Against All Odds, Week 5: Misery
Michigan State’s season has taken a frightening turn, but the numbers suggest the Spartans may still have a path out of their misery and into the postseason
It was a miserable weekend for Michigan State football. After a promising start to the season offered a few glimmers of hope, the Spartans hit a patch of ice in the Wisconsin backwoods and ran off the road.
Michigan State’s 31-3 loss to Wisconsin was painful to watch, and the prospects for a quick recovery appear grim. After four straight seasons of disappointment, the start of Big Ten play has left fans feeling isolated and alone, trapped in the same bleak story.
To some, the program feels broken, with no escape from the bottom of the Big Ten standings in sight. Is this the fate awaiting Michigan State, or is there still a way out of this nightmare?
Despite the horror show unfolding before us, I believe some hope remains, and a few plot twists may still be in the works. Michigan State was soundly defeated in Madison, but a few plays breaking the other way could have made the game much more competitive.
What if Jacob Harris drops the bobbled touchdown pass on third down of Wisconsin’s opening possession? What if Michigan State converts even half of its fourth-down attempts in Wisconsin territory? It would not have taken much for the game to remain competitive deep into the second half.
As we will see below, there is also reason to believe the Spartans’ past three opponents are all better than expected. Pat Fitzgerald noted after the game that Notre Dame, Nebraska, and Wisconsin are a combined 14-1, with the Badgers’ loss to the Fighting Irish the lone blemish.
What if all three are top-15 teams? Based on their performance on the field so far, my computer says that is exactly where they belong.
Yes, the current situation is terrifying, but the Spartans must remain patient, work hard, persevere, and, most importantly, heal their wounds.
At full strength, Michigan State has enough talent to compete with at least the middle tier of the conference. The Spartans, however, are far from healthy.
If Michigan State is going to bring this dark, bowl-less chapter to an end, it needs key players such as Jordan Hall back on the field as soon as possible. If that happens, there is still time to break free from this misery.
If they cannot, the Spartans can at least take solace in knowing they are not the only Big Ten team in the state to start conference play 0-2. Unlike another team nearby, MSU did not begin the season ranked with Big Ten title or playoff expectations.
It may be cold comfort, but misery loves company.
Week 5 Betting Results
Let us review last week’s Bad Betting Advice, beginning with the overview in Figure 1. For an explanation of how to read the figure, see the Week 1 edition of Against All Odds.
Figure 1: Week 5 results showing actual point differentials relative to the opening spread.
Ten teams overachieved in Week 5, including Wake Forest, Kansas, Memphis, Alabama, Arizona, and, unfortunately, Wisconsin.
No teams notably underachieved and still won, largely because every significant underachiever lost outright.
After four weeks with fewer upsets than expected, chaos came out of the shadows in Week 5, when 20 favored teams lost outright.
Table 1 summarizes the 20 upsets and compares them with last week’s picks. As we will see, the computers took a sledgehammer to the odds this week.
Table 1: Week 5 upsets based on the opening line, compared with last week’s upset projections.
The week’s biggest upsets were Purdue (+9.5) over Illinois, Minnesota (+6.5) over Michigan, NC State (+6) over Louisville, Temple (+6) over South Florida, Pittsburgh (+5.5) over Virginia Tech, Baylor (+4.5) over Arizona State, Missouri (+4.5) over Florida, Kentucky (+3.5) over South Carolina, Northwestern (+3.5) over Penn State, and Florida State (+3) over Virginia.
My algorithm finally had a good week, going an impressive 10-4 (71%) and lifting its year-to-date record to 18-26 (41%), much closer to its historical success rate of 43%.
FPI (7-0) and SP+ (4-0) were both perfect on upset picks this week. FPI is now 10-11 (48%) on upset picks for the year, and SP+ is 7-13 (35%).
Table 2 shows the computers’ results against the opening spread.
Table 2: Results of the highlighted Week 5 picks against the opening spread.
The computers also displayed an uncanny ability to see the spreads clearly.
My computer went 6-3 (67%) on its recommended picks, bringing the year-to-date record to 17-17 (50%). FPI was 5-2 (71%), but its year-to-date record remains below .500 at 21-26 (45%). SP+ has had the best season in this category. It went 3-2 (60%) this week, improving its overall record to 22-17 (56%).
Across all 59 games, my algorithm went 31-28 (53.5%) ATS, identical to FPI’s record. Both algorithms are now 196-194 (50.3%) for the season.
SP+ had an excellent week, going 37-22 (63%) and improving its year-to-date record to 205-185 (53%).
Table 3 shows the Week 5 point-total results.
Table 3: Results of the recommended Week 5 point-total wagers.
The good picks are still shining through. My computer went 3-1 (75%), improving its year-to-date record to 28-22 (56%). SP+ went 1-1, bringing its season total to 7-4 (64%).
My computer went 30-26 (54%) on the full slate of total bets this week, while SP+ posted a dismal 20-36 (36%). Year to date, both computers remain below .500 across all games. My algorithm is 160-163 (49.5%), while SP+ is just 153-170 (47%).
Updated Big Ten Odds and Expected Wins
After Week 5, I reran the full-season Monte Carlo simulation using updated power rankings and their current uncertainty. Table 4 shows the latest Big Ten odds, and Table 5 shows the conference’s updated win-distribution matrix.
All rankings beside team names refer to my computer’s power rankings, not either national poll.
The Week 5 betting results are in, but the most important numbers are still lurking behind the curtain. Subscribe to Spartans Illustrated for the updated Big Ten race, Michigan State’s possible escape route to the postseason, and the latest playoff projections from around the country.
