Just a few days ago, it seemed likely that much of the focus this week was going to center on if the Michigan State Spartans were going to make the NCAA Tournament and if so, could they avoid the First Four pair of games reserved for the “last four teams in.”
But, MSU went out and beat the No. 2 ranked Michigan Wolverines on the last day of the regular season to all but clinch a ticket to Big Dance. Basically all prognosticators agree that even if the Spartans drop their first Big Ten Tournament game to the Maryland Terrapins, the Spartans will see their names in the bracket on Sunday night.
However, what seed is MSU likely to receive? Who are some likely opponents? What is the best and worst case scenarios? I would like to explore these questions today. But before we do that, let’s get an update on the “newly” created results-based metric, the NEW Index.
This weekend, I introduced my “performance relative to Normalized Expected Wins” metric, as a simple, transparent, and more accurate replacement for the NCAA NET metric. Briefly, the NEW Index measures the number of expected wins (based on spreads derived from Kenpom efficiency data) that an average power five team should receive if that average team were to play any given team’s schedule. It then compares the actual number of wins for the team in question to that expectation. This results is then divided by the total number of games played to get a marginal win percentage.
My initial analysis of the current expert NCAA Tournament projections suggests that the NEW Index is more accurate than the NET in selecting and seeding teams. It will be interesting to see how it does in the days leading up to Selection Sunday. For now, here is a brief update on the current NCAA Tournament resumes for the top-10 Big Ten teams, including the updated NEW rankings.
The projected seeds are based on the current information on the bracket matrix website and the “NEW Seeds” are based simply on a direct application of the rankings to the s-curve seed list. In other words, the top four teams in the NEW Index get No. 1 seeds, the teams ranked No. 5 to No. 8 get No. 2 seeds, etc.
The NEW Index is in complete agreement with the experts on the seeds for Illinois and Michigan (No. 1 seeds) as well as for Iowa and Ohio State (No. 2 seeds). Interestingly, the NEW Index is currently very high on Purdue and has the Boilermakers also as a No. 2 seed while most bracketologists have them as a No. 4 seed. While I have a hard time believing that the Big Ten would get five out of the top eight seeds, I think that a No. 3 seed is in play for Purdue if the Boilermakers can make the Big Ten semifinals.
As for the rest of the conference, there is some disagreement between the NEW Index and the current seed projections. The NEW Index is more favorable for Wisconsin (No. 6 instead of a No. 7 seed) and Rutgers (No. 7 instead of a No. 9 seed), but it is less favorable for Maryland (No. 12 seed instead of a No. 10 seed).
As for Michigan State, following the Spartans’ win over the Wolverines, MSU shot up to No. 24 in the NEW Index, which would be good enough for a No. 6 seed. This is certainly quite a difference and I would tend to believe the experts current projection of a No. 11 seed is much more accurate. However, it would not shock me if the committee winds up placing the Spartans a little higher on the seed list than folks like Joe Lunardi will be projecting on Selection Sunday.
The formula to calculate Michigan State’s (or any other team’s) NEW ranking is so simple that it is possible to estimate the future ranking of each team based on future outcomes. For example, what happens if MSU were to lose to Maryland or what would happens if MSU were to win the Big Ten Tournament? I can perform these simple calculations and see where the Spartans might wind up. Those are summarized here:
- If MSU loses to Maryland => NEW ranking drops to No. 36 (No. 9 seed)
- If MSU beats Maryland, and losses to Michigan => ranking improves to No. 22 (No. 6 seed)
- If MSU beats Maryland and Michigan, but loses in the semifinals to Ohio State => ranking improved to No. 17 (No. 5 seed)
- If MSU beats Maryland, Michigan, and Ohio State, but loses in the finals to Illinois => ranking improves to No. 11 (No. 3 seed)
- If MSU wins the Big Ten Tournament with wins over Maryland, Michigan, Ohio State, and Illinois => ranking improves to No. 8 (No. 2 seed)
First, I know that this sounds ridiculous. There is no way that MSU could go from a team barely on the bubble to a No. 2 seed over the span of a week...right? But, think about the resume the Spartans would assemble in this process. MSU would finish the Big Trn Tournament with nine quad-one wins and would have gone 6-1 combined against teams currently ranked in the top six of the seed list in the last 10 games of the season.
What would the committee do with MSU in that situation? Only 11 teams total had nine or more quad one wins prior to the tournament in 2019 on a full schedule. Besides, if MSU did somehow accomplish this feat, they would very clearly be the hottest team in the country. NO ONE would want to face the Spartans. How low of a seed could the committee actually give them without disadvantaging other teams in the field?
So, what is MSU’s ceiling? We have no actual way of knowing what the committee would think of such a run, but I think that a No. 4 or No. 5 seed would be a strong possibility with a Big Ten Tournament Title. But, as I have calculated, the odds of MSU winning four games in four days to claim the title are quite low (one to 15 percent at the absolute best).
As for the other scenarios, it is impossible to predict what the committee will actually do, but I have the following guesses, assuming no major upsets that would lower the value of each win:
- MSU loses to Maryland => No. 11 or No. 12 seed
- MSU beats Maryland, but loses to Michigan => No. 9 seed or No. 10 seed
- MSU beats Maryland and Michigan => No. 7 seed or No. 8 seed
- MSU loses in Finals = No. 6 or No. 7 seed
- MSU wins the Tournament => No. 4 or No. 5 seed
That is a crazy amount of variation, but I think that it is fairly reasonable.
With this uncertainly, there is a wide range of possible opponents that Michigan State could face in the first few rounds of the NCAA Tournament. That said, it is possible to take a quick look at the field to get an understanding of what the bracket might look like.
One visual tool that I find useful is to compare the efficiency of each team to the historical Kenpom efficiencies of teams with the same seed. That comparison is shown below in Figures 1 and 2 based on the seeds for each team as listed by the bracket matrix.
In these plots the historical average and standard deviation for the Kenpom adjusted efficiency margin of each seed, from No. 1 to No. 16 is shown with the blue dots with the error bars. I have also added the efficiency for each team projected to be in the 2021 field in the position of their current projected seed. This allows us to quickly see if a team is strong or weak relative to past teams with the same seed.
For example, if we look at the No. 1 and No. 2 seeds, we see that the teams in line to receive these seeds are almost all above average or well above average. The only team that sticks out as being simply average is Alabama. The Crimson Tide currently project as the weakest of the No. 2 seeds.
With this introduction, let’s look at a few possible scenarios for MSU in more detail.
MSU draws a No. 12 seed
I wanted to first say that I personally think that it is pretty unlikely for Michigan State to wind up as a No. 12 seed. Based on the current distribution of seeds, in order to wind up as a No. 12 seed, the Spartans would likely need to drop back down into the “last team in” category, and I just don’t think that would happen, even if MSU losses badly to Maryland, which I do not expect.
That said, if this were to happen, MSU would likely face one of the other bubble teams in the First Four:
As Figure 2 shows, all of these teams have pretty similar efficiencies that are all slightly above average for No. 11 or No. 12 seeds. MSU should beat all of these teams easily, as long as the Spartans play at a level similar to what we have seen over the past few weeks. The only exception might be Georgia Tech, who looks to be a little tougher, but who I also do not expect to be in the First Four.
As Figure 1 shows, almost all of these teams also seem to be above average relative to historical efficiencies. The lone exception is Oklahoma State. So, if the Spartans were to find themselves in this part of the bracket, Kenpom data suggests that the Cowboys would be the easiest opponent in this group.
MSU draws a No. 11 seed
If Michigan State loses to Maryland, this might be the most likely scenario, and honestly it might be the one that puts the Spartans in the best possible position to make a March run.
Possible first round opponents (No. 6 seeds): USC, Tennessee, Oregon, and Oklahoma
While all four of these teams would likely be favored over MSU, none of these teams particularly scare me. USC and Tennessee have the stats of above average No. 6 seeds, but Oregon and Oklahoma both look potentially ripe for an upset.
Possible second round opponents (No. 3 seeds): Arkansas, Texas, Kansas, and West Virginia
One other thing that sticks out about Figures 1 and 2 is that most teams in 2021 are projecting to be above average. Personally, I doubt that this is true in reality. I think that this may be a symptom of the condensed schedules and storage of inter-conference games.
That said, to the extent that we can trust any of this data, Figure 1 does imply that as a group the No. 3 seeds are a bit weak. Of those four teams, Arkansas looks slightly better, but honestly I would be very happy with any of those teams as possible second round opponents. Furthermore, I think that we would see a lot of prognosticators on Sunday night putting MSU into the Sweet 16 with this draw.
MSU draws a No. 10 seed
If MSU wins only one game in the Big Ten Tournament, a No. 10 seed seems likely, in my opinion.
Possible first round opponents (No. 7 seeds): Wisconsin, Clemson, BYU, and San Diego State
The bracketing principles released by the NCAA clearly says that MSU cannot face a Big Ten opponent such as Wisconsin in the first round, so the Badgers are off the table. Based on Kenpom data BYU and San Diego State are relatively strong, but Clemson is a possible weak link. They would be an opponent to look for if MSU were to wind up on this line.
Possible second round opponents (No. 2 seeds): Alabama, Iowa, Ohio St., and Houston
Here is another situation where the bracketing rules may come into play. Michigan State played both Iowa and Ohio State twice in the regular season. As a result, MSU cannot face those teams until the Sweet 16. Therefore, if MSU does draw a No. 10 seed (or a No. 7 seed) and if this seed list holds, the Spartans would be lined up to potentially face either Alabama or Houston in the second round.
Based on the efficiency data, facing Alabama projects to be an easier draw than Houston. That said, Houston played a very weak schedule this year and I have my doubts that the Cougars are actually as good as their efficiency numbers imply. I would be okay to face either team in the second round, if that situation were to come to pass.
MSU draws a No. 8 or No. 9 seed
If Michigan State does beat Maryland, it is certainly possible that the Spartans would get getting bumped up to the No. 9 or No. 8 line in the tournament. This in not ideal, as the Spartans would almost certainly draw a No. 1 seed in the second round were they to advance. While MSU has already gone 2-1 against other projected No. 1 seeds in the last two weeks, I would prefer not to face another one this soon.
Maryland and Rutgers are also floating around near this part of the bracket, but once again it would violate bracketing principles to face a Big Ten team in the first round. Of these possible opponents, many of them appear to be above average, but Missouri and Virginia Tech stick out as the weakest of this bunch, based on Figure 1 and 2 above.
Possible second round opponents (No. 1 seeds): Gonzaga, Baylor, Michigan, and Illinois
Interestingly, the bracketing principles prevent MSU from being placed in Michigan’s region as a No. 8 or No. 9 seed (as they played twice already), but there is the possibility that MSU could be placed in the same region as Illinois, since those teams only faced off once. Based on the tremendous flexibility that the committee has this year with no consideration for geography, this would gross incompetence and malpractice on the part of the committee... but I cannot rule that out entirely.
That said, if MSU is to draw a No. 8 or No. 9 seed, it is basically a coin flip between being placed in a bracket with Baylor or Gonzaga. Which option is better? Baylor? Maybe? That would be my choice, but this would be a very disappointing draw.
MSU draws a No. 7 (or higher)
If MSU beats Maryland and then beats Michigan again, I think that it is possible that MSU will pull themselves out of the No. 8/No. 9 line and up to the No. 7 line. I am not sure of this, but this is what I am thinking. This would once again put MSU in a position to possible make a little noise in March.
Possible first round opponents (No. 10 seeds): UCLA, Maryland, and Rutgers
In this analysis, MSU is a No. 10 seed as well so there are only three other teams in this group, and two of them MSU cannot play in the first round. Obviously at least one other team, such as Virginia Tech would have to fall to the No. 10 line if MSU were to move up. Neither UCLA or the Hokies scare me, particularly.
As a No. 7 seed, Michigan State would line up to face a No. 2 seed in the second round so the analysis above holds true. If MSU rises to a No. 6 seed, the Spartans would face a No. 11 seed, which in general are the teams listed above as on the bubble and potential First Four teams. If MSU were to somehow make it up to the No. 5 or even No. 4 line, the Spartans would likely face another No. 4 or No. 5 team in the second record.
To put a bow on this, there are a lot of good scenarios for Michigan State and a few bad ones. The best draw, to me, would be if MSU were to either wind up as a No. 6 or a No. 11 seed. Either way, the Spartans would not have to play any team seeded better than No. 3 in the first weekend and the projected No. 3 seeds right now all look relatively weak.
Furthermore, I seriously doubt that the committee would place Michigan State in the same Region as Iowa or Ohio State if they could avoid it (and they should be able to... it’s not that hard). This means that even the No. 2 seed in that Region would not be that scary. I am not saying that MSU would even win the first game in this scenario, but I do think that it provides the best odds for success.
If MSU were to draw a No. 7 or No. 10 seed, it would almost be as good. This would ensure that the Spartans are not paired with Iowa or Ohio State and the other No. 2 seeds might be ripe for an upset.
The worst case scenario is for MSU to wind up as a No. 8 or No. 9 seed. Unfortunately, I think that the odds are 50-50 that this is where the Spartans wind up, especially with a win over Maryland.
In just a few days, all of the speculation will be over and we will have our first official NCAA Tournament bracket in 24 months. Just hearing the intro music to the Selection Show is bound to bring a tear to my eye. Once the bracket is set, it will be time to break it all down.
Stay tuned and Go Green.