Leaves continue to fall off trees, days continue to get shorter, and the College Football Playoff continues to approach. This week I thought it'd be interesting to look at which teams control their own destiny regarding the College Football Playoff.
I added logic to the model that checks each simulation to see if a) a team wins the rest of their games and b) that team is selected for the CFP per my new and enhanced CFP model. From there, conditional probability lets us figure out which teams control their own destiny,
The likelihood a team makes the CFP if they win out (win the rest of their games) can be calculated by dividing the likelihood they win out AND make the CFP, by the overall likelihood they win out. For example, in 11.6% of simulated college football seasons, Florida wins the rest of their games (including the SEC title game if appropriate, excluding any bowl games). In 10.7% of simulated seasons, Florida wins the rest of their games AND qualifies for the CFP. That means we can say that 91% (10.7% / 11.6%) of the time that Florida wins the rest of their games, they are selected for the CFP.
I calculated these #s for any team with a reasonable shot at the CFP and threw them in the table below.
- Bama, Stanford, and Notre Dame all control their destiny as 1-loss teams
- All the undefeated teams in power 5 conferences control their own destiny
- Mid-majors need lots of help all over, as do mediocre 1-loss power 5 teams
- Both my model and Vegas lines favor unranked USC (slightly) over #3 Utah. Not sure what that says about AP voters
- Lots of potential for "upsets" this week, with lots of weaker ranked teams travelling to play stronger unranked teams:
- UCLA has a 66% chance to upset #20 Cal
- USC has a 55% chance to upset #3 Utah
- East Carolina is 48% to upset #22 Temple
- Virginia Tech is 43% to upset #23 Duke
- Georgia Tech is 32% to upset #9 Florida State
Everyone loves an upset.