clock menu more-arrow no yes mobile

Filed under:

Against All Odds, Week 10: On the Margin

The Michigan State Spartans’ margin for error has been small all year. In West Lafayette on Saturday, MSU’s number finally came up.

NCAA Football: Michigan State at Purdue Trevor Ruszkowski-USA TODAY Sports

Let us not forget where we started this season. Back in the summer, the expectations for the Michigan State Spartans were limited. My annual summer analysis suggested that MSU was likely to win only four or five games total.

The odds of winning eight games or more was only eight percent and the odds of getting to a record of 9-3 or better was right around three percent. So-called experts picked Michigan State to finish last in the Big Ten East. The Spartans were marginalized.

Former Michigan State head coach Mark Dantonio used to say that football is “a game of inches.” Through the first eight games of the season, current MSU head coach Mel Tucker’s team had managed to find those inches. The Spartans made the winning plays when they mattered in the come-from-behind wins against Miami, Nebraska, Rutgers, Indiana and Michigan.

But, the margin of error for this team is small.

If a team is ranked outside of the top-50 at the beginning of the season, there is usually a reason. Coach Tucker and his staff did an excellent job of preparing the schemes and retooling the roster via an influx of marginalized talent out of the transfer portal. But, it is not like the Spartans have a dozen four-star prospects to roll off of the bench when needed.

Football is a fickle game. So far this year, the Spartans have been reasonable healthy and reasonably lucky. However, sometimes that oblong ball simply bounces the wrong way. By the time a team reaches November in Big Ten country, bumps and bruises tend to accumulate.

With several Spartans dinged up on Saturday, and with a lot of the 50-50 calls and bounces not going Michigan State’s way, the remaining Spartans essentially needed to be close to perfect. There was no margin for dropped passes, misplaced throws or fumbles. The defense needed to get stops on third downs, especially in the red zone. Those things simply did not happen. The final margin on the scoreboard highlighted what our eyes told us all night — Purdue was the better team on Saturday.

Now, all the Green and White can do is to try to move forward. At 8-1, there is still a lot to play for. MSU still controls its own destiny — in the Big Ten East, the New Year’s Six and even the College Football Playoff. Everything is still on the table.

Week Ten Results

As is my weekly tradition, Figure 1 below shows the results of all 59 games involving two FBS teams in Week 10.

Figure 1: Results of Week 10 showing the actual point differentials relative to the opening spread.

The teams that overachieved notably versus the opening spread this week include Wisconsin, Utah, East Carolina, Old Dominion, North Texas, Oklahoma State and Iowa State. As for the teams that had more marginal performances and underachieved, that list includes Alabama, Cincinnati and Ball State.

The upsets relative to the opening lines in Week 10 are summarized below in Table 1.

Table 1: Upsets based on the opening Vegas line compared to the upset projections from Week 10.

In total, there were 17 upsets relative to the opening Las Vegas spreads, which is slightly more than expected. While Michigan State was the highest-ranked team bitten by the upset bug, the Spartans were nowhere near the top of the board this week. Florida, Minnesota, California and Oregon State all lost as double-digit favorites.

My algorithm only got two upset picks correct out of seven total (29 percent), which brings the year-to-date performance down to 35 percent. The FPI did much better, going three-for-four (75 percent) and is sitting at a marginally better 37 percent success rate for the year.

Table 2 below gives the results of the computers’ picks against the spread (ATS) for Week 10.

Table 2: Results of the highlighted picks versus the spread in Week 10.

My computer was 4-4 for the week, bringing the year-to-date performance to just under 62 percent. For the first time in six weeks, the FPI was above .500 and actually got four picks out of five correct (80 percent). But, this strong performance only brought the FPI up to 50-50 for the year. Overall, the computers combined have been correct just under 55 percent of the time this year.

Overall, my computer was 30-29 ATS for Week 10, while the FPI did better at 35-24 (50 percent). However, both computers are still under .500 for the year. My machine is now at 256-277 (48 percent) while the FPI is marginally better at 258-275 (48.4 percent).

Updated Big Ten Odds and Expected Wins

Following the results of Week 10, I have re-run the full-season Monte Carlo simulation using the updated power rankings, including the current uncertainty in those rankings, in order to update the season odds for each team. Table 3 below gives the update for the Big Ten conference.

Table 3: Updated Big Ten rankings, expected wins, strengths of schedule and season odds.

Despite the Spartans’ upset loss, Michigan State is still very much in the thick of the Big Ten East race. Overall, Ohio State has a full game lead in the standings and the edge in the odds at 47 percent. Michigan (31 percent) is close behind, with Michigan State (22 percent) just a few percentage points back.

In the Big Ten West, there is now a four-way tie for first place between Wisconsin, Minnesota, Iowa and Purdue, as all of those schools are now 4-2 in conference play. Don’t be fooled, however, as this is now very much the Badgers’ race to lose. My math gives Wisconsin a 69 percent chance to return to Indianapolis.

Wisconsin holds the head-to-head tiebreaker against Iowa and Purdue and has only lost to Michigan and Penn State out of the East. Minnesota (13 percent) still controls its own destiny, but the Gophers will need to survive road tests at Iowa and Indiana and then (most importantly) beat Wisconsin in Minneapolis in the final weekend of the season.

The updated Big Ten win distribution matrix is shown below in Table 4.

Table 4: Updated Big Ten win distribution matrix after Week 10.

Michigan State’s expected win total dropped a bit to 9.68. More precisely, the Spartans’ odds to run the table are now at 16 percent and the odds to get to at least 10 wins are at 57 percent. Exactly nine wins has odds of 38 percent while losing out has only a five percent chance.

A closer look at Michigan State’s remaining schedule

With only three games remaining, I will forgo my usual bar chart figure and simply give the projected spread and win probabilities for the Spartan’s three remaining games as a bulleted list:

  • Maryland at Michigan State (-19.0), 92 percent odds to win
  • Michigan State (+8.0) at Ohio State, 29 percent
  • Penn State at Michigan State (+1), 48 percent

Basically, there is one likely win, one likely loss and a toss-up for Michigan State.

The actual line for next weekend’s game with Maryland opened with Michigan State as a 14-point favorite, which corresponds to an 84 percent chance that MSU gets to 9-1.

With three games remaining, there are still eight possible scenarios for Michigan State. One scenario where MSU is winless, one scenario where MSU goes undefeated, three scenarios where MSU finishes 9-3 and three scenarios where MSU finishes at 10-2.

But, only three of those scenarios has a probability over about five percent. One of those low probability scenarios is the situation where Michigan State loses out and finishes at 8-4. In that scenario, it is hard to judge how far MSU would fall in the bowl pecking order, but an invite to the Las Vegas Bowl or Music City Bowl would seem likely.

Based on the numbers above, the two most likely scenarios involve Michigan State beating Maryland, but losing to Ohio State. The results of the Penn State game will decide if MSU finishes at either 9-3 or 10-2.

If the Spartans do finish at 9-3, a New Year’s Six Bowl berth would be possible, but unlikely. Michigan State would likely wind up playing in either the Citrus Bowl or the Outback Bowl in that scenario.

But, if Michigan State is able to beat Penn State and get to 10-2, a berth in the New Year’s Six is very likely. If Ohio State wins out and makes the College Football Playoff, then it mainly comes down to whether Michigan beats Penn State next weekend. If Michigan and Michigan State both finish at 10-2, it will come down to their relative position in the final College Football Playoff poll as to which team will go to Pasadena.

I would love to tell you that the head-to-head win over Michigan would convince the committee to send the Spartans back to the Rose Bowl. I personally do not think that is how that decision would go down. If both teams are 10-2, Michigan State would likely fall to either the Peach or Fiesta Bowl, which is a pretty nice consolation prize.

Of course, there is also the roughly 15 percent chance that Michigan State beats Ohio State and wins out. In this scenario, the Spartans will return to Indianapolis. A win there would place them once again in the College Football Playoff. A loss would also result in a Peach or Fiesta Bowl berth (as the Big Ten West Champion would be sent to the Rose Bowl).

National Overview

Finally, let’s take a quick look around the country at the rest of the action. I am going to take a slightly different approach this week. Table 5 shows the division, conference, and season odds for all of the remaining contenders in the rest of the FBS.

Table 5: Updated rankings, expected wins, strengths of schedule and season odds for the remaining FBS contenders.

Despite Wake Forest’s loss to North Carolina, the race in both ACC Divisions is status quo, primarily because the Wake Forest/North Carolina game was considered a non-conference game, which did not impact the conference standings. As a result, Wake Forest and Pittsburgh still hold the edge in odds with North Carolina State, Virginia and Miami all still hanging around.

In Big 12 action, Oklahoma State and Iowa State won by large margins over West Virginia and Texas and thus hold down the top-two spots in the conference in my calculations. Oklahoma remains in first place in the standings, but third place in my calculations. Baylor dropped to fourth following an upset loss to TCU.

In the Pac-12, Oregon effectively eliminated Washington in the North, and Oregon State did themselves no favors by being upset by Colorado. This leaves Washington State as the only real threat to the Ducks. In the South, Utah has the division all but locked up, especially after they pounded Stanford last week.

In SEC action, Texas A&M’s defeat of Auburn means that the Aggies are the only real team with a shot to prevent the Alabama/Georgia showdown in the SEC Championship Game. Texas A&M will likely need Auburn to upset Alabama in order for this disruption to occur.

In Group of Five action, there were no upsets among the contenders, so the New Year’s Six leaderboard is largely unchanged. The slot is still Cincinnati’s to lose (47 percent) with Texas-San Antonio (31 percent) or Houston (11 percent) most likely to steal the bid if the Bearcats falter.

As for the College Football Playoff, my current top 15 leaderboard lists the top four teams as Georgia, Oklahoma State, Notre Dame and Michigan, with the Spartans currently down at No. 10. That said, I think that my strength of record calculation is currently giving a more reasonable estimation of the real standings. Table 6 below gives the results of this calculation.

Table 6: Strength of record calculation, based on my power rankings and strength of schedules

Michigan State is currently No. 6 in strength of record, just ahead of Michigan. The most notable aspect of both my playoff odds and strengths of record calculations is that my machine is giving more respect to Oklahoma State and Notre Dame than the humans are right now. We will see if that trend continues.

Against all odds, I have reached the end for this week. Until next time, enjoy, and Go State, beat the Terrapins!