Not so Crazy (Elite) Eight

In a nod to the NCAA Tournament Selection Committee and to the dominance of four 1-seeds during the season, it would seem that this year there will not be too many surprises at the top of the bracket and the numbers bear this out.

Using the criterion established in the first part of this post in selecting which teams are most likely to reach the Elite Eight (i.e. record vs. top 50 RPI, undefeated vs. RPI 100+, > 6-4 record in the 10 games preceding the NCAA Tournament, top 15 in either Ken Pomeroy's adjusted offensive efficiency or adjusted defensive efficiency), 7 teams (North Carolina, Tennessee, Kansas, Wisconsin, Georgetown, Memphis, and Xavier) perfectly match the qualities of the group that made up 2007's Elite Eight. This is somewhat problematic because 7 doesn't make 8 unless you're this guy and because Georgetown and Wisconsin are slated to play each other in the Sweet Sixteen Round (thus only 1 of them will reach the Elite Eight), leaving 2 spots to be filled by teams with some blemishes according to this rubric.

North Carolina, Tennessee, Kansas, and Memphis, not surprisingly, are the best bets to win each of their first 3 games of the NCAA Tournament. They are 1, 2, 5, 3 in RPI, respectively, and have dominated RPI Top 50 foes to the tune of 34-7 and, with the exception of Tennessee, are in the Top 5 in efficiency ratings. Accordingly, of this group, I believe that Tennessee will have the most difficulty in advancing to the end of the Regionals because they face a stern tests against Butler and Louisville.

Moving on to the 5th Elite Eight team, Xavier may be somewhat of a surprise as they are projected to face Duke in the Sweet Sixteen. Xavier advances here because Duke has lost 4 games in the past month (to Wake Forest, Miami, North Carolina, and Clemson), which is disturbing because before the Wake Forest game, Duke had only lost once the entire year and because Xavier is clearly superior to 3 of the 4 teams on that list. Additionally, the Selection Committee has finally taken Duke's dick out of its mouth and for the first time in recent memory, has sent the Blue Devils out West, far from the homecourt advantage they have been afforded when playing in Charlotte, Greensboro, or Raleigh.

In the potentially epic Wisconsin-Georgetown throwdown, both teams bring equally impressive performances during the regular season. Wisconsin has the slight edge because of 3 factors: (i) the game will be played in Detroit, easily accessible to the minions of Badger-faithful who have a reputation for traveling well to neutral sites; (ii) Wisconsin has won its last 10 games, including a relatively easy romp through the Big 10 Tournament; and (iii) Defense wins championships (or, in this case, a berth to the Elite Eight). Wisconsin's adjusted defensive efficiency, measuring points allowed for every 100 possessions, block rate, steal rate, turnover percentage, defensive rebounding percentage, etc., is #1 in Division I.

Now to the two remaining spots to be filled by a marginally "flawed" team. Initially, before the brackets were released, I eliminated UCLA and Texas for the sole reason that they were not perfect against teams with an RPI over 100, the logic being that a single loss to a team that they should not only beat, but beat convincingly, is indicative of a dearth in focus or a susceptibility to playing to their opponents' level and would leave them vulnerable to losing once in the first 3 games of the Tournament. A loss to Washington or Missouri (albeit a road loss) is borderline inexcusable in much the same way that it is inexcusable to finish with a sub .500 record against the Top 50 in RPI. But, like UNLV last year, UCLA and Texas will be the exceptions to the rule because they are strong by all the other measures of a team's success.

I know these picks are not necessarily sexy (no seed lower than 3 will make it past the Sweet Sixteen) or may not be a real revelation given the strength of the contenders, but I've done this more as an experiment to determine a system for making better predictions and providing some sensical way to interpret the data we're bombarded with during March Madness. 1 or 2 Elite Eight teams wrong is a failure and will require a broader study of NCAA Tournament history.
4 or more teams wrong and I'll beg you to never bring this up again and I promise to never to write authoritatively about college basketball.

For the record, for the rest of the bracket, I think this year will be devoid of 12-5 upsets, the 6-11 and 7-10 games will be split by the higher and lower seeds, 8s will beat 9s (3-1), and the 1-seeds will all advance to the Final Four for the first time since the 1985 when the bracket expanded to accomodate 64 teams. Oddly enough, because this prediction is at odds with an established historical trend, if any of the 1-seeds are going to be upset it's going to occur in the Elite Eight.

In the most unpredictable playoff of them all, only one thing is for sure. We'll all know how wrong I am on March 28th.