Last week in this space, I presented the results of my annual football preview. I performed the complete analysis using the original, full schedule. But, as expected, the original schedule has now been wiped away. In perhaps a desperate, last ditch attempt to save college football, the Big Ten officially announced a new, 10-game schedule this past week. The other conferences are announcing similar moves. The original schedule is no more... probably.
Everything is now reset. Assuming, for now, that this new schedule will actually get played out... at some point, what does the Big Ten race look like now? What has changed? Are MSU’s chances to finish over .500 or even win the East Division or conference better or worse? Fortunately, I can use the same analytic tools as before with the new schedule now in place. So, what’s new?
Let’s begin with a review of the changes to the schedule, which are summarized below in Table 1.
As was already announced, the Big Ten cancelled all non-conference games, and added a 10th conference game to the schedule of all 14 Big Ten schools. At this point, I think that it is most informative to clarify the changes that were made. Table 1 shows the new opponent that was added to each team’s schedule. Big Ten East teams all added a road game against a Big Ten West team, and vice versa.
In addition to added game, a few additional changes were made. Most MSU fans by now have heard about the change in location of the contest with Michigan from East Lansing to Ann Arbor. Further details from the Big Ten suggest that this change was made in large part to balance the home and away schedule such that MSU now gets a home game with either Michigan or Ohio State each year for the foreseeable future.
That is good news for MSU season ticket holders going forward. In order to complete this change, a third team had to also adjust its schedule with Michigan and MSU as well. That team is Indiana, who will now come to MSU this year again, but will get the chance to host Michigan. They likely don’t care. In the West, a similar three-team schedule shift was put into place involving Purdue, Wisconsin, and Nebraska. It is not clear why this change was made other than it added “flexibility” to the schedule... OK...
Updated Strength of Schedule
Looking at Table 1, an interesting pattern appears in the selection of the extra team added on each schedule. By looking at the preseason consensus rankings, the selection of the added team has the look of the NCAA Basketball tournament “S-curve.” In other words, they took the strongest team in the East (Ohio State) and added the weakest team in the West, not already on their schedule (Purdue). The No. 2 team in the East (Penn State) got Illinois. Michigan (No. 3 in the East) gets Northwestern.
MSU already had Minnesota, Iowa, and Northwestern on the schedule, so the Spartans get Nebraska. Minnesota already had Maryland on the schedule, so they add Indiana. Wisconsin, as the clear No. 1 in the West gets Rutgers. Almost by default, Iowa gets Maryland. It is so obvious now that I see it, that I am not sure why I haven’t seen this reported previously. It had to be intentional.
In any event, I can use the same methodology that I used before to recalculate the conference strength of schedule. Figure 1 below compares the strength of schedule of each Big Ten team using the new schedule and comparing it to the previous schedule. As a quick reminder, my strength of schedule metric represents the expected number of wins that an average (ranked roughly No. 25) power five team would have with the same schedule. So, more “wins” indicates an easier schedule.
As the graph shows, in the new Big Ten schedule, there are four teams with a noticeably easier road: Minnesota, Penn State, Ohio State, and Wisconsin. An average power five team would likely win about six-and-a-half games with any of those teams’ schedule. These same four teams made up the top four in the easy schedule category in the old schedule as well, but Wisconsin moved into first place overall due to the addition of a home game with Rutgers. Note in general that these teams have the built-in advantage of not having to play themselves.
For the middle eight or so teams in the conference, including both MSU and Michigan, the revised schedule difficulty is comparable at approximately 5.6 wins. Michigan’s schedule is slightly easier (5.81) than MSU’s schedule (5.57), but only by 0.2 games, which is a small difference. In a relative sense, Iowa made the biggest move in the schedule rankings, moving from 13th place up to 6th place by adding a home game with Maryland.
At the bottom of the rankings are Maryland, Illinois, and Purdue. They all had schedules on the harder end of the spectrum to begin with (in part because they also do not get to play themselves) and then they added games at Iowa, vs. Penn State, and vs. Ohio State. That is not ideal. Over the full scale, Wisconsin’s revised schedule projects to be over one-and-a-half wins easier than Maryland’s schedule.
Updated Conference Odds
Using the updated schedule, I was able to re-run my Monte Carlo simulation of the new Big Ten season, using the same consensus preseason rankings, and by taking into to account the historical uncertainty in those preseason rankings. The results of that simulation are shown below in Table 2.
Table 2 has a lot of detail. For each team, I show the updated expected win total, the raw strength of schedule numbers (as shown in Figure 1) and the updated odds for the division and conference races. The values in parenthesis either shows the change from the original simulation, or the conference rank for the strength of schedule numbers.
The most striking result from Table 2 is that the odds for the conference races really don’t change that much. The biggest change in the table is Wisconsin’s odds to win the West nudged up by a little over one percentage point. Iowa and Penn State saw similar, small improvements in odds. Considering the “S-curve” strategy that appears to have been used to modify the schedule, it is no surprise that “the rich got richer” in a relative sense.
It should also be noted that the three-team-venue shuffle on the western side of the conference between Purdue, Nebraska, and Wisconsin seems to have had a small, but visible effect on the West Division odds. Purdue now has to travel to Madison and added Ohio State, which is bound to make things harder. In contrast, Nebraska now gets to host Wisconsin and added MSU as a home game.
In regards to MSU, the new schedule is a net negative, without question. But, the raw odds to win the division and conference don’t drop by that much. For the Division Title, the odds moved from 1 in 60 to about 1 in 70. For the odds to win the whole Conference, they moved from 1 in 150 to about 1 in 180. Down the road in Ann Arbor, the Wolverines odds for each race dropped by about half a percentage point.
Updated Win Distribution
Finally, the simulation results also can be used to project the odds that each team will win any given number of games. The updated win distribution table is shown below in Table 3.
Similar to the original simulation results, the odds suggest that it will be challenging for any team to run the table in a 10-game Big Ten schedule. Even Ohio State’s odds are less than 20 percent (due in large part to the existence of scenarios where Ohio State is not as good as expected, due to injuries and other unknowns). Even with this uncertainty, it’s 50-50 that Ohio State wins at least nine games. For Penn State, those odds are about 40 percent. For Wisconsin, it’s 35 percent.
As for MSU, the expected win calculation suggests MSU will win 3.87 games. Practically, and more specifically, this translates to a 55 percent chance that MSU wins at least four games, and a 45 percent chance that MSU wins three games or fewer. So, that’s not great.
It is a bit tricky to make a direct comparison between the previous schedule and the new one, but a reasonable benchmark is the odds to win half of the games on the schedule. In the old schedule, which included some more winnable non-conference games, MSU’s odds to win six game were about 45 percent. In the new conference-only schedule, the odds to win five games fall to 36 percent.
As for the more optimistic scenarios, the data suggest that MSU’s odds to go 7-3 or better are right at 10 percent. The odds to go 8-2 or better are 4 percent. In these scenarios, MSU would certainly need some good breaks. A solid quarterback would likely need to emerge, as well as likely some pleasant surprises on the defensive side of the ball. MSU would need to stay very healthy and would need to get some good bounces. A little bad luck for the opponents in places like Ann Arbor, Columbus, or State College would also help.
But, history has shown that this type of chaos happens all the time. Sports are unpredictable, which is why they are so much fun. The way I see it, the results of my simulation simply tell us how much luck a team like MSU needs to have a satisfying, or even a very satisfying season. It sure seems like we are all due for some good news and good fortune.
All this number crunching and analysis, however, remains agnostic to the reality that the true odds of a college football season this fall seem to be falling. As players continue to opt out, the preseason rankings (which form the basic input to my simulation) are becoming less and less reliable. It seems fair to say that more uncertainty about the future of college sports exists today than at any point in the last 75 years.
All that we can do as fans is to wear and mask, social distance, and try to be nice to each other. As for the future, it remains unknown. As for me, as always, I choose to be optimistic.
Stay safe and Go Green.