March Madness: Using Advanced Rankings To Select The Final 4

By PAUL NEWBERRY

AP National Writer / March 17, 2013

Who else has spent countless hours breaking down the field while attempting to fill out their bracket? If your anything like me then you definitely have, and you probably won’t submit your final bracket until Thursday morning.

For most bracket pools, your Final 4 picks and National Championship picks receive the most points, and they will separate the winners from the losers. Today I’m going to help make your selections a little bit easier by eliminating some teams using the Kenpom Ratings.

I’ve reviewed every Final 4 and National Championship team from the past 10 years and have found trends that will help you separate the true contenders from the pretenders. Obviously there are never guarantees when it comes to the NCAA Tournament and every year has it’s own surprises so don’t put $100 into a bracket pool and then blame me because you didn’t win.

Final 4 Capable Teams

Jed Conklin/AP

The first major trend that I noticed with Final 4 teams, is that there has only been one Final 4 team since 2003 that has had an AdjO ranking higher then 50. That was Louisville last year. However, Louisville had the No. 1 overall ranking for AdjD. So I eliminated every team from the field that had a AdjO ranking greater then 50, unless they had an elite defense ( Top 5 AdjD ranking).

There was also a very similar trend using AdjD. Only two teams have had an AdjD Ranking of 50 or higher and made the Final 4. Marquette was one of them in 2003, but their AdjO ranking was No. 1 overall. Virginia Commonwealth in 2011 was the 2nd team. VCU had a defensive ranking of 86th and an AdjO ranking of 38.

Since VCU was the only team out of 40 that either wasn’t ranked in the Top 50 of AdjD or didn’t have a AdjO ranking in the Top 5, I went ahead and threw them out.

I also looked at KenPom’s Strength of Schedule . There has only been one team since 2003 with a SOS worse than 70 that has made the Final 4. That was George Mason in 2006 when they had a SOS ranking of 93. Again since they were an extreme outlier, I deleted them from the trends list and decided to eliminate all teams this year with a SOS outside of the Top 70.

They final thing I looked at was seeds. Out of the last 44 Final 4 teams, only three schools (2006 #11 George Mason, 2011 #11 VCU, 2011 #8 Butler) have reached the Final 4 with a 6th seed or worse. That means 93% of the Final 4 teams over the last 11 years have been a 1-5 seed. Sure years like 2011 happen but is a 7% chance worth the risk?

The last thing I did was eliminate all teams with a Pyth Rating of .9000 or lower. Once again, only George Mason, VCU, and Butler have made the Final 4 with a Pyth Rating of .9000 or worse.

You are then left with 13 teams that meet all requirements above. Now I know 13 teams is still a lot of teams to pick from when deciding your Final 4, but we still eliminated 80% of the field.

Notable teams missing:

– Gonzaga (1 seed): Eliminated due to a 96 SOS. A lot of the experts have said Gonzaga doesn’t deserve a 1 seed because of their schedule, well it hurts them here as well. Gonzaga is a great team, but if they make the Final 4 this year, they will have the worst ranked SOS of any team to make the Final 4 since 2003.

– Marquette (3 seed): Eliminated due to a Pyth Rating below .9000. The Golden Eagles had a Pyth Rating of .8805 which falls below the threshold I’m looking for in a Final 4 team. Only 3 out of the past 40 Final 4 teams (7.5%), have made the Final 4 with a Pyth Rating worse than .9000.

 Which teams can make the title game?

Can the Gators be consistent enough to make a run in the tournament?

The next step was to find a trend for teams who made the national championship game. First, you should only project teams with a 3 seed or better to make the title game. Butler in 2010 and 2011, has been the only team in the past 10 years to make the title game that wasn’t a 3 seed or better.

I’m also going to delete the 2011 Butler team completely because they were major outlier as far as the trends are concerned. Once you delete them, all of the other 19 teams to make the national title game had a AdjD ranking in the Top 20.  This would eliminate Duke, Michigan, Syracuse, and Miami.

Another trend was that 100% of the national title teams, including the 2011 Butler team, had a AdjO ranking of 50 or better. This would eliminate Wisconsin and Georgetown.

The next thing I noticed, was every national title team, besides the 2011 Butler team, had a AdjO ranking in the Top 25 or had a Top 5 defense.  This eliminates New Mexico.

SOS didn’t eliminate anyone at this level, but the 2008 Memphis team with a SOS ranking of 64 was the worst out of any team to make the title game in the past 10 years. So who’s left?

Who will be the 2013 National Champion?

(Photo: Dennis Wierzbicki, USA TODAY Sports)

The biggest trend from National Championship teams, is every single one of them has been ranked in the Top 20 of both AdjO and AdjD. This would eliminate Kansas and Michigan State, who falls outside of the Top 20 in offense.

Another thing to look for is a SOS in the Top 50 and being a 3 seed or greater. 100% of the teams to win a title over the last 10 years has fallen into both of these categories. So if you believe in all of these trends and rankings, the 2013 National Champion will be one of the following teams. Do you agree or disagree? And good luck to everyone with their brackets!