So, I get this is a totally pointless exercise.  But I have nothing else to rant and rave about.  So, why not work on my bracketology.  I will probably continue to refine this as the week goes on.   I am thinking that I will come up with a quick estimate over the weekend, and continue to refine my choices based on more educated decisions.

Remember this exercise is meant to predict who the selection committee would pick – not who I think should play in the tournament.  I probably would have a few more rules in place.  For example, I always had the belief that if you don’t finish your season with at least a non-losing record, you should not be in the tournament.  Sure, a team like Purdue or Indiana might have a high enough NET ranking to make the tournament.  But if you went 9-11 in your conference and are the 10th and 11th seed from that conference, you have not earned a bid.  All the major conferences have tough schedules – shouldn’t I lean towards taking Arizona State who went 11-7 in the Pac 12, or South Carolina who went 10-8 in the SEC.  Or why not reward a small conference team like Northern Iowa who won their conference title, beat a team that likely is in the field on the road (Colorado) during non-conference play, but got upset in their tournament.

This year, there were 43 teams that went 50% or better in what I consider the 6 power conferences (ACC, Big East, Big 10, Big 12, Pac 12, and SEC).  If you give 6 of them the automatic bids, that means you would still not have enough at-large spots to give all these teams an invite.  And if I give all these teams the spots, I have no room for a ranked school like BYU who had a great season in a small conference.  So, why would a team with a losing record in conference deserve a spot over any of those teams.

But that is not how the committee works – a team can make up for a bad conference record if they dominated their non-conference schedule.  It actually also helps out when evaluating teams that had injuries like Arkansas, who went 7-11 in conference because of the 5 game losing streak without their 2nd leading scorer.  It also helps out with evaluating conferences with non-balanced schedule (Indiana had to play the three Big 10 co-champs 5 times, while Michigan only had to play them 4 times – would their records have been reversed if they had each other’s schedule).  So, strict rules manage to force the committee not to make exceptions.  I am torn on this, but I can understand it.

So, lets determine the bubble size and who doesn’t have to worry about the bubble.

It is important to understand the Selection Committee process.  The process starts by each of the 10 members picking the 36 teams they think should be in the tournament do matter what – and any team on 8 of the ballots is put into the field, while anyone with 3 ballots get put into the “under consideration” list.  In my mind, this “under consideration” list is the bubble.

Out of curiosity, I wanted to see if I could use the computers and polls to create my own committee.  I looked at this to see how it would do on last year’s bracket, and found some interesting results.  Using this to predict the bubble eventually fails – the committee eventually picks differently from the computers (otherwise, teams like 63rd ranked Arizona State or 73rd ranked St. John’s don’t get the last two at-large bids).  However, it does have an interesting result.

I looked at the NET ranking, the old RPI, the 5 computer rankings on the team sheets (KPI, SOR, BPI, Pomeroy and Sagarin), the AP poll and the Coaches poll and figured out which teams were in the top 36 of each of the 9 rankings.  There were 28 teams that managed to be agreed upon by all 9 rankings (which in retrospect, is actually a lot – would have expected the rankings to disagree on at least a couple teams here and there…..)  Anyways, in could be just a coincidence – and is totally a small sample size for observation (since I don’t know that I want to go back and do this for multiple years).  But those 28 teams were the top 28 seeds in the tournament.  That is right – if you were in this list, you were one of the top 7 seeds in a region, and if you were a team that missed on a couple lists (for example, VCU or Utah St that made 6 of the 9) were 8th or worse.

Kind of makes sense – it all the computers and all the polls agree they are one of the top 36 teams, they should be one of the top 36 teams.  But I think it is interesting that the committee didn’t have a single team break into the list.  If they didn’t think that NET ranked 33 team NC State was one of the 36 at-large teams, I would have thought that at least one team would break into these top 28….

There is another piece of bracket predicting facts that probably ties to this a little.  I only remember two teams in history (Utah St in 2004 and SMU in 2014) that were in the top 25 of the polls in the last week and did not get an at-large bid.  As you can see, both of these schools are from small conferences and were upset in their conference tournaments (by teams the committee obviously felt were not tournament-worthy).

So – lets apply our rules to the 2020 bracket.  This year, the first lock (the 9 rankings in the top 36) didn’t agree – makes sense.  Nothing is going to be easy about 2020.  But there are still 21 teams – they will start our locks (in order of NET ranking)…..

• Gonzaga – WCC auto-bid
• Kansas – projected Big 12 auto-bid
• Dayton – projected Atl 10 auto-bid
• San Diego State
• Baylor
• Duke
• Michigan State
• Louisville
• BYU
• Florida State – projected ACC auto-bid
• Creighton – projected Big East auto-bid
• Oregon – projected Pac 12 auto-bid
• Villanova
• Seton Hall
• West Virginia
• Maryland
• Butler
• Houston
• Kentucky – projected SEC auto-bid
• Wisconsin – projected Big 10 auto-bid
• Auburn
• Ohio State (honorary mention 22nd team) – they were on all but the RPI list (37), and since the committee says they don’t use RPI any more, it is hard to say that they would notice this discrepancy…….

The AP/Coaches Poll Rule also suggests that Virginia, Illinois and Iowa are safe – as the remaining top 25 teams that were left out.  OK – so 25 locks with 17 needing an at-large bid.

There are also all the auto-bids, the way I did this was taking the top remaining seed in each of the conference tournaments (or the team that won their conference tournament if it happened before the madness begun).

So, right now, the bubble is big – still another 19 teams to select……