Week Nine Recap (Trick or Treat!)

Spooky things have a tendency of happening on Halloween. Michigan State once had a terrifying trip to Minneapolis on Halloween night back in the John L. Smith era. In the lead up to this week’s game in Ann Arbor, all signs were pointing to another nightmarish afternoon for the Spartans at the hands of our eternal nemesis: the archvillain Michigan Wolverines.

The opening line in some places was a scary 25-points, which gave MSU only about a four percent chance of pulling an upset. An upset of that magnitude happens only slightly more than once in any given season, which is only a bit more frequent than a blue moon (which also took place Saturday night). 

A blue moon on Halloween? Now that’s rare. That happens only about once every 19 years, which corresponds roughly to the frequency of Michigan winning the Big Ten, beating Ohio State, or generally being relevant nationally. Maybe they should call that a “maize-and-blue moon?” But, I digress...

For those paying attention, there were signs that perhaps MSU’s trip to the Big House would not be the horror-show that many, including I, predicted. On Friday afternoon, the line suddenly dropped to 21.5-points. Then, on Friday night Minnesota played at Maryland and looked more like a MAC team than the ranked Big Ten West contender that we expected a few weeks back. Was it possible that Michigan simply wasn’t as good as the Wolverines appeared?

Even if that were true, we all saw what happened last week. Even if Michigan was overrated, it seemed like the Spartans required a near miraculous improvement, especially in the trenches, just to keep it close. Unless someone in the Chemistry Building was secretly working after hours on a Ironman-style suit or super soldier serum, it seemed futile.

And yet, somehow the Spartans went into Michigan’s house and took all of the candy.

The raid on the candy bowl was led by a young freshman wide receiver named Ricky White. At times he looked like Superman out there which makes sense as he sometimes wears a block “S” on his chest. Also in the party was Rocky Lombardi, this year dressed as Rocky Balboa. All afternoon, he took some punches, but also landed his fair share of body blows to the Wolverine secondary. 

The Spartan defense was led by a masked Antjuan Simmons, whose Mandalorian costume was so convincing that Wolverine quarterback Joe Milton didn’t seem to recognize him at all. Even Coach Harbaugh got into the act and decided to come to the game as the Joker. You could tell by the way he failed miserably to both correctly wear a mask in combination with a headset and to properly manage the clock in late-game situations. It’s his superpower.

When the clock hit all zeros in the fourth quarter, the Spartans had pulled the ultimate trick on the Wolverines. By beating the over 20-point spread, MSU scored the biggest upset over the Wolverines in the series this century, and likely the biggest since 1990. Meanwhile the Spartans got the treat of being able to bring the Paul Bunyan Trophy back to East Lansing for the ninth time in the last 13 tries. 

The win is like feasting on a selection of king-sized Reese’s Peanut Butter Cups, Kit Kits, and Butterfingers while your oafish, prematurely balding older brother who thought he was too cool to trick-or-treat only got a bag full of circus peanuts, stale candy corn, and a couple of soft apples. Victory is indeed sweet.

Just to put the current era in perspective, MSU’s nine wins since 2008 now equal the total number of wins that MSU accumulated during the previous 40 years from 1968 to 2007. Michigan fans had convinced themselves that the recurring nightmare that was Mark Dantonio (dressed as Freddy Krueger) was finally over. After last week, it was all but a foregone conclusion in Ann Arbor. 

But now, Mel Tucker did something that even Coach Dantonio failed to do in Year One: earn a victory over the Wolverines. Our maize-and-blue clad neighbors spent the summer counting stars on websites and repeating the numbers “44-10” to each other until they had convinced themselves that the rivalry was over and that they had won. 

But, it’s not over. It will never be over. If last Saturday’s game is any indication, Coach Tucker is going to be on the sidelines in East Lansing for a while and the Wolverine’s Nightmare on South Main Street will continue for the foreseeable future. MSU’s 21st century renaissance wasn’t just a one-off thriller. It’s looking more like a bona fide franchise, and sometimes, the sequel is even better than the original.

Whatever you do, Wolverine fans, don’t fall asleep.

Week Eight Results and Betting Review

Figure 1 below gives a visual summary of the overall results of Week Nine compared to the opening Vegas lines:

Figure 1: Summary of the results of Week Nine relative to the opening Vegas lines

This week, the favored teams that overachieved the most relative to the opening Vegas line include Cincinnati, Tulane, Duke, Auburn, San Diego State, Oklahoma, West Virginia, and UCF.  Meanwhile Tulsa and Clemson both underachieved significantly, yet still managed to win.

Table 1 below summarizes the upset results from Week Nine.

Table 1: Upset results from Week Nine relative to the computer upset picks

For the second time in two weeks, MSU was involved in the biggest upset of the week. This time, it was much more pleasant. In total, there were 11 upsets relative to the opening Vegas line, which is dead-on the number predicted in my weekly simulation. 

As for the computer upset predictions, my algorithm went three-for-five this week (60 percent) bringing the year-to-date tally to 7-9 (44 percent). The FPI went four-for-seven (57 percent) and now stands at 7-3 (70 percent) since I started tracking its performance last week.

Table 2 below summarizes the result of my suggested bets for the week. Two of the original games had to be removed from the board due to cancellations, leaving only six predictions total.

Table 2: Summary of the results for Week Nine's recommended bets

Both computer systems had a 50 percent success rate this week (1-1 for my algorithm and 2-2 for the FPI) resulting in a combined performance of three-for-six. After two weeks, my computer is 2-4 for recommended bets (33 percent) while the FPI is an impressive 8-5 (62 percent). This brings the combined results to 10-9 (53 percent). 

Note that the three highest confidence picks (once the cancelled Wisconsin game was removed from the list) also happened to be three games where the predictions were correct. 

Also, it is of course both ironic and hilarious that my failed pick for the week was for Michigan to cover the 23-point spread versus MSU. On Sunday, I had a long heart-to-hard drive talk with my computer about last week’s predictions. The desktop wanted to express its deepest regrets to Spartan Nation and has agreed to spend the week working on prime factoring of very large numbers as penance.

Updated Expected Wins and Season Odds

A week ago, things seemed very dire. While most people expected MSU to finish with a record of around 3-5, the loss to Rutgers left Spartan fans seriously questioning whether MSU would go 0-8. Now, MSU’s prospects for the 2020 season seem much brighter. But, how bright are they?

A team is typically never as bad as their last big loss and never as good as their last big win. MSU has now had one of each. With only two games in the book during a global pandemic, there is still a lot that we do not know about MSU, the Big Ten, and college football landscape. But, we do know a bit more than we did last week.  

In the face of this continued uncertainty, I always lean on the cold, hard numbers to keep things in perspective. In this spirit, Table 3 shows the updated power rankings, expected win totals, strengths of schedule, division, conference, playoff, and national title odds based on a 50,000 cycle simulation of the remaining games in the 2020 season.

Table 3: Updated Big Ten expected win, strengths of schedule, and season odds after Week Nine

As for MSU, the Spartans did shoot up 23 slots in my power rankings to No. 60, but this is still a bit below the position at the beginning of the season. Basically, the one awesome win does not quite offset the one terrible loss in the realm of mathematics, even if most MSU fans would trade a win over Michigan for a loss to Rutgers any day. MSU’s expected win total did jump up by a win-and-a-half to 2.88. 

As for the season odds, the best that I can say is that MSU still has a very, very small shot at glory. The odds of a Big Ten East title are at 0.71 percent while the odds to win the Big Ten stand at 0.21 percent. MSU even won the national title in a grand total of five simulations out of 50,000 (0.01 percent). In other words, yes, there still is a chance.

For the rest of the Big Ten, Ohio State now has a stranglehold on the East, with an 87 percent chance to return to Indy. The Buckeyes also hold the best odds in the Big Ten to both make the playoffs (59 percent) and to win the national title (17 percent). Indiana is solidly in second place in the East with a 10 percent chance to win the division, while Michigan’s odds dropped to a shade below two percent.  

In the West, the race is certainly harder to handicap due to the COVID-19 situation in Madison of which my computer is completely unaware. Based on the raw math so far, the Big Ten West now looks like a dead heat between Wisconsin (46 percent odds) and Northwestern (42 percent). While I am not personally convinced that the Wildcats are actually good, they have ascended to No. 13 in my power rankings and also have the easiest overall schedule in the West. 

Additional detail about the number of wins to expect for each Big Ten team is shown below in Table 4.

Table 4: Updated Big Ten win distribution matrix following Week Nine

As shown above, MSU’s expected win total is back up to a value approaching that projected in the preseason. That said, the most likely single result is that MSU will finish with only one more win at 2-6. However, the odds that MSU finishes with a record of 3-5 or better (58 percent) is now higher than the odds of finishing 2-6 or 1-7. 

Meanwhile, in Ann Arbor, the Wolverine’s most likely result is a 3-5 finish in Big Ten play with only 25 percent chance to finish over 0.500. That said, with Wisconsin on the schedule is just two weeks, there is still a lot of uncertainly with the remaining schedule.

What's Next for MSU

Similar to last week, the expected win distribution matrix for MSU can be deconstructed by looking at the odds for MSU to win each of the remaining seven scheduled games. The simulation provides these odds, which can then by used to back out a projected spread. Here are the updated projections following Week Nine’s action:

• MSU (+8.5, 28 percent) at Iowa. Although Iowa is 0-2 overall, the Hawkeye’s two losses were by a total of five point to teams that are currently undefeated (Purdue and Northwestern). Sure, it is possible that none of those teams are actually any good and that MSU will continue to play the way that they did against Michigan. If so, I like MSU’s chances. But, the truth is that no one really knows. The actual line seems to have opened at Iowa -9, so once again, my simulation is quite accurate.
• Indiana at MSU (+6.5, 32 percent). Indiana so far looks like the real deal and a potentially scary match-up. The Hoosiers draw an angry Michigan team next week, which might play to MSU’s favor. That said, MSU still projects as an underdog at home right now.
• MSU (-9.5, 76 percent) at Maryland. While the Terps managed to steal a win this weekend over the Gophers, my computer was not impressed. It feels like Minnesota is just really bad and thus MSU’s odds at College Park actually went up a bit since last week.
• Northwestern at MSU (+8.5, 28 percent). Last week, Northwestern projected as a double digit favorite over MSU, so it is encouraging that the projected line is below 10 points now. 
• Ohio State at MSU (+19, 9 percent). MSU’s odds here are improved a bit, but staying within three touchdowns on the Buckeyes this year still looks really hard. Then again, I said that last week about Michigan...
• MSU (+15.5, 14 percent) at Penn State. Will the season even make it this far? I don’t have the odds for that projection, but MSU’s odds to win at State College have gone up 10 percentage points since last week.

While MSU only projects to be favored in the road game at Maryland from here out, the next three easiest games on the schedule (at Iowa, vs. Indiana, and vs. Northwestern) now all certainly look winnable. The odds that MSU beats at least one of those three teams is about 65 percent. If MSU continues to play the way that they did this past weekend, I think that MSU’s numbers will continue to improve. 

Playoff Landscape

This week I wanted to show my current odds for various teams to make the college football playoffs (assuming that we make it to the end of the season). Table 5 shows the 23 remaining teams that I calculate still have greater than a one percent change to make the playoffs. Right now, the usual suspects of Ohio State, Alabama, Clemson, and Georgia reside in the top-four.

Table 5: Playoff odds for teams with over a one percent chance following Week Nine

Finally, I have also estimated the odds that each of the Group of Five teams will earn a spot in the New Year’s Six Bowls. Right now, Cincinnati and Marshall are at the top of my leaderboard.

Table 6: Updated odds for the Group of Five slot in the New Years Six Bowls after Week Nine

That’s all for now. Enjoy the win, Spartan fans. We get to enjoy a yet another full year of bragging rights. Until next time, Go State, beat the Hawkeyes!