Saturday, April 7, 2018

How Football Ruins Everything and How Basketball Can Solve their Class Crisis



My buddy and I were sitting around watching Ben Richardson drain threes in the faces of many Wildcat defenders when he inquired about my bracket projections and whether or not Loyola would have gotten in if they had not automatically qualified.  This is obviously a very baited question, as we can only know what the numbers tell us, not what other people perceive the numbers to mean, regardless of agenda.  Much like 2017, I believed that the MVC regular season champion profiled favorably to the other 11 seeds (factoring in a neutral loss to NIA), and the efficiency numbers had them well ahead of Middle Tennessee, Louisville, and NC State.  That said, I doubted that the committee, when faced with that decision again, would send Loyola to Dayton.  It is a shame.  In the early days of the “play-in”, it appeared that mid-majors were seen as the preferred matchup for these games, with the likes of BYU/Iona, LaSalle/Boise.  Here is a quick table.  For this drill, the Big East and AAC are not Power 5 and are binned apart in the “mid” category.
Clemson
70–52
UAB
VCU
59–46
USC
South Florida
65–54
California
Brigham Young
78–72
Iona
La Salle
80–71
Boise State
Saint Mary's
67–54
Middle Tennessee
North Carolina St.
74–59
Xavier
Tennessee
78–65 (OT)
Iowa
Dayton
56–55
Boise State
Mississippi
94–90
Brigham Young
Michigan
67–62
Tulsa
Wichita State
70–50
Vanderbilt
Kansas State
95–88
Wake Forest
Southern California
75–71
Providence
Saint Bonaventure
65–58
UCLA
Syracuse
60–56
Arizona State

“Mid-Majors” made up all the participants in 2013.  With a quick run down, beginning in 2011: 2/4, 3/4, 4/4, 1/4, 3/4, 2/4, 1/4, 1/4. 17/32 isn’t terrible, right?  In head-to-head games, the BCS leads the Mids 5-4, with two of those losses from the Big East.  (Are the Big East and American Athletic Mid-Majors?  In terms of athletic revenue, absolutely.  By performance and visibility, of course not?  But the NCAA sees dollars first.

So it is easy to see where the Twitter noise is coming from that the mid-majors are getting the shaft.  2 of the last 8, and those teams were Saint Bonaventure and Providence, who represent conference many would have considered majors.  Not to beat dead Sun Devil, but this just looks like the NCAA and networks thinking “name” schools from big conferences will draw more revenue than the next best mid-majors.  OK, then… open up the 10/11 lines to some deserving mid-majors, instead of packing in the 10th best teams from the Power 5.

Back to the original question.  So we determined that Loyola was likely screwed.  And while they would certainly be more competitive than DePaul, I doubt that they are crossing conference lines any time soon.  Then we asked the question: how would NCAA Championship look if the football selection was used?

We pulled out the numbers and, no, of course Loyola missed.  Here is what we came up with:

1
Virginia

Kansas
2





4
North Carolina

Purdue
3

In addition, the Top 12 were ranked: 

5
Tennessee
6
Duke
7
Arizona
8
Michigan State
9
Villanova
10
Cincinnati
11
Michigan
12
Gonzaga

Tough break for the non-power 5.  So the methodology requires you scrap the conference tourney as it occurred, because football only selects its “division” champs for post season play.  Duke would have lost to UVA (by the numbers) and then fallen behind UNC in the metrics pile, but edging out unconvincing conference champs in Tennessee and Arizona.  Villanova? Xavier? Cincinnati? Gonzaga? No shot.

But there is a solution out there.  One that brings the competitive mid majors back into the fold, and gives a 29-5 mid major the same chance a 19-15 major has of being selected.

Expand the field.

I was a HUGE naysayer disrupting the symmetry of the 64 team tourney.  I saw little benefit, and recognized it as a slap in the face to the less endowed schools and conferences.  When it moved to 68, and they balanced the earned autobids with at larges playing in the First Four, I was a little less frustrated.  The format has grown on the public.  Here is the solution.

There are 32 automatic bids, seeded 1-8 in each region.  Here is how that would have looked this year.

Auto




1
Virginia
Villanova
Kansas
Cincinnati
2
Michigan
Gonzaga
Arizona
Kentucky
3
Loyola
New Mexico State
Buffalo
San Diego State
4
Davidson
South Dakota State
Murray State
UNC Greensboro
5
Montana
Charleston
Bucknell
Marshall
6
Stephen F. Austin
Wright State
Iona
Penn
7
Lipscomb
Georgia State
UMBC
Fullerton
8
Radford
Texas Southern
LIU Brooklyn
NC Central

There are 64 at-large bids, and how they are seeded is open to interpretation.  They could go in as 9-24, and simply fill in behind the auto-seeded teams.  This could put a 1 seed at a major disadvantage competitively to the top at large team.

The other option would be to pod-off the auto-seeds, and have the at large field play into that.  This would maintain the competitive balance and allow the committee to reward the best at-large teams without impact the best auto teams.  Here is what the At Large seed list would have looked like, shown below.  Fair, balanced, certainly more inclusive.  Sure, Saint John’s got bumped because they were sub-.500. Rider would be squawking about getting left out, much the way MTSU and SMC had beef this year.  Now the tricky part is figuring out how to mesh these two tables.

At Large




1
North Carolina
Duke
Xavier
Purdue
2
Tennessee
Michigan State
Wichita State
Clemson
3
Houston
West Virginia
Texas Tech
Auburn
4
Nevada
Ohio State
Texas Christian
Arkansas
5
Texas A&M
Miami FL
Seton Hall
Florida
6
Southern California
Butler
Rhode Island
Missouri
7
Texas
Providence
Oklahoma
Florida State
8
Kansas State
Alabama
Saint Bonaventure
Virginia Tech
9
Creighton
Saint Mary’s
UCLA
Middle Tennessee
10
NC State
Louisville
Baylor
Western Kentucky
11
Syracuse
Notre Dame
Oklahoma State
Marquette
12
Penn State
Arizona State
Boise State
Nebraska
13
Utah
Georgia
Oregon
BYU
14
Mississippi State
Vermont
Temple
Old Dominion
15
Louisiana
LSU
Northeastern
Maryland
16
Washington
South Dakota
Belmont
Stanford

On the S-Curve for the T-Factored Ratings, there is a pretty steep drop between auto bid winners Marshall and Stephen F. Austin, which means the at-large field should mostly work in behind there.  But for simplicity, let’s just run the bracket seeding the 64 at-large teams in directly behind the 32 seeded teams.



The first thing noted is that- yikes- the 1 seeds are exposed.  Virginia would end up with a very tough road to the Sweet 16, having to beat the winner of the Alabama/UCLA game, followed by likely Xavier.  And through a quick survey (and I’m sure KenPom could back these up) that the 9 seeds would have a much better chance of reaching the Sweet 16 than the 1’s, just because their win probability in their second game would be closer to 100%, and given that they are only marginally worse (if not better) than the 1 seeds in their brackets. This puts many top 10 matchup in the Round of 32 (if they even survive) and dilutes the final product.  Just by my calculations, only Villanova would maintain a slight edge over UNC, as Kansas would fall behind Duke and Cincinnati behind Purdue.  

In an effort to drive these games into the Sweet 16 and Elite 8, here is a suggestion: reverse seed the at-large pool again.  This would line up the Top 2 at-large bids against the 5 and 6 auto seeds.  The seed weight of each pod still totals 75, indicating that bracket is balanced, although still a little unfair.  The 1 seed will still play the same level of opponent as the 8 seed, indicating that the 8 seed is at a huge advantage to the 4/5 seeds, who will likely play a Top 10 team in the second round.  This also makes it very difficult to meet the NCAA matchup criteria strictly using seed lines, as the 15-18 lines (at-large 8-11) will always be in the same pods; Louisville had to drop pretty far.



Finally, we arrive at what I feel is the best compromise: the Staggered Seeds.  The 1 seeds represent teams that: were excellent in excellent conferences, then proceeded to demonstrate that by earning the automatic bid by besting the field again.  Now, say Pittsburgh win the ACC tourney this year… they don’t automatically move to a preferred line because of their ACC status.  They would be seeded in where they fall amongst the auto bid pool, which would be ~7 seed.  The 1 seeds are protected and play the weakest teams in the field, who happen to autobids.  The path to the Sweet 16 is much easier for the 1-2 seeds.  The 3/6 seeds also have to like their lay out.  



Nobody is saying that 96 teams is the solution.  Or 68 teams for that matter.  The NCAA is going to make decisions that are in their interests, not the interests of their 351 D I member schools.  Disruption means risk to their cash cow.  This proposal puts all 32 conferences on more equal footing for opportunity.  In the current set up, there is virtually no chance for teams outside the top 10 conferences to earn an at large bid, even if they run the table.  To them, the first 27 games are meaningless; the only ones that count are the last three.  In the power 5, the conference tournaments are essentially meaningless in terms of NCAA tournament selection for the majority of teams; their fates are sealed aside from a few bubble teams or stinkers that make a run (2008 Georgia?).