Week 8 of College Football is in the books prompting the first round of Playoff Committee rankings! Until then, we’ll just have to rely on the AP poll, Forward Progress, and internet arguments to settle on which team is the best in the league. This week of games ended up fairly boilerplate with no major upsets as opposed to last week’s bloodbath. The computer didn’t predict any Top 25 upsets either, which is good for the algorithm but not as good for the viewers. In any case, there were some semi-major mispredictions that I’ll outline below.
- Memphis - 42 vs. Houston - 38
- Forward Progress predicted a win for the Cougars (by less than a point but a win nonetheless) but instead visitor Memphis won in Houston.
- Louisville - 31 vs. Florida State - 28
- Both of these teams have had very strange seasons thus far. Florida State has had a precipitous dropoff from last season on the back of several hurricane cancellations and Louisville’s athletic department is currently under investigation by the FBI for illegal payments to the basketball coach by equipment manufacturer Adidas. In both cases I think the algorithm isn’t getting a clear picture of what is going on as it predicted a 10 point win for FSU.
- Oregon - 14 vs. UCLA - 31
- The computer predicted a close win by Oregon but intead UCLA won by more than 2 touchdowns. I don’t exactly know what Forward Progress sees in Oregon since they haven’t exactly shown any hidden strengths but it was a miss nonetheless.
On to the rankings!
|Team||Predicted Wins||Ranking Date||Rank Movement|
|20||North Carolina State||208||2017-10-22||1|
Alabama remains at the top of the list and will probably stay there until they lose. Clemson is hanging on at number 2 despite their loss and may not leave that spot until last year’s championship game falls out of the data model entirely. Penn State has climbed two additional spots after a decisive win against Michigan. Ohio State, Oklahoma, and newcomer Virginia Tech(!?) are fighting for fourth. VT did have a blowout against North Carolina this weekend which likely explains the 7 spot jump. Notre Dame has also moved up after a good win against USC while Georgia, Miami, and Oklahoma State fall after lackluster wins or byes.
Overall the algorithm is hovering at around 65%-75% accuracy for calling game outcomes and it will be exciting to see how it compares to the Playoff Committee’s rankings next week.