What a week. Students are going to Leavey more to avoid the insufferable heat wave rather than study, the Southeastern Conference has a record 10 teams in the Associated Press Top 25 Poll and the Pac-12 overall had arguably the biggest opening week flop in recent memory. But fear not, because the Trojans are alive and kicking, and we’re back with more pick’em predictions!
Now I can tell your burning question right now is this: “How did the model do last week?” Well, thanks for asking —we had an 80 percent success rate last week! As good as that is, here at the Daily Trojan we like to keep things exciting, and we’ve been working on some changes to our model so we don’t have to include a pesky caveats section.
Updates to the Model:
Now that play is underway, we finally have data from this season we can incorporate into our model. But before we get ahead of ourselves, we’re still including the 2014 data in our model. Why? Well, throwing the 2014 data out in favor of Week 1’s data is foolish from a statistical standpoint because the small sample size of Week 1’s data would violate a basic assumption of our regression model. Thus, whether it is because they played against a Division-1AA team or because inclement weather canceled their game (or both — yeah, LSU we’re talking about you) some Football Bowl Subdivision (FBS) teams still do not have 2015 data for games against other FBS teams. For these teams we will continue to use 2014 data from games against other FBS teams because the talent difference between FBS and Football Championship Subdivision teams can result in inflated observations that would skew the model, except for the rather rare occasion (c’mon now, Washington State).
While the new data is exciting, what we’re really hyped about is our strength of schedule measure. We’re aggregating the weekly Football Power Index rankings of the opposing teams each team has played to create a variable that accounts for the level of difficulty of each team’s past games. After all, there is a big difference between scoring 40 points or earning 500 yards on offense against an unranked team and doing the same against a top 25 team.
One final cool thing that we added in is the spread for each game. For the less betting inclined, the spread is an estimate of the point gap between two teams in the final result. For example, the spread for the USC-Idaho game this week is -21. This can be interpreted as saying that our model predicts USC to win by three touchdowns over Idaho (not bad, right?).
With all that methodological mumbo-jumbo out of the way, let’s take a look at the matchups this week.
|Day||Time||Home Team||Away Team||Home Team Odds||Away Team Odds||Spread|
|Thu||05:00 PM||Western Kentucky||Louisiana Tech||33.06%||66.94%||6.5|
|Fri||06:00 PM||Utah||Utah State||96.82%||3.18%||-19.5|
|Sat||01:00 PM||Wyoming||Eastern Michigan||88.84%||11.16%||-16|
|Sat||01:30 PM||Iowa State||Iowa||51.68%||48.32%||-1|
|Sat||04:00 PM||Arkansas State||Missouri||1.51%||98.49%||19.5|
|Sat||04:00 PM||Florida||East Carolina||99.97%||0.03%||-21|
|Sat||04:00 PM||Southern Methodist||North Texas||43.38%||56.62%||2|
|Sat||04:00 PM||Texas A&M||Ball State||99.71%||0.29%||-20.5|
|Sat||04:30 PM||South Carolina||Kentucky||49.47%||50.53%||0|
|Sat||05:00 PM||Indiana||Florida International||75.37%||24.63%||-11|
|Sat||05:00 PM||Michigan State||Oregon||65.9%||34.1%||-7|
|Sat||05:00 PM||Nebraska||South Alabama||99.4%||0.6%||-20.5|
|Sat||05:00 PM||New Mexico||Tulsa||59.92%||40.08%||-4.5|
|Sat||05:00 PM||New Mexico State||Georgia State||43.07%||56.93%||2.5|
|Sat||05:00 PM||Southern California||Idaho||99.99%||0.01%||-21|
|Sat||07:15 PM||Air Force||San Jose State||56.03%||43.97%||-3|
|Sat||07:15 PM||Brigham Young||Boise State||69.84%||30.16%||-8.5|
|Sat||07:30 PM||Nevada-Las Vegas||UCLA||0.61%||99.39%||19.5|
|Sat||07:30 PM||Stanford||Central Florida||79.05%||20.95%||-12.5|
|Sat||08:30 AM||Florida State||South Florida||99.88%||0.12%||-20.5|
|Sat||09:00 AM||Michigan||Oregon State||93.28%||6.72%||-18|
|Sat||09:00 AM||Penn State||Buffalo||93.15%||6.85%||-18|
|Sat||09:00 AM||Texas-San Antonio||Kansas State||13.86%||86.14%||14|
|Sat||09:00 AM||Wisconsin||Miami (Ohio)||95.44%||4.56%||-19|
|Sat||09:30 AM||Clemson||Appalachian State||95.51%||4.49%||-19|
|Sat||09:30 AM||Syracuse||Wake Forest||64.38%||35.62%||-6.5|
|Sat||12:00 PM||Texas Tech||Texas-El Paso||92.29%||7.71%||-17.5|
|Sat||12:30 PM||Colorado State||Minnesota||9.32%||90.68%||16|
|Sat||12:30 PM||Georgia Tech||Tulane||98.07%||1.93%||-20|
|Sat||12:30 PM||Mississippi||Fresno State||99.68%||0.32%||-20.5|
|Sat||12:30 PM||Ohio State||Hawaii||99.94%||0.06%||-21|
|Sat||12:30 PM||Rutgers||Washington State||82.54%||17.46%||-13.5|
|Sat||12:30 PM||Virginia||Notre Dame||12.42%||87.58%||15|
After complaining of a week of cupcake openers, we’re finally rewarded with a redemption match with both teams eyeing a playoff spot down the road. If you only watch one game this weekend (not counting the USC game, of course), I suggest you watch this one. Michigan State will be led by veteran quarterback Connor Cook, and even without Marcus Mariota, Oregon has plenty of talented players to keep Michigan State busy.
In short, these are two talented teams. Each has a few question marks, especially defensively after shaky openers. But both are more than capable of earning an early slot in the College Football Playoff field. Our model has Michigan State as the favorite to win but it’s looking like this one could go either way.No. 19 Oklahoma at No. 23 Tennessee (Spread: -19.5)
Oh boy, a Big 12 vs. SEC matchup, ESPN must be so excited. But to be fair, both teams showed they have offenses capable of putting up good numbers. Tennessee quarterback Joshua Dobbs threw for 205 yards, ran for 89 yards and produced three touchdowns (two passing, one running). Additionally, running backs Jalen Hurd and Alvin Kamara combined to rush for 267 yards and score five touchdowns. Oklahoma’s quarterback Baker Mayfield threw for 388 yards and had three passing and one running touchdown, proving Oklahoma’s not a team to take lightly.
Both of these former powerhouses are looking to jump back into the limelight, and with their offense-oriented game plans, expect to see plenty of points in this one. But Tennessee’s dual threat offensive threat makes them the favorite in our model for this match up.
Idaho at No. 8 USC (Spread: -21)
If you’re a good Trojan fan, I suspect you already took a look at the odds for this week’s game. If you haven’t, look now. Our model has USC as a 21-point favorite, and that’s a conservative estimate compared to Vegas odds, which have us as a 43-point favorite.
Regardless, this game will be a good tune-up match before we begin conference play. Though as any Trojan fan who went through last season knows, overconfidence can come back to haunt you.
Data provided by cfbstats.com
Team logos courtesy of sportslogos.net