesday night’s playoff rankings, I think my assumption about
TCU is correct, and the Bama – OSU question is still an open one. The only
other major difference between my ranking and the last committee ranking is
that they have Wisconsin a few slots lower than I do. I am not sure that this
makes a huge difference.
Anyway, if I may ask you to accept these rankings, this list is all that is needed to predict the final ranking (and estimate Bowl
pairs) based on all possible outcomes of the four Power 5 Championship games
that actually matter (sorry Pac 12). There are a total of 16 combinations. If
you know the point spread for each game (and I do) then you can also calculate
the probability of each possible outcome. As it turns out, that is precisely
what I did. To discuss the various combinations, let's discuss each conference in turn.
Group of Five / Independents:
At this point, I think that it is pretty clear that Notre
Dame is out of the race for a NY6 Bowl, but I do hear that the Camping World
Bowl is quite nice. As for the Group of Five, this is clearly down to the
winner of the Memphis vs. UCF (-7) game. Incidentally, my spreadsheet likes the
Knights by 13 points. I have been saying and assuming all year that the Group
of Five Representative would play in the Peach Bowl, but as I dove into each
scenario this week, I actually don’t think that is true anymore. Based on my
rankings, if TCU loses to Oklahoma, I think the AAC champ is likely to play TCU
in the Cotton Bowl. If TCU does beat Oklahoma, Penn State would draw the AAC
Champ and in this case, the Peach Bowl does make sense. Based on the playoff rankings, one could make
the argument that even Washington could pass TCU and grab the last NY6 slot. If
that does happen, I think they wind up in the Cotton Bowl anyway, but I would
argue that I think TCU has a better resume than the Huskies.
Pac 12: Stanford vs.
USC (-3, 58% chance to win)
Some people are making the argument that USC could sneak
into the playoffs with a win over Stanford. I do not agree. Basically, no matter how I slice it, the
winner of this game goes to the Fiesta Bowl and the loser is out of the NY6
(and likely into the Alamo Bowl, which I am sure is an experience that they
will never forget). Incidentally, my spreadsheet likes Stanford in the upset by
2 points.
ACC: Miami vs. Clemson (-7.5, 70% chance to win)
This one is easy as well. The winner of this game will make
the playoffs and the loser will play in the Orange Bowl, most likely. If Clemson wins, they
will finish the #1 team (based on Tuesday night’s rankings) and will play in
the Sugar Bowl. If Miami wins, I have them in the Rose Bowl in 6 of the 8
scenarios. For the record, my spreadsheet likes Clemson by 9.
Now, the bizarre scenario
that could happen is if Clemson were
to lose, yet only fall only to 4th place. In this scenario, 2 ACC teams would make the
playoffs and a vacancy would be created in the Orange Bowl. By contract, I believe
the Orange Bowl would need to take the next highest ranked ACC team, which is
9-3 Virginia Tech. This would also bump one other team off the bottom of the
NY6 ladder (likely either TCU or Penn State). There is an extremely low probability
scenario where the Playoff includes the 2 ACC teams and 2 SEC teams (the SEC Champ and Alabama). This would also mean that TCU beats
Oklahoma. In this scenario the loser of the SEC Title Game is likely ranked below Wisconsin (who would also need to lose for this scenario to come to pass). In this world (if it exists), Penn State would fall to the Outback
Bowl and MSU would wind up in San Diego in the Holiday Bowl. Based on the spreads, there is a 6% chance for ACC, Big Ten, and Big 12 games to end in this result and you would also need to assume that a B1G Champ Ohio State team would finish ranked below Clemson and Bama, which I think is not likely, but possible.
Big 12: TCU vs. Oklahoma (-7, 69% chance to win)
A week or so ago, I was thinking that the winner of this game might be in the playoffs either way, but the more I thought about it, it seemed clear that this statement only applies to Oklahoma. When the playoff rankings came out this week with TCU already below Penn State, the writing seems to be on the wall. If Oklahoma does win, as expected, they are certainly in the playoffs. If the Sooners get upset, in all scenarios I have them falling to the Fiesta Bowl to face the Pac 12 Champ. As for TCU, as I looked through all 16 scenarios, every single one of them has the Horned Frogs in the Cotton Bowl and the opponent varies. I suppose there is a very slim chance that TCU could blow out Oklahoma and rise to #4. In this case, Alabama would most likely by the team that loses out. Incidentally, despite the TD spread in Vegas, my spreadsheet likes the Frogs by 3.5 points in the upset.
SEC: Georgia vs. Auburn (-3, 58% chance to win)
The results of the SEC Title game seem very clear: the winner is in and the loser is out. The winner would wind up in either the Sugar or Rose Bowl depending on the other 3 teams in the playoffs with them. As for the loser, unlike Clemson, I see no chance that they remain in the Top 4. Once again, my spreadsheet is being a contrarian here and likes Georgia by 5. And then, there is Alabama. The single biggest question of the weekend is will Ohio State jump Alabama if they beat Wisconsin? I think the answer is yes. If this is true, then as long as Oklahoma beats TCU, Alabama stays at #5 and winds up playing in the Orange Bowl against the ACC Loser. In this set of scenarios, I would expect the loser of the SEC Title Game to be slotted into the Peach Bowl to face the loser of the Big Ten Championship (with PSU playing the Pac 12 Champ in the Fiesta Bowl and TCU playing the AAC Champ in the Cotton Bowl).
If TCU were to beat Oklahoma, it seems extremely likely that Alabama would slide up into the playoffs. In these scenarios, it gets a little messy as the SEC Loser and Big Ten Loser would then get slotted into the Orange Bowl and the Cotton Bowl. The higher ranked team would be placed in the Orange Bowl and the lower ranked team would wind up playing TCU in the Cotton Bowl. Based both on my rankings, if Ohio State wins the Big Ten and my assumption is correct that they make the playoffs, Wisconsin would still be the higher ranked team and they would play in the Orange Bowl. If Ohio State loses, it seems likely that they would drop into the Cotton Bowl.
If the committee in their less than infinite wisdom does opt to take Alabama over a Big Ten Champ OSU, the Buckeyes, as B1G Champs cannot play in the Orange Bowl and would likely be slotted into the Peach Bowl. In this case, Wisconsin would have to be ranked above the SEC Loser and placed in the Orange Bowl, otherwise the committee would have a hard time avoiding a Big Ten - Big Ten Bowl match-up.
Big Ten: Wisconsin vs. Ohio State (-5.5, 65% chance to win)
Through the discussion above, I think that I have covered pretty much every combination. If Wisconsin wins, as my spreadsheet suggests that they will (by 2), then the current playoff rankings suggest they are in the playoffs. As stated above, I personally think OSU is in the playoffs with a win as well, but that is not a sure thing. In general, as long as the Big Ten Champ is not ranked #4, they will play in the Rose Bowl, as the Good Lord intended. As mentioned above, as long as Alabama does not make the playoffs, the Big Ten Loser seems likely to wind up in the Peach Bowl to face the SEC Loser. If Bama does make the playoffs, Wisconsin (assuming they lose) seems headed for the Orange Bowl while Ohio State seems likely to play in the Cotton Bowl.
Wisconsin will enter the Big 10 Title game as the #4 team in the country. This is lower than I have them and I have to wonder if they will stay that way with a win over OSU. I find it hard to justify a 2-loss SEC Champ (for example) being ranked ahead of an undefeated Big Ten Champ who just beat the mighty Buckeyes. Then again, perhaps this the committee's evil plan. If I were them, my #1 goal for the 2nd to last poll is to set up a situation where they can have the flexibility to put teams in the match-ups that they want in the final poll. Depending on the match-ups on the 1-3 line, I could see the committee "tweaking" things with some level of plausible deniability. They can keep Wisconsin at #4 if they like the match-ups, or they can slide them up if needed, which the argument that they needed to see the Badgers beat a good team on the neutral field in order to become believers
As for Penn State, as stated above, their fate seems also tied to the Big 12 Title Game. If Oklahoma wins, I think they face the Pac 12 Champ in the Fiesta Bowl. If TCU wins, Oklahoma would drop to the Fiesta Bowl, leaving Penn State in the Peach Bowl to play the AAC Champ. I suppose that it is possible that OSU could look really bad against Wisconsin and actually drop below Penn State in the final standings. I seriously doubt that they would drop below Washington and out of the NY6, but in this scenario, I would simply swap OSU and PSU's bowl placements.
Add It Up
Of the 16 total scenarios that I mapped out, the most likely one (predictably) is the one where all the favorites win (Clemson, Oklahoma, Auburn, and Ohio State). The odds of this scenario occurring are not as high as you might think. I have them at only 18.2%. If we assume the committee would take OSU over Alabama, the Bowl pairing would likely be:
Sugar Bowl: #1. Clemson vs. #4 Ohio State
Rose Bowl: #2 Auburn vs. #3 Oklahoma
Orange Bowl: Alabama vs. Miami
Peach Bowl: Wisconsin vs. Georgia
Fiesta Bowl: USC vs. Penn State
Cotton Bowl: TCU vs. UCF
But, does the committee really want a Clemson / Ohio State rematch? Would they consider nudging the Buckeyes up to #3 considering a win over Wisconsin would seem on paper to be more impressive than a win over TCU? How much of a role will margin of victory play? All good questions.
Now, if the committee does take Alabama over Ohio State, the pairings would likely be:
Sugar Bowl: #1. Clemson vs. #4 Alabama
Rose Bowl: #2 Auburn vs. #3 Oklahoma
Orange Bowl: Wisconsin vs. Miami
Peach Bowl: Ohio State vs. Georgia
Fiesta Bowl: USC vs. Penn State
Cotton Bowl: TCU vs. UCF
Now, my spreadsheet actually likes upsets in 4 of the 5 conference championship games, where for some reason it only likes Clemson to win. Based on the Vegas spreads, I calculate this scenario as only having a 3.2% chance of happening, but if this were to come to pass, I would have the playoff pairings as follows:
Sugar Bowl: #1. Clemson vs. #4 Alabama
Rose Bowl: #2 Wisconsin vs. #3 Georgia
Orange Bowl: Auburn vs. Miami
Cotton Bowl: TCU vs. Ohio State
Fiesta Bowl: Stanford vs. Oklahoma
Peach Bowl: Penn State vs. UCF
Finally, since I can calculate the probability of each of the 16 scenarios, I can also calculate the probability of each team playing in each bowl. That I show in the following matrix. Keep in mind that these probabilities assume that my ranking of teams is correct, and I have no way to handicap what the committee will actually do:
Well, that about does it. I think that this is likely my final post in this series for the 2017 season, although I may have some other stats or observations to share. I hope that my musings over the past several months have been informative and entertaining. Until next time. Go Green!
A week or so ago, I was thinking that the winner of this game might be in the playoffs either way, but the more I thought about it, it seemed clear that this statement only applies to Oklahoma. When the playoff rankings came out this week with TCU already below Penn State, the writing seems to be on the wall. If Oklahoma does win, as expected, they are certainly in the playoffs. If the Sooners get upset, in all scenarios I have them falling to the Fiesta Bowl to face the Pac 12 Champ. As for TCU, as I looked through all 16 scenarios, every single one of them has the Horned Frogs in the Cotton Bowl and the opponent varies. I suppose there is a very slim chance that TCU could blow out Oklahoma and rise to #4. In this case, Alabama would most likely by the team that loses out. Incidentally, despite the TD spread in Vegas, my spreadsheet likes the Frogs by 3.5 points in the upset.
SEC: Georgia vs. Auburn (-3, 58% chance to win)
The results of the SEC Title game seem very clear: the winner is in and the loser is out. The winner would wind up in either the Sugar or Rose Bowl depending on the other 3 teams in the playoffs with them. As for the loser, unlike Clemson, I see no chance that they remain in the Top 4. Once again, my spreadsheet is being a contrarian here and likes Georgia by 5. And then, there is Alabama. The single biggest question of the weekend is will Ohio State jump Alabama if they beat Wisconsin? I think the answer is yes. If this is true, then as long as Oklahoma beats TCU, Alabama stays at #5 and winds up playing in the Orange Bowl against the ACC Loser. In this set of scenarios, I would expect the loser of the SEC Title Game to be slotted into the Peach Bowl to face the loser of the Big Ten Championship (with PSU playing the Pac 12 Champ in the Fiesta Bowl and TCU playing the AAC Champ in the Cotton Bowl).
If TCU were to beat Oklahoma, it seems extremely likely that Alabama would slide up into the playoffs. In these scenarios, it gets a little messy as the SEC Loser and Big Ten Loser would then get slotted into the Orange Bowl and the Cotton Bowl. The higher ranked team would be placed in the Orange Bowl and the lower ranked team would wind up playing TCU in the Cotton Bowl. Based both on my rankings, if Ohio State wins the Big Ten and my assumption is correct that they make the playoffs, Wisconsin would still be the higher ranked team and they would play in the Orange Bowl. If Ohio State loses, it seems likely that they would drop into the Cotton Bowl.
If the committee in their less than infinite wisdom does opt to take Alabama over a Big Ten Champ OSU, the Buckeyes, as B1G Champs cannot play in the Orange Bowl and would likely be slotted into the Peach Bowl. In this case, Wisconsin would have to be ranked above the SEC Loser and placed in the Orange Bowl, otherwise the committee would have a hard time avoiding a Big Ten - Big Ten Bowl match-up.
Big Ten: Wisconsin vs. Ohio State (-5.5, 65% chance to win)
Through the discussion above, I think that I have covered pretty much every combination. If Wisconsin wins, as my spreadsheet suggests that they will (by 2), then the current playoff rankings suggest they are in the playoffs. As stated above, I personally think OSU is in the playoffs with a win as well, but that is not a sure thing. In general, as long as the Big Ten Champ is not ranked #4, they will play in the Rose Bowl, as the Good Lord intended. As mentioned above, as long as Alabama does not make the playoffs, the Big Ten Loser seems likely to wind up in the Peach Bowl to face the SEC Loser. If Bama does make the playoffs, Wisconsin (assuming they lose) seems headed for the Orange Bowl while Ohio State seems likely to play in the Cotton Bowl.
Wisconsin will enter the Big 10 Title game as the #4 team in the country. This is lower than I have them and I have to wonder if they will stay that way with a win over OSU. I find it hard to justify a 2-loss SEC Champ (for example) being ranked ahead of an undefeated Big Ten Champ who just beat the mighty Buckeyes. Then again, perhaps this the committee's evil plan. If I were them, my #1 goal for the 2nd to last poll is to set up a situation where they can have the flexibility to put teams in the match-ups that they want in the final poll. Depending on the match-ups on the 1-3 line, I could see the committee "tweaking" things with some level of plausible deniability. They can keep Wisconsin at #4 if they like the match-ups, or they can slide them up if needed, which the argument that they needed to see the Badgers beat a good team on the neutral field in order to become believers
As for Penn State, as stated above, their fate seems also tied to the Big 12 Title Game. If Oklahoma wins, I think they face the Pac 12 Champ in the Fiesta Bowl. If TCU wins, Oklahoma would drop to the Fiesta Bowl, leaving Penn State in the Peach Bowl to play the AAC Champ. I suppose that it is possible that OSU could look really bad against Wisconsin and actually drop below Penn State in the final standings. I seriously doubt that they would drop below Washington and out of the NY6, but in this scenario, I would simply swap OSU and PSU's bowl placements.
Add It Up
Of the 16 total scenarios that I mapped out, the most likely one (predictably) is the one where all the favorites win (Clemson, Oklahoma, Auburn, and Ohio State). The odds of this scenario occurring are not as high as you might think. I have them at only 18.2%. If we assume the committee would take OSU over Alabama, the Bowl pairing would likely be:
Sugar Bowl: #1. Clemson vs. #4 Ohio State
Rose Bowl: #2 Auburn vs. #3 Oklahoma
Orange Bowl: Alabama vs. Miami
Peach Bowl: Wisconsin vs. Georgia
Fiesta Bowl: USC vs. Penn State
Cotton Bowl: TCU vs. UCF
But, does the committee really want a Clemson / Ohio State rematch? Would they consider nudging the Buckeyes up to #3 considering a win over Wisconsin would seem on paper to be more impressive than a win over TCU? How much of a role will margin of victory play? All good questions.
Now, if the committee does take Alabama over Ohio State, the pairings would likely be:
Sugar Bowl: #1. Clemson vs. #4 Alabama
Rose Bowl: #2 Auburn vs. #3 Oklahoma
Orange Bowl: Wisconsin vs. Miami
Peach Bowl: Ohio State vs. Georgia
Fiesta Bowl: USC vs. Penn State
Cotton Bowl: TCU vs. UCF
Now, my spreadsheet actually likes upsets in 4 of the 5 conference championship games, where for some reason it only likes Clemson to win. Based on the Vegas spreads, I calculate this scenario as only having a 3.2% chance of happening, but if this were to come to pass, I would have the playoff pairings as follows:
Sugar Bowl: #1. Clemson vs. #4 Alabama
Rose Bowl: #2 Wisconsin vs. #3 Georgia
Orange Bowl: Auburn vs. Miami
Cotton Bowl: TCU vs. Ohio State
Fiesta Bowl: Stanford vs. Oklahoma
Peach Bowl: Penn State vs. UCF
Finally, since I can calculate the probability of each of the 16 scenarios, I can also calculate the probability of each team playing in each bowl. That I show in the following matrix. Keep in mind that these probabilities assume that my ranking of teams is correct, and I have no way to handicap what the committee will actually do:
Well, that about does it. I think that this is likely my final post in this series for the 2017 season, although I may have some other stats or observations to share. I hope that my musings over the past several months have been informative and entertaining. Until next time. Go Green!
Comments
Post a Comment