OSU Odds to BCS Title Game: 60%
By Chris Stassen
[Editor's Note: For those of you not familiar with the name Chris Stassen, Chris is the developer of the software that is the-Ozone Forum and has been a valued contributor to our site. Chris is also a great sports fan a pretty fair country mathematician.
Chris has been calculating the odds of OSU reaching then national championship game this year. The day after OSU's loss to Illinois, Chris had the odds at under 2 per cent, assuming an OSU win over Michigan. Without that assumption, the odds were less-than one per cent.
Below is his current statistical analysis of OSU's chances of getting to New Orleans.]
Using Vegas line, and a research paper on college football simulation for the odds of winning (e.g., how likely a 3-point favorite is to win a game), I get 60% OSU to New Orleans. Here's the detail:
There are two teams slated to be in front of Ohio State when the dust settles: West Virginia and Missouri. Each has just one game remaining. The chances of Ohio State jumping West Virginia, or some other
team jumping Ohio State, seem so low as to not be worth considering at this point. The results this past weekend were a mix of favorable (LSU losing, Missouri winning as Kansas had a better shot at Oklahoma) and unfavorable (West Virginia winning a game they were very likely to win, but their only realistic chance at losing).
According to Stern, "On the Probability of Winning a Football Game":
1 point favorite - 53% likely to win
3 point favorite - 59% likely to win
5 point favorite - 64% likely to win
7 point favorite - 69% likely to win
9 point favorite - 74% likely to win
14 point favorite - 84% likely to win (not in original table, but given by sigma=14)
28 point favorite - 98% likely to win (ditto)
(I) West Virginia
West Virginia has Pittsburgh at home left on their schedule. They are prohibitive favorites and have almost no chance to lose barring an upset of USC-Stanford-type proportions.
West Virginia -27 vs Pittsburgh (0.97 West Virginia wins [P1])
Missouri plays Oklahoma at a neutral site, in the Big XII title game next week. Missouri is a slight underdog.
Missouri +3 vs Oklahoma (0.41 Missouri wins [P2])
Ohio State needs one of the two above things:
(a) West Virginia winning (97%, P1), and
(b) Missouri winning (41%, P2)
... to not happen.
Here are the 4 combinations:
(A) P1*P2 (40%) Ohio State finishes 3rd in BCS
(B) P1*(1-P2) (57%) Ohio State finishes 2nd to West Virginia
(C) (1-P1)*P2 (1%) Ohio State finishes 2nd to Missouri
Total for 2nd: 58%
(D) (1-P1)*(1-P2) (2%) Ohio State finishes 1st in the BCS
Making BCS title game requires finishing in the top two BCS, so that chance is the sum of scenarios (B) through (D), or 60%.
(I) Matchups and total team odds
Here are the probabilities for each possible matchup:
57% Ohio State vs West Virginia
40% West Virginia vs Missouri
2% Ohio State vs someone else (Georgia, LSU, etc.)
1% Ohio State vs Missouri
And here is each individual team's chance to make the game (these total 200% because there are two teams in the game):
West Virginia - 97%
Ohio State - 60%
Missouri - 41%
Someone else (both West Virginia and Missouri lose) - 2%
(II) Vegas vs Sagarin-Predictor
The percentage goes up to 71% if using Sagarin-predictor ratings to create
pointspreads, instead of Vegas lines. For earlier versions of this calculation that were
posted to the O-Zone Fan Forum, I had to come up with pointspreads for games that were
only hypothetical, such as Kansas vs Texas in the Big XII title game. Since no Vegas line
existed for those games, I used Jeff Sagarin's Predictor ratings instead. However, the
research paper by Stern is based on Vegas lines, so 60% is probably the better number.