Instead of picking players from various teams to see how well they do each week, my fantasy football is a little different. I have played a simulated expanded college football season and am about to start again with a second season.
This grew in part out of an idea I first had in April 2008 in a comment on the PFR blog.
The original plan took the 120 FBS teams and divided them into 10 conferences of 12 teams each, primarily keeping conference lines when possible. Some of the mid-majors got moved around, and the Big East was demoted to second tier and renamed Football East. No independent teams were allowed. Each team played a schedule of 29 games :
- 2 games against teams in their division
- 1 game against the other division in their conference, with a second game against the team ranked in the same position
- 1 game against each team in one division of an affiliated conference (PAC-10 & WAC, Big XII & MWC, Big Ten & MAC, SEC & C-USA, ACC & Football East)
- 1 game against each of 2 teams in the other division of the associate conference
- 4 nonconference games against teams on the same level in the same position in the other conference groups
- Each conference had a championship game at the end of the season between the winners of the two divisions.
So for example, Ohio State (Big Ten South) played Iowa (Big Ten North) twice, the rest of the Big Ten North once, the Big Ten South twice, the MAC Ohio division once, Central Michigan and Northern Illinois once each, with nonconference games against Florida, Texas, Virginia Tech, and USC.
The results of the games were determined as follows.
I researched the scores of all games played by FBS teams from 2007 to 2009, including bowl games, and rejecting games against FCS or non-D I teams. My source for the data was the ESPN website, since Sports Reference's College Football page was not available yet. Mean and standard deviations were computed for total points for&against, and point differential, considering home and away games separately. Neutral-site games, such as bowl games or traditional neutral site games (Texas-Oklahoma, Georgia-Florida, Army-Navy, Auburn at Alabama, etc.) were considered away games for both teams. If a team played its bowl game on its home field it was counted as a home game.
The normal random variable feature of the Microsoft® Excel™ spreadsheet program was used to determine the final scores. A team's score for a game was a (A+B+C-D)/4, where A, B, C, and D were normal random variables representing the following quantities.
- A was total points for and against by that team.
- B was total points for and against by the opponent
- C was net margin of victory for that team
- D was net margin of victory for the opponent
Let us consider an example game of USC at Notre Dame. USC in road games had an average sum of 49.33 with standard deviation of 16.99. Their average difference was 14.57 with S = 20.62. So this means their average road game was a (49.33+14.57)/2 to (49.33-14.57)/2 victory, or 31.95 to 17.38. For Notre Dame home games, the sum was 47.86 (S = 10.82) with a difference of 1.00 (S = 17.70), leading to a typical score of 24.43-23.43. Taking the raw numbers and no randomness, we find the score is USC (49.33+47.86+14.57-1.00)/4, NDme (47.86+49.33+1-14.57)/4, leading to a score of USC 27.69, Notre Dame 20.905, which is recorded as USC 28, NDme 21.
Scores were rounded to the nearest whole number; rare single digit scores such as 1 or 4 were rounded to an appropriate football score, and negatives were rounded to 0. In case of a tie, the game was determined to have gone into overtime; the fractional part of the numbers determined which team received the victory and which the Overtime Loss (reckoned in the standings as a tie). Example : Assume the USC-NDme game came out USC 25.08594, NDme 25.43757. Both of these scores round to 25, so the game goes into overtime. Notre Dame's fractional part is greater than USC's, so Notre Dame is credited with the win, USC the Overtime Loss, and the score is recorded as 25 OW-25 OL.
Standings were as follows:
|Championship: USC 29, Oregon 20|
|Championship: Utah 37, Fresno 28|
|Championship: Missouri 38, Texas 32|
|Championship: Air Force 26, TCU 15|
|Championship: Iowa 22, Ohio St 19|
|Championship: Ohio 36, Central Mich 17|
|Championship: Florida 30, Alabama 13|
|Championship: Troy 25, Louisiana Tech 16|
|Championship: Boston College 19, Virginia Tech 16|
|Championship: Connecticut 32, East Carolina 31|
The Playoffs will be discussed in a later post.