Quick Odds Update: On to something NEW

When it comes to Michigan State basketball, I am the eternal optimist. I never count out a Tom Izzo-led team until the final buzzer. While a big part of this attitude is simply my personality, the facts tend to justify my rosy attitude. 

Afterall, Michigan State has won a Big Ten regular season or tournament title in exactly half of the seasons going back to 1998. Coach Izzo has made the Sweet 16 at a higher rate than that (61 percent of the time) and has advanced to the Final Four in a little over a third of all of the NCAA Tournaments in which he has appeared.

Furthermore, as I have shown, it is not all uncommon for the Spartans to hit the skids in late January and early February. But the typical pattern is for Michigan State to recover and surge by now. I fully expected that this Spartan team had the goods to still be in the Big Ten race in the final week of the regular season. I was wrong. With the loss to Illinois over the weekend, it is time to shift focus to simply getting better in preparation for the postseason.

As we await the tip off for tonight's contest with Iowa, let's take a look a quick look at the updated numbers for the rest of the Big Ten season, starting with the updated enhanced Big Ten standings shown below in Table 1.

Table 1: Enhanced Big Ten standings as of Feb. 22, 2022

Michigan State is currently sitting in sixth place in the standings. The Spartans are a half game behind Rutgers and a full game ahead of both Iowa and Michigan. In addition to the game tonight in Iowa City, Michigan hosts Rutgers on Wednesday night. The results of these two games will help to sort out the final placement of those four teams in the Big Ten standings. 

Note that Michigan State has now slipped into negative territory in the "luck" metric (-0.13 games), but the Spartans are nowhere near as unlucky as Iowa (-2.04 games). On the other side of the coin, the very lucky Wisconsin Badgers (+3.06) and Rutgers Scarlet Knights (+2.92) may do overdue for a late-season slump.

The updated Big Ten win distribution matrix is shown below in Table 2. 

Table 2: Updated Big Ten expected wins and win distribution matrix as of Feb. 22.

As we approach the finish line of the regular season, some separation in the standings is apparent. The top three teams (Purdue, Illinois, and Wisconsin) are in the top tier, with Ohio State now a game back and comfortably in fourth place. 

The next four teams (Rutgers, Iowa, Michigan State, and Michigan) are closely packed with expected win totals between 11 and 12. Indiana and Penn State are a full two to three games back in expected wins, and then there is small gap that separates the bottom four teams in the conference: Northwestern, Maryland, Minnesota and Nebraska.

The raw numbers suggest that despite Michigan State's tough schedule to close out the regular season, there is still a 74 percent chance that the Spartan win at least two more games (to finish at 11-9 or better). There is also a 36 percent chance for MSU to get to 12 wins and a nine percent chance that the Spartans get to 13 wins. At least 13 wins will likely be needed for any chance to sneak back into the top four teams in the conference.

Table 3 below gives the updated odds at least share the Big Ten regular season title as of Feb. 22.

Table 3: Updated odds to win or share the Big Ten title and the win distributions for those winning teams.

The story being told by Table 3 is very similar to the one told by the win matrix. There is about a 50-50 chance that at least one team from the group of Purdue, Illinois, and Wisconsin will win out and claim at least a share of the regular season title. If not, a 15-5 record will almost certainly be good enough to hang a banner, most likely concurrent with one or two other Big Ten teams.

Ohio State is still in the race, but they are starting to fade with odds now down to seven percent. Rutgers and Michigan State are the only other conference teams to record a conference title in any of the 100,000 simulations that I ran overnight. Rutgers' current odds are about 1-in-750 while Michigan State's title odds are around 1-in-1,500. In other words, yes, I am saying that there still is a chance.

Jockeying for Seeds

With the Big Ten regular season race coming to a close, it is time to shift focus onto the Big Ten and NCAA Tournament. To this end, Table 4 shows the updated Big Ten Tournament seeds matrix.

Table 4: Big Ten Tournament seeding odds as of Feb 22, 2022.

As it stands now, Michigan State is most likely to earn the No. 7 seed in the Big Ten Tournament with the No. 8 seed very much a possibility if the currently projected favored teams all win out. The top five and bottom six seeds are starting to look sorted out, but seeds No. 6 to No. 8 are all very much in play.

To this end, tonight's game with Iowa is critical, as it is the only meeting between the two schools and the result could become a seeding tiebreaker at the end of the season. A few simulations suggest that a win over Iowa would mean a 76 percent chance that Michigan State ends up with a top six seed in the Big Ten Tournament. A loss would mean a 71 percent chance that the Green and White drop to the No. 7 or No. 8 seed.

As for the NCAA Tournament seeding, the recent slide from the Spartans' record has also caused Michigan State to slide down the seed list of the various online mock brackets. Both ESPN and CBS now have the Green and White projected as a No. 6 seed.

When it comes to "bracketology" there are a variety of different metrics that are used to try to seed the teams into the final bracket. In the coming weeks, college basketball fans will start to hear more and more about NET rankings and "Quad One" wins. 

While these metrics have their value, I have concerns about the usefulness of the NET ranking system specifically. For this reason, last year I created my own results-based ranking system that I refer to as the NEW index (because it is based on Normalized Expected Wins). 

The idea is actually quite simple, as I explained in detail last year. The NEW index measures the number of wins that a team actually has relative to the expected wins for an average power five team playing the same schedule. 

Tables 5 and 6 below introduce the current top 70 teams, based on my NEW rankings. Big Ten teams are highlighted in blue. Michigan State is highlighted in green, and various teams from mid-major conferences have their names shaded in light green. For reference, the current NET and Kenpom rankings are shown, along with the initial seed list released by the tournament selection committee over the weekend.

Table 5: Initial NEW rankings and selected "nitty gritty" data as of Feb. 22, 2022 for the top 35 teams

Table 6: Initial NEW rankings and selected "nitty gritty" data as of Feb. 22, 2022 for the teams ranked 36 to 70.

There is a lot of information in these two tables, but I will simply point out a few things. In general, the NEW rankings correlate well with the NET rankings and with the initial seed list presented by the selection committee. I would also argue that the NEW index does a better job than the NET in ranking some of the teams.

For example, the selection committee had both Purdue and Wisconsin placed higher on their seed list than either teams' NET or Kenpom ranking would suggest. Both teams score higher on my NEW index. In contrast, Houston currently is ranked in the top five of the NET and in the top 10 in Kenpom. However, the selection committee mentioned over the weekend that Houston is currently in a position to earn a No. 5 seed. The NEW index comes to the same conclusion by ranking the Cougars No. 19.

As for Michigan State, the Spartans are currently ranked No. 30 in my NEW index, which is very similar to the current NET ranking of No. 29 and Kenpom ranking of No. 27. What is useful is that I can use my NEW index to project where Michigan State might wind up depending on how the rest of the season plays out.

In the absolute worst-case scenario that the Spartans lose out (and finish at 18-14 overall), I can project that Michigan State would fall in my NEW index to around No. 65. This is right in the area of the rankings where Michigan and Indiana are currently located. This would place the Spartans squarely on the tournament bubble with eight consecutive losses and 10 losses in the final 11 games. 

While many of those losses would be to other tournament teams, this nightmare scenario would likely send the Spartans to the NIT. Fortunately, the odds of this scenario playing out are also quite low (perhaps around two percent).

If Michigan State finishes out with two more regular season wins and then goes 1-1 in the Big Ten Tournament (which the math suggests is the "expected" result), the Spartans NEW rankings would stay about the same in the lower 30s. I would project that in this scenario, Michigan State would likely earn a No. 6 or No. 7 seed. If MSU can win three of the next five games, I think that a No. 5 seed starts to be more likely.

Finally, if we would like to consider the very optimistic scenario where Michigan State gets hot, wins out, and somehow wins the Big Ten Tournament, how high of a seed might the Spartans be able to claim? Based on my projections, in this case the Spartans could rise to as high as No. 10 in my NEW rankings. This would likely result is a No. 3 seed or even a No. 2 seed depending on what happens elsewhere.

That said, this result is honestly less likely than the one where MSU tanks and fails to make the tournament at all. But, as March approaches, as MSU fans it is OK to dream big. The past 25 years have given us permission to be optimistic.

That is all for today. Until next time, enjoy, and Go State, beat the Hawkeyes!