For the first time since the NCAA Tournament has been seeded the Selection Committee has released the overall rankings of all seeds otherwise known as the S-curve. While the NCAA Selection Committee ranks the teams this way certain rules dictate match-ups be changed. After the jump is the S-curve in full.
[table id=207 /]
Looking at the way the brackets were setup and the S-curve the first question that arises is how the #2 seeds ended up where they did. According to the S-curve, Kansas is #5, Duke is #6, Ohio State is #7 and Missouri is #8. This is speculation on my part but here is what I think happened.
The first obvious change to the top eight which had to be made was Duke as the #6 which would have put the Blue Devils as the #2 seed in Midwest where UNC is the #1 seed. That match-up is not going to happen so Duke needed to be moved. Off hand it would make sense to flip Duke and Kansas but that would put Duke as the #2 seed in the West where Michigan State is #1. Duke and MSU already played once this season so the committee likely would want to avoid that as a regional final. Flipping Duke and Ohio State was possible but it looks like the disposition of Florida State as a #3 seed in the East may have prevented that. FSU should have been the #3 in the Midwest but likely changed because that was UNC's region. If FSU is the #3 in the East that means Ohio State has to stay the #2 there. Duke and FSU cannot be #2 and #3 together since the NCAA prohibits match-ups between conference schools prior to the regional final unless it cannot be avoided(such as Marquette vs Syracuse last year.)
The solution? Duke was sent to Atlanta which is closer to home but also with the #1 overall seed. Missouri couldn't be moved to #6 because that would be giving a lower seed a favorable geographic placement. Since Kansas was the #5 overall it makes more sense to place them in St. Louis over the #8 overall Missouri who was sent West with the #4 overall Michigan St. Ohio St. remained it in its "natural" position was the #7 overall making them the #2 in the East to overall #2 Syracuse.
The #3 seeds were also interesting. Baylor was assigned to the South as it should have been. Marquette should have been gone to East but instead went to the West. FSU was moved from the Midwest to the East and Georgetown from the West to the Midwest. Likewise the #4 seeds were also shuffled. According to the S-curve, Wisconsin should have been the #4 seed in the Midwest but UNC played the Badgers already. Louisville was the #4 in Kentucky's region but those two are rivals and had played this season. Michigan was lined up to be the #4 in Michigan State's regional and Indiana the #4 in the East. In other words everyone had to be moved and the end result was Kentucky getting Indiana as the #4 in the South despite the fact the two played this season. This was probably an attempt to still give UK the lowest seed possible. Wisconsin had to avoid both UNC and Michigan St so they were placed in the East. After that Michigan went to the Midwest and Louisville to the West.
The real takeaway from this is how much the committee shuffles the S-curve for whatever reason whether it be to avoid certain match-ups, the geography and conference affiliations. Here is how the top four seeds in each region look according to the S-curve as compared to how they ended up.
[table id=208 /]
Had the committee gone with the S-curve the Midwest would have been a repeat of the ACC Tournament with UNC, Duke and FSU as the top three seeds. There would have also been rematches from the regular season and conference foes playing each other all over the bracket. As it stands now only one rematch is lined up which was unavoidable lest the #1 overall seed be given a tougher match-up.
Needless to say, for all the grief the committee gets, even sorting out the top 16 seeds in the S-curve is complicated and I am not even well versed on most of the rules in play. Multiple this by four and consider all the other factors of including or excluding teams and it is certain a tough task.