This was designed by Jack Cackler, a doctoral student in Biostatistics at Harvard University.
To make winter travel plans with better information, this site graphically illustrates of all of the bowls a team might play in, how likely they are, and where and when each bowl will be.
Click the radio buttons to switch between viewing by team or by bowl, and mouseover any bar to see the chance of that event occurring.
Feel free to contact me at email@example.com.
With only a week of games and the Army-Navy game to go, the bowl schedule is mainly settled out. The final projections gel nicely with many of my earlier predictions, though it's been a learning experience, and I've added a few features to make it more accurate along the way. This week I've tried to go get every edge case to make sure predictions are reasonably accurate. I've tried to clean up the layout to make it easiest to use.
There are currently exactly 70 teams eligible for a bowl. Georgia Tech is unique in that, if it wins, it will go to a BCS bowl, and if it loses, it may not even be eligible with a 6-7 record. Given that Miami-Florida and North Carolina are both ineligible this year, Georgia Tech has petitioned for a waiver, which will likely be accepted, granting them eligible for a bowl whether or not they beat Florida State. Connecticut and Pittsburgh are also on the bubble, and if they win may be eligible. This leaves up to three teams that are eligible without a bowl game, and which could leave Central Michigan, Western Kentucky, Bowling Green, and Baylor on the chopping block.
There are two main priors that were added in manually. One of the main limitations of this algorithm is that, even controlling for conference strength, the algorithm only assigns teams based on strength, and not based on external factors that may sway a bowl. For example, if Nebraska beats Wisconsin to earn a trip to the Rose Bowl, Northwestern would be the third team in the Big Ten, and the model predicts that the Outback Bowl would select them. Despite a good showing this year by the Wildcats, I doubt the Outback Bowl would pass up a chance at the lower-ranked Wisconsin, who has twice the fan base. It would be difficult to control for all of these cases, but a few others that come to mind are the Cotton bowl preferring Texas A&M to LSU, who ended up ranked higher or the Alamo Bowl taking USC or UCLA over Oregon State.
One I did control for was the Sugar Bowl. With the BCS Championship assuredly featuring Notre Dame and the winner of Alabama and Georgia, the Sugar Bowl will have the first pick of teams to replace the SEC Champion. In many cases, the highest rated available team was Oregon, but especially since the Sugar Bowl did not have an SEC team last year, it's almost certain they would take an SEC team. This would almost certainly be Florida, though there is a slim chance, should Alabama lose, that the Sugar Bowl will pick them over Florida. I find this unlikely, as Florida is already ranked higher in the computers, and if Alabama loses to Georgia I imagine they'll drop in the polls, too. So as a prior, Florida gets the SEC spot for the Sugar Bowl. As an additional note, the SEC championship is played in the Georgia Dome in Atlanta, which is technically not home field for Georgia. Given the proximity to Athens, I did give Georgia home field advantage in the SEC Championship, which boosted their chances from 24.897% to 32.393%.
The second prior addresses a thorn in the model all season: the chance that an MAC team ends up with a BCS bid. Both Kent State and Northern Illinois are 11-1, and are currently ranked 17 and 21. With both Rutgers and Louisville losing this past weekend, the winner will end up higher than the Big East champion, and therefore only needs to be in the top 16 to guarantee a spot. Of the remaining games, six other games are relevant to what place the MAC champion ends up in.
Oklahoma (11), Nebraska(12), Florida State (13), Oregon State(15), UCLA (16), and Texas (18) all play this week. If any two of them lose, I have to imagine they would drop below the winner of the MAC championship, as all are playing unranked opponents except for UCLA, who is just barely ahead of Kent State. Thus, I added a prior that, if any two of the above teams lost, the MAC champion would get a BCS Bowl, but if one or zero lost, they would not. This may not completely cover all possibilities, but I believe it to be a pretty good assessment of the situation. Currently, Oklahoma has a 74.066% chance of beating TCU, Nebraska has a 56.318% chance of beating Wisconsin, Florida State has a 85.684% chance of beating Georgia Tech, Oregon State has a whopping 99.997% chance of beating Nicholls State, UCLA has a 24.487% chance of beating Stanford, and Texas has a 28.878% chance of beating Kansas State. A MAC fan's best hope lies in rooting for at least Stanford and Kansas State to win, and in total the MAC Champion is projected to make a BCS Bowl 83.8% of the time.
With Oregon and Florida taking the open spots in the Fiesta and Sugar Bowl, and whoever wins the Big East taking the last spot in the Orange Bowl, in those 83.8% of samples, the MAC Champion will end up in the Sugar Bowl (again, the Sugar Bowl could decide that Rutgers/Louisville would be a better choice, but this model does not account for that). In the remaining 16.2%, almost all of the time the final spot is taken by the second place Big Twelve team, either Kansas State or Oklahoma. A microscopic .6% of the time, Clemson may have a chance, but in order for this to happen, Oklahoma has to lose, so that Clemson can jump Oklahoma in the rankings, and the other five aforementioned teams have to win so that the MAC does not get a BCS Bowl. This could happen, but is highly unlikely.
These predictions were generated largely based on the pure points predictor ratings created by Jeff Sagarin, and will be updated weekly.
For each game remaining in the schedule a probability of winning was determined based on the rating of each team and home field advantage.
Based on these predictions, each game was sampled 10,000 times and for each sample, conference championship games and subsequently bowl games were computed.
Overall ratings were given by an Elo rating system used by the computer rankings in the BCS standings. For each team, the number of times it ended up in each bowl was counted and is given as a percentage. The maximum standard deviation of any estimate is .5%. Bowls that occurs .1% of the time or less are grouped as "Other", and are theoretically possible, but highly unlikely
There is also a prior distribution for each conference's poll ranking versus computer ranking, essentially adjusting each team's rating up and down by a number of points depending on two factors. The first was the overall strength of the conference, derived from the Central mean of their Sagarin Pure Points rating. The second was, looking back at the end results since 2005, how teams poll rankings were compared to their computer rankings. These were averaged by conference, and conferences were maximally adjusted up or down half a game against a neutral team.
In other words, an ACC team that went 10-2 would be comparable with an MAC team that went 11-1 against the same teams, as the ACC is an average conference that does very well in polls compared to computer rankings, and the MAC is a mediocre conference that does not do well in polls.
The precision of the end result will be lower, and so you can uncheck the "Use Poll Data" box to view results purely by computer predictions, but the results are probably much closer to what will actually happen.
Each team that plays this week is also displayed with their imputed likelihood of winning this week's game.
The predictions are robust, but have several limitations that could impact accuracy. The prior I used has been trained on 8 years of data, but may have limited accuracy..
As such, these predictions will over-rank teams with good numerical rankings but poor poll standings.
While most bowl seats have contracts with conferences, in some cases a bowl may choose any available team if no team they are contracted to is eligible.
These predictions simply assign the highest rated team not yet in a bowl, but there may be instances in which a bowl might pick a lower-ranked team that they believe has a larger fan base.
Analysis coded in R, graphs constructed with the D3 Library.