Eliminating Contenders To Find A Winner For The 2019 NCAA March Madness Tournament

By | March 19, 2019

The NCAA Men’s Basketball Tournament is upon us once again and as always I will be breaking down 5 ways you can eliminate some top seeds from consideration. In picking potential final 4 teams and champions I find it much easier to eliminate teams from consideration to narrow down the selections. The criteria we will use to do this are as follows

1. Preseason AP Rankings vs Current Rankings
2. 3 Point Shooting
3. Points Per Game
4. Coaches Success
5. Scoring Margin

1. Pre Season AP Rankings can predict tournament success

The following comes from a Nate Silver article in the NY Times before the 2011 tournament.

Since the tournament field expanded to 64 teams in 1985, there have been 41 cases in which a school that was unranked to start the season entered the tournament ranked in the A.P. top 10 (excluding a couple of cases where the team was ineligible for tournament play). These schools, as you might expect, were seeded highly: eight were No. 1 seeds and another 17 were seeded No. 2.
Absolutely none of these teams have made the Final Four, however: they are 0-for-41. Instead, they have been the victims of some notorious upsets, like the No. 2-seeded South Carolina Gamecocks losing — by 13 points! — to the No. 15 seed Coppin State in the opening round of the 1997 tournament, and the No. 3 seed Wisconsin Badgers being one of Davidson’s victims in 2008. Three of the No. 1 seeds that fit this description — Michigan in 1985, St. John’s in 1986, and Cincinnati in 2002 — lost in the Round of 32.

Since this article, I have kept track of all qualifying teams.

2011 Notre Dame (2) lost in 2nd round to Florida
2012 Florida State (3) lost in 2nd round to Cincinnati
2013 Georgetown (2) lost in 1st round to Florida GC
2013 Miami (2) lost in 3rd round to Marquette
2014 Villanova (1) lost in 2nd round to Uconn
2014 Iowa St (3) lost in 3rd round to Uconn
2015 Maryland(4) lost in 3rd round to Kentucky
2015 N. Iowa (5) lost in 3rd round to Louisville
2016. Xavier (2) lost in 2nd round to Wisconsin
2016. Oregon (1) lost in Elite 8 to Oklahoma
2016. West Virginia (3) lost in 1st round to Stephen F. Austin
2016. Miami (3) lost in 3rd round to Villanova
2017. Baylor (3) lost in the 3rd round to South Carolina
2018. Virginia (1) lost to 16th seed UMBC in its first game
2018. Michigan (3) lost in the championship game

The record is now 1-55 for these teams to reach the Final 4. Last season Michigan became the first team to buck this trend. They did so by miraculously getting out of the rd of 32 game against Houston. Houston had a chance to go up 4 with a few seconds left but missed 2 FT’s and Michigan answered with a 30 ft 3 point shot at the buzzer to win by 1. A painful game for me as I had Houston in a rather large Calcutta style pool.

Here are the preseason rankings for 2018-19

Current Teams in the top 10 not ranked in the preseason

3W Texas Tech

2. 3 Point Shooting

The 3 pt shot has become extremely important in both the NBA and college basketball.

Since 2007 no team who shoots under 30% from behind the arc has won a game in the tournament.

No teams qualify this season

Since 2007 only 5 teams shooting under 33% from 3 have made the elite 8. If you are not going to shoot the 3 well then you need to defend it well. In 2012 both Louisville and Ohio State made it to the final 4 with sub 33% 3 point shooting but both defended the 3 as good or better than they shot it. Louisville gave up 30.3% from 3 and Ohio State 32.5%. 2015 Kentucky was sub 33% but they defended it well at 32.5%.  2016 North Carolina broke this trend a bit with a poor 3pt shooting team that gave up 36% from 3.

Teams to be leery of
1E Duke – 30.2%
8E VCU – 30.7%
13E St. Louis – 30.8%
16E UNC Central – 31.1%
16W Prarie View – 31.5%
10E Minnesota – 32.1%
3E LSU – 32.3%

What is interesting here is how many teams in the East bracket are on this list including the number 1 overall seed Duke. When we look further into the numbers Duke is the 9th best team in the country at defending the 3 at 29.3%. Potential 2nd rd matchup VCU also defends the 3 really well. They give up 27.6% from behind the arc which is 3rd best in the country.

3. Points Per Game

Playing defense is great but you will need to keep pace with great offenses come tournament time.

80 Teams have scored 73 PPG or less coming into the tournament and had a 1-4 seed. 11 made the final 4 with no champions.

Three seasons ago there was a change in the rules. The shot clock went from 35 to 30 seconds. This has caused more possessions and as such higher scoring. The average ppg per team since 2010 are as follows

2010-11 : 68.28
2011-12 : 67.25
2012-13 : 66.72
2013-14 : 70.10
2014-15 : 66.76
2015-16 : 72.29
2016-17 : 72.57
2017-18 : 72.98
2018-19 : 71.77

Scoring has increased by about 6.6%. If we add 6.6% of 73 to 73 we get an average ppg of 77.8. This is now the number needed to be exceeded in order to be a contender. To be conservative we will round down to 77.5 PPG. In 2015-16 Villanova was at 77.2 ppg entering the tournament and won it all. We want to be careful if teams are really close as this stat has obviously changed over time.

Here are the 1-4 Seeds and their PPG

The teams coming up short are
1S Virginia
2MW Kentucky
2W Michigan
3W Texas Tech
3S Purdue
3MW Houston
4E Virginia Tech
4W Florida St
4MW Kansas
4S Kansas St

4. Coaches Success
Coaches are very important to the success of a team come tournament time. The great coaches tend to outperform their expectations.

This is taken from Pete at Bracket Science

Using tourney appearances and Elite Eight trips, I’ve come up with the following taxonomy of coaching types in the tourney:

Rookies – making their first trip to the tourney
Novices – 2-5 tourney trips with no Elite Eight runs
Prodigies – 2-5 tourney trips with at least one Elite Eight run
Snake-bit – more than five trips with no Elite Eight runs
Flashes – more than five trips with one Elite Eight run
Destined – 6-10 trips with more than one Elite Eight run
Veterans – more than 10 trips with 2-4 Elite Eight runs
Legends – more than 10 trips with more than four Elite Eight runs

I’m focusing this analysis on one through six seeds, the seeds most likely to advance in the dance. A standard PASE analysis on the eight classes of coaches turned up these results:

2014_Coach_Taxonomy

As the chart shows the “snakebit” coaches are ones you will want to stay away from picking to make a deep run.  Last season Cincinnati and Purdue fit this category and fell short of their seed expectations as 2 seeds.

This year’s teams with Snake Bit coaches with a 1-6 seed are:

6E Maryland
3S Purdue

 

5. Scoring Margin

Using the PASE (Performance above seed expectations) stat from the now-defunct Bracket Science the number one determinant of being able to outperform your seed expectation is scoring margin.

129 teams have been a 1-4 seed and come into the tournament with a 15+ scoring margin. 51 have made the final 4 (39.5%) and 18 have won it all (14%) including Villanova last season (Virginia was at 14.2 and lost in their first game)

 

Of the 157 teams to get a 1-4 seed and have a scoring margin under 10 just 2 have won a title and 11 have gone to the final 4.

Last season
1W Xavier Out in Rd of 32
2W UNC Out in Rd of 32
3S Tennessee Out in Rd of 32
4S Arizona Out in Rd of 64

Teams under the 10ppg margin this season
3S Purdue
3E LSU
4W Florida St
4MW Kansas
4S Kansas St

Summary

31 Champions since 1985 have been a 1-4 seed with just 3 seeded higher. So for a potential champion, we should eliminate all teams seeded above 4.

From the above, we can eliminate the following top 4 seeds from championship contention;

Team (corresponding number from above that eliminated them from consideration)

1 Seeds
1E Duke (2)
1S Virginia (3)

2 Seeds
2MW Kentucky (3)
2W Michigan (3)

3 Seeds
3W Texas Tech (1,3)
3S Purdue (3,4,5)
3MW Houston (3)
3E LSU (2,5)

4 Seeds
4E Virginia Tech (3)
4W Florida St (3,5)
4MW Kansas (3,5)
4S Kansas St (3,5)

If we go by these numbers we have eliminated 12 of the top 16 teams as potential champions. I did mention in the 3 point section that while Duke does qualify as a potential upset victim because of poor 3 pt shooting they do hold opponents to a 3 pt% lower than what they shoot. We have seen in the past that this formula can work and since Duke is the favorite to win it all they could very well win it all. However, if you are participating in a bracket pool with a large number of entrants you may want to look to a team other than Duke.

The teams we have left are
1W Gonzaga
1MW UNC
2E Michigan St
2S Tennessee

Let’s have a look at where the country is picking on ESPN

Almost 40% of brackets have Duke as a champion. If you have 100 entrants in your pool and you pick Duke to win it all you are going to be -EV on that bracket. The reason is Duke has about a 19% chance to win the championship according to 538.  The Vegas odds show about a 29% chance of Duke winning it all. That number is juiced a bit as well. Regardless of what you look at, there is no way Duke wins this tournament 4 out of 10 times as the public is picking. Because of this, you want to be short Duke and find value elsewhere. Of course, Duke may win and you will lose all of your brackets but even if you pick Duke and they win the odds are good that you will not win your bracket anyway. If you insist on picking Duke (perhaps you are an alum) then you need to differentiate yourself elsewhere in the pool. Pick some early upsets and perhaps look for a long shot or 2 to make the final 4 and championship game.

With 1 team from each bracket remaining, I will probably be putting in a bracket with a final 4 that looks exactly like the above (I play a few different brackets to be diverse and increase my chances of winning). It is a bit chalky but without Duke, it will be different than most brackets in most pools.

In this bracket, I will eliminate Gonzaga based on the 83rd best Strength of schedule. They typically have a weak SOS coming out of a mid-major conference and this season is no different.

The way my bracket pool works is you get the seed multiplied by the rounds value. So a 1 seed gets half the points a 2 seed gets for each win. With that said I will go with Michigan St. and Tennessee in the final game with Michigan St winning it all. I think this gives me enough differential from other Michigan St winners. In a 100 person pool, I expect there to be 5-8 Michigan St champions picked. ESPN has it at 4.7% but knowing my pool there will be a few more than that. The percentage is still low enough that if they do win it all my chances of cashing are high (top 5 cash). I will be adding a few more brackets with some other champions. I am going to do more analysis in a future article on possible Cinderella picks. That will help you if you take mostly favorites in the final 4.

As always good luck and let me know who you are picking!

Stats

Leave a Reply

Your email address will not be published. Required fields are marked *