How unpredictable has this year’s NCAA men’s tournament been? The most lopsided win of the Elite Eight belonged to an 11 seed that advanced to its first Final Four appearance in 55 years. At the same time, the three national semifinalists other than Loyola of Chicago — Kansas, Villanova and Michigan — are among the most successful programs in college basketball history. Together, they fill out a Final Four that nicely mixes the traditional with the unexpected.
Loyola is the most compelling story of the tourney, thanks to a run that slowly turned them from a fun Cinderella into something much more special. But they’re not merely a George Mason or Virginia Commonwealth that’ll be easy to knock off, despite the identical seed numbers. According to our Elo ratings — a measure of a team’s strength that is based on game-by-game results — the Ramblers were the 25th-best team in the field going into the tournament, compared with No. 44 for George Mason in 2006 and No. 52 for VCU. (Loyola is a little more like LSU in 1986, which ranked 32nd in the field before the tournament and became the first double-digit seed ever to make a Final Four.) Going back to the start of the 64-team era in 1985, Loyola’s pre-tourney Elo ranks sixth-best among the 14 Final Four teams who were seeded seventh or worse:
The best low-seeded Final Four teams
Best pre-tournament Elo rating for Final Four teams seeded seventh or lower, 1985-2018
So you can build a case that the Ramblers are much better than the typical low-seeded Final Four squad. Still, history has not been kind to Cinderellas in the tournament’s third weekend. In fact, the Final Four is right around when the clock strikes midnight: Teams seeded ninth or worse are 0-6 all-time in the national semifinal.
Michigan is an especially interesting opponent for the Ramblers, given that both teams are riding double-digit winning streaks (Loyola has won 14 straight, and Michigan has won 13) and that the chief knock on Michigan’s otherwise stellar tournament play has been a lack of difficult opponents. If they do beat Loyola, the Wolverines will become the first team in history to make it to the national title game without facing a single team seeded better than sixth. But that probably overstates how easy Michigan’s path was: The average Elo of the Wolverines’ opponents is not notably low by Final Four standards,1 and what’s more, it’s not even the lowest of 2018. (Villanova has faced a much easier path to the Final Four, in terms of its opponents’ average Elo ratings.)
Villanova’s path was easier than Michigan’s
Pregame Elo ratings for Villanova’s and Michigan’s 2018 NCAA tournament opponents, by round
|Villanova Opp.||Elo Rating||Round||Elo Rating||Michigan Opp.|
|Radford||1552||Rd. of 64||1693||Montana|
|Alabama||1795||Rd. of 32||1934||Houston|
|West Virginia||1972||Sweet 16||1892||Texas A&M|
|Texas Tech||1988||Elite Eight||1948||Florida State|
Either way, Michigan’s defense has been dominant — it’s holding opponents to 38 percent shooting from the field in the NCAAs thus far — and in some ways the Wolverines will be facing a smaller and less-heralded version of themselves in the defensive-minded, slow-paced Ramblers. Plus, for all the talk of Loyola’s last championship coming in 1963, it’s also been nearly three decades since Michigan, behind Glen Rice and coach Steve Fisher, last cut down the nets — a somewhat surprising title drought for a marquee program like UM. Our prediction model gives Michigan a 69 percent chance of advancing to its second title game in six years under John Beilein.2
Meanwhile, for all the chaos of this tournament, chalk prevailed in the other regional finals, producing a pair of No. 1 seeds in Kansas and Villanova. It’ll be the tournament’s 15th battle of top-seeded teams in the national semifinals, and the first since Wisconsin-Kentucky in 2015.
Kansas had its hands full with Duke in the Midwest region’s titanic 1-vs-2 clash. According to our excitement index, which measures how thrilling a game was by computing the average change in win probability on each play, the Jayhawks and Blue Devils tied for the second-most heart-pounding game of the entire tournament so far, trailing only Michigan vs. Houston from the Round of 32. It was the kind of game Kansas hasn’t tended to come out on top of in recent years — it’d lost a regional final in each of the past two seasons, and was 1-3 in chances to go to the Final Four since last winning the championship in 2008.
When we looked at March Madness’s best coaches — in terms of exceeding the record we would expect a team to earn in their tournament games based on Elo — the statistical contrast between Kansas’s Bill Self and Duke’s Mike Krzyzewski pretty much summed up the narrative for the two coaches. While each had guided tournament teams with roughly the same seed on average (Self’s teams had an average seed of 2.7, while Coach K’s were at 2.2), Self had played almost exactly to his expected record over the years (+0.1 wins, 188th-best among all coaches), while Krzyzewski had outpaced his expectations by 5.8 wins (seventh-best) since 1985.
But Self and the Jayhawks found redemption behind the outstanding play of Malik Newman, who scored 32 against the Blue Devils, including all 13 Kansas points in overtime. We weren’t fully sold on the Jayhawks before the tournament began — we thought there was a glimmer of hope for Penn to upset them in the first round — but KU has now worked itself up to the fourth-highest power rating of any team in the nation, giving itself a 23 percent title probability according to our model.
Villanova remains our favorite to win the tournament, however, with nearly a 50 percent chance against the field. Unlike Kansas versus Duke, the Wildcats were comfortably in control of their Elite Eight matchup with Texas Tech for practically the entire game, helping it produce the lowest excitement index of any regional final (even lower than Loyola’s rout of K-State). Aside from some midgame jockeying with West Virginia in the Sweet 16, Nova’s victories have seldom been in doubt on the road to the Final Four.
Even when the Wildcats’ offense has sputtered — and against the Red Raiders, they were not impressive by any means, generating their fifth-fewest points per possession in any game this season — their defense has come through. During the tournament, only West Virginia has managed to crack 95 points per 100 possessions against Villanova, and even the Mountaineers were held to their sixth-worst offensive game of the season. Against the Red Raiders, the Wildcats’ two most dynamic players –Mikal Bridges and Jalen Brunson — combined to shoot 7-for-24 from the floor (including 0-for-9 from deep), and yet the defense was strong enough to help Nova not just overcome problems on offense, but cruise to victory.
A Villanova win would put an uncharacteristically routine capper on what has been a wild season in college basketball. According to ESPN’s Stats & Information Group, the Wildcats won their 134th game in the last four seasons when they beat Texas Tech, breaking the Division I record for the most wins by a program in a four-year span. If they win a couple more, they’d become the fourth program to win at least two championships in a three-year span since 1985.
But it feels premature to speculate about that. This year has already offered some of the most unexpected moments in tournament history. So who knows, maybe the tourney still has a few surprises left in its final weekend.