First things first, let's set the table with the latest updated RotoWire bracket for the 2018 NCAA Tournament.
If youíre a big college basketball fan, this is your favorite time of the year. If youíre not a college basketball fan, this is still one of the better times of the year. The NCAA tournament is less than a week away and brackets will soon be analyzed way too much. Before we get to the actual bracket, itís still fun trying to predict what the NCAA committee will do.
In my five years of doing Bracketology, this season may be the toughest of figuring out where the committee is coming from. In addition to the questions surrounding advanced analytics and RPI usage, the bubble is loaded with a surprising amount of teams that have resumes filled with quality wins. That includes the recently maligned Oklahoma and Arizona State.
Both the Sooners and Sun Devils are two teams the general public is bashing and stunned that either team is in contention for March Madness. According to Bruce Rasmussen, the chair of the NCAA tournament selection committee, games at the beginning of the season are just as important at games at the end. So while both of those teams have struggled in the last couple months, they also have some huge wins from non-conference play.
Oklahoma won at Wichita State, TCU and in the Staples Center against USC. Its worst loss was at Iowa State (RPI 137) or on a neutral floor against Oklahoma State (RPI 85, in Big 12 tourney). Arizona State beat Xavier and Kansas State on a neutral floor, as well as Kansas and Utah on the road. Throw in home wins over USC and UCLA (both projected to make the tourney) and not many can compete in strength of wins. Its worst losses were at home to Stanford (RPI 83) and at Oregon State (RPI 129).
To compare, there are numerous teams with lesser wins and worse losses that are on the bubble. I havenít been a St. Maryís backer all season because it purposely didnít schedule tougher games in non-conference. The Gaels best wins were at Gonzaga and BYU, but they also lost at San Francisco (RPI 138) and in their two toughest non-conference games on a neutral court to Washington State (RPI 175) and Georgia (RPI 81). Fifteen of their wins came in what the committee considers Quadrant 4. Gonzaga is the only other team close to that number outside of the lower mid-majors. Even St. Bonaventure has just 10 Quad 4 wins and six more in Quad 3.
Most recently, Middle Tennessee entered bubble conversations after losing to Southern Miss on Thursday. Its best wins are against fellow mid-majors (Murray State, FGCU, Western Kentucky) and lost against its toughest opponents (Auburn, Miami, USC). The Blue Raiders have won tournament games each of the last two years so the public is familiar with them, making them a more popular team. But according to the resume, they donít deserve to make it, no matter what the RPI is (32). Those pining for Middle Tennessee would also have to go for Boise State, which has a similar resume, although worse RPI (52).
There are cases like this all across the bubble. The committee has to decide what matters most. Whether itís between rankings like RPI and KenPom or quality of wins and losses, there will be a line drawn somewhere. Throw in injuries and it gets even more confusing. Notre Dame is probably a tournament team at full strength, but it still lost at home to Ball State and on a neutral court to Indiana with everyone healthy. Almost half of the teams in the tournament and on the bubble have had a major player miss a game or two due to injury.
Itís the same case when seeding teams already in the tournament. Teams like Michigan State and Cincinnati donít have a bad loss, but how much does a lack of good wins mean? North Carolina lost at home to Wofford, but it also has the most Quad 1 wins in the country. Then thereís Missouri, which finally got Michael Porter Jr. back in its lineup. After watching the Tigers on Thursday, Iím not changing their seeding at all.
After Thursdayís games, Michigan State lost a Quad 1 win (Maryland dropped), while Middle Tennessee gained one (Old Dominion moved up). With that, MTSU has more Quad 1 wins than MSU. Again, MTSUís best win is at Western Kentucky, according to RPI. If you look at some of these numbers with a microscope, the flaws start to shine.
Iíve said before that I prefer a combination of KenPom and Strength of Record, but comparing teams by simply looking at numbers on a page usually doesnít work. Everyone loves the blind resume test, but those are close to useless unless you dive into the schedule of each team. Simply listing Quadrant records doesnít show the difference between beating Virginia and Villanova or winning on the road at Rider, UNC Greensboro or Utah Valley. There are flaws in every metric or stat, RPI and KenPom included, and most of the time examining the full resume is the best way to go.
The bubble changes by the day during conference tournaments, so keep an eye on my twitter (@RotoZdroik) and RotoWire for updates. This is where I stand as of Friday morning.
IN: St. Bonaventure. Providence. Arizona State. Alabama. Louisville. Baylor.
ALIVE: St. Maryís. Middle Tennessee. Marquette. Notre Dame. Oklahoma State.
OUT: Syracuse. Nebraska. Penn State. Boise State. Utah.
If you want my reasoning on where I have certain teams, leave a comment or send me a message. Iíve gone in depth over the last month in regards to every bubble team and at this point with two full days of games left, things could easily change in the next 24 hours.
*Stats from KenPom, RPI and SOR are as of March 9.
**For the stats used, KenPom and RPI are most relevant in the conversation and I replaced BPI with Strength of Record (SOR) because BPI is extremely similar to KenPom these days. The combination of KenPom and SOR is what I believe to be the best and easiest metrics to use when looking at teams. It doesnít overrate mid-majors for beating up on weaker competition, which is KenPomís biggest problem.
To view a compilation of all bracket predictions in the world, check out the Bracket Matrix. This bracket started in 2014 and has been among the most reliable the last three years.