Elite 8 and Final Four Picks

These rounds are all about the best teams taking control.

The Cinderella slippers usually fall off in the Sweet 16.  Double-digit seeds 4-25 in this round since 2003.  Also, picking the top 2 seeds to advance to the Elite 8 is right about a third of the time, 18 times of 60 since 2002.  13 of the 18 teams that entered the tournament with an adjusted efficiency margin over 30 won a Sweet 16 game.  This round is all about the best teams taking control as 4-6 of the kenpom top 10 and 7-8 of the kenpom top 26 usually survive to the Elite 8.

In the top half of the bracket, the top seed has won 31 of 42 times against a 4 or 5 seed.  Of the seven 4/5 seeds that upset a 1 since 2010, six underdogs were in the top 20 in adjusted offensive or defensive efficiency.  Of the 1 seeds that fell, five were ranked worse than 16 in either offensive or defensive efficiency.

In the bottom half, a similar pattern can be found with 2 v 3 or 6 seeds.  The 2 seed has won 21 of 29 times against a 3 or 6, with six of those upsets by a team that specializes in one side of the ball.  In 3 v 7 seeds, the match-up is surprisingly close, each seed winning 4 times.  In all but two of the 3/6/7 upsets, the underdog’s speciality matched opposite the favorite’s speciality (offensively focused underdog v. defensively focused favorite or vice versa).

Going from eight to four teams provides even more significant trends.  Over two-thirds of all Final Four teams have been top 3 seeds, with six 4-seeds, four 5-seeds, and eight from 5, 7, 8, 9, 10, and 11 seeds since 2002.  Over a third were conference tournament champions.  Final Four teams are likely to have been top 3 in their conference during the regular season unless from the old Big East or the new ACC.  Typically, the Final Four teams consist of three from the top 10 preseason poll and one from the top 25.  Also, at least two and usually three of the kenpom top 10 advance to the Final Four.  If a mid-major rides to the Final Four, they should be in the top 10, or at least the top 12 conferences in kenpom.  Finally, FF teams should be in the top 20 in either adjusted offensive or defensive efficiency.

How Sweet It Is

Picking the top four teams to the Sweet 16 has only been right 7 times.

Picking the top four teams to the Sweet 16 has only been right 7 times, just over 10% of tournament regions since 2002.  Only about half of the top 25 kenpom teams and those teams with an adjusted efficiency margin over 25 survive the first weekend.

But if you are looking for major upsets, you should focus on the bottom half of each region.  1 seeds have only lost to eight 8/9 seeds since 2002.  Only two 12 seeds have beaten a 4 seed.  It is much more sensible to focus on a 2 seed and a 3 seed to fall early.

Indeed, in the last seven tournaments, seven 2 seeds and seven 3 seeds have lost in the second round.  2 seeds typically are those teams who did not win their power conference regular season title but won the conference tournament or vice versa.  They are the almost-perfects, the elite but not the best.

In preparing for this post and this entire exercise of making a perfect bracket, nothing was as shocking as this factoid.  In the last 15 years, there have been 14 2-seeds that won a power conference tournament – not a single one lost a second round game.  Now, obviously, 2012 Missouri and 2016 Michigan St., conference tournament winners in their own right, lost in the first round in unexpected fashion.  But if you see a conference tournament winner, pen them into the Sweet 16.

Conversely, 7-10 upsets have come from mid-major conference tournament winners or teams that lost in their conference’s semis or finals.  Only two 7-10 upsets came from a team that lost earlier than the semis since 2002: 2011 Florida St., which lost in the quarters by 1, and 2016 Wisconsin, which lost its first game and won over Xavier on a buzzer beater.

3 seeds versus 6 or 11 seeds are a bit more unpredictable.  Some 3 seeds seem so vulnerable, it feels like they should lose, but they don’t: like 2013 Florida, which went 1-7 in “A” rated games according to kenpom, or 2015 Oklahoma, which had two losses to teams that finished the year ranked 80th or worse in adjusted efficiency margin.  Other 3 seeds seem so strong, they should not lose but did: 2014 Creighton which had no bad or “B” rated losses.  Only about 62% of 3-6 matchups since 2002 resulted in the better team on kenpom winning.  Add in 11 seeds and that percentage barely rises to 67%.  Good, but we need more data.

The only consistent trend that I can see comes from a smaller sample size.  Of the last seven 3 seeds to have fallen, only one lost to defensively focused opponent – an over-ranked 2012 Florida St. team.

Next up is the seemingly trickier 4-5 matchup.  In looking at the data, picking the favorite in kenpom works more often than not and when it doesn’t, it has usually been because the underdog’s defense or offense was better (numbers favor defense slightly more).  Very rarely did I see an underdog 4 or 5 beat a similarly composited opponent.  Beyond that, it becomes tough call.  However, keep in mind that about half of all regions since 2002 have involved a second round matchup that wasn’t a 4-5 seed affair.  Selecting a 12 or 13 seed to win the first round game should make this second round choice easier.

Finally, we have 1’s versus 8’s and 9’s.  As stated near the top of this post, this type of upset is pretty rare but it used to be rarer.  Five 8/9-1 seed upsets have occurred since 2010.  Compare that to 2005-2009 when there were four straight tournaments without a second round upset of a 1 seed.  Three of recent upset winners were mid-majors, three offensive-focused, four out of five were in their conference tournament final, all five ranked in the top 45 teams, all five played an opponent rated 7th overall or better with a top 15 offense in kenpom, and all five had at least eight “A” or “B” ranked wins.  So while it is unlikely that a 1 seed will fall, it has been happening consistently as of late.  Another loss in the second round this year would solidify this type of upset as something that has to be included in a perfect bracket attempt going forward.

Finding a March Madness Miracle

Dunk City highlights the excitement of an increase in 13, 14, and 15 seed upsets in the last couple of years.

In the last decade of tournaments, there has been at least one upset from a 13, 14, or 15 seed.  Of the 24 such upsets since 2002, half have occurred in the last five years.  What is especially remarkable is how consistent several data points are (based on kenpomeroy‘s pre-tournament adjusted efficiency stats) for these March Madness miracles.

All but two 2, 3, and 4 seeds that were upset had less than 14 rank difference between their adjusted offensive and defensive efficiency.  Raising that number to a 25 rank difference still leaves us with 17 of 24 (71%) favorites who have a significant difference in how they play depending on whether they have the ball or not.

Similarly, only one underdog in these matchups had less than an 18 rank differential (2014 Mercer) between offense and defense.  Only one underdog had a negative adjusted efficiency margin and defensive rank above 145 (2012 Norfolk St. for both).  On offense, only three 13 or 14 seed upsets had an adjusted offensive efficiency rank above 135.  Only five 13, 14, or 15 seeds had an adjusted efficiency margin of 5.5 or less and a rank worse than 105 .  Three of those teams were 15 seeds.

Also, only two teams came from conferences with an overall adjusted efficiency margin worse than negative ten, three teams total worse than negative 8.5, and five total worse than negative 6.

Our parameters for a miracle upset:

  • 2, 3, or 4 seed favorite with a 14-25+ rank difference in adjusted offensive efficiency and adjusted defensive efficiency.
  • 13, 14, or 15 seed underdog with a 18+ rank difference in adjusted offensive efficiency and adjusted defensive efficiency.
    • Positive adjusted efficiency margin overall, better than 5.5, and rank, better than 105.
    • Adjusted defensive efficiency should not be beyond rank 145.
    • Conference affiliation matters to a minor extent, but should be no worse overall than negative 6.