Cracketology: Reveal Reset

The NCAA Selection Committee released their midseason Top-16 on Saturday. Let's rundown the teams that were included, then see if there are any takeaways from the reveal.

1-Seeds: Auburn (1), Alabama (2), Duke (3), Florida (4)

Auburn, Alabama, and Duke were the easy selections. The only question was who would be the last 1-Seed. The Committee went with Florida. As we said in our preview, the Gators were a team we thought could jump up to the last 1-seed if they included Tuesday results. Clearly they did.

2-Seeds: Tennessee (5), Texas A&M (6), Purdue (7), Houston (8)

The first three were expected, the only surprise with Houston is they were that far back. The Cougars don't have a great collection of wins to support their elite metrics, which has to be what cost them. With the Arizona win and Texas Tech, Iowa State, and Kansas still on the schedule, there's room for them to move up.

3-Seeds: Iowa State (9), Kentucky (10), Wisconsin (11), Arizona (12)

We had Iowa State, Kentucky, and Wisconsin here, though in a different order. Arizona was our first 4-Seed. Most likely this means the Arizona win at BYU Tuesday was included.

4-Seeds: Texas Tech (13), Michigan (14), Kansas (15), St. John's (16)

We had Michigan State in and Michigan out, but as mentioned in the preview, which Michigan school is a protected seed would tell us which games were included. If this comes out Sunday or Monday, it would've been MSU. Because it was Michigan, we know their win over Purdue was factored in.We also had St. John's too high at 12. The loss to 'Nova on Wednesday probably knocked them down, but this spot indicates they were probably always a 4-Seed, indicating that being Big East leaders didn't help them much (similar for Michigan).

In Consideration: Michigan State, Marquette

These teams were said to be right there for the last spot in the Top-16. Both have winning records in Q1 and losses in Q2, which stands out against Ole Miss, who was not mentioned but came into the Reveal better metrically but with with a losing Q1 record and no Q2 losses.

Takeaways

• Wins Over Metrics: Houston had better predictive metrics than Florida, Tennessee, Texas A&M, and Purdue, but was behind them with only one Q1A win (the teams ahead all have at least three). Similarly, Kentucky was on a 3-Seed with a bevy of good wins despite being a metric 4-Seed. This is the exact opposite trend we saw last year.
  
• Focus on Resume (Metrics): The resume metrics were better seed predictors than the predictive metrics. SOR (5 exact, 13 +/-2 spots, 2.3 average away) and WAB (4 exact, 13 +/-2 spots, 2.2 average away) were the metrics closest to the Selection Committee's seeds. BPI was worst (0 exact, 7 +/- 2 spots, 4.6 average away).
  
• RAP is Still King: The average of the two averages was still the best predictor, with 7 exact, 15 within 2 spots, and an average of 1.3 away. If you used RAP as a strict guideline, 16 of the 18 teams mentioned would have been on the correct seed line, with only Kansas and Kentucky flipped.
• Marquette's Future: Chairman Bubba Cunningham mentioned Marquette as "right there" for the #16 overall spot. A 4-seed is still a very realistic goal, while a 3-seed is possible if they get hot (think 7 more total wins). I do think getting back to a 2-seed is probably unrealistic unless they get a lot of help.
  
• Championship Week Might Matter: We put the bracket together with Monday's data because historically, the Selection Committee hasn't included later in the week results. It is clear they at least used Tuesday data and likely used Wednesday results as well. Might this indicate that Championship Week results on Thursday, Friday, and maybe even Saturday will be factored into the final bracket? It seems more possible if this Committee is capable of processing data beyond what's immediately available when they walk into their meeting room.

So what did this mean further down the bracket? We feel confident with Memphis as a 6-seed thanks to their Q1 wins, even though their metrics would be an 8-seed. On the bubble, we looked closer at Q1 win percentages and teams with Q1-3 winning records. Vanderbilt, Arkansas, and Georgia were all in our last field, but now all are below .300 WP% in Q1 and have losing records in Q1-3. As a result, the SEC is down to 11 schools, with Texas, George Mason, and UC Irvine replacing them. Texas' late win over Kentucky lifted them into the field. BYU also moved in with their hot streak of late (including a win over K-State last night), as did Wake Forest after winning at SMU.

For the Big East, Xavier losing at Villanova was their last Q1 opportunity, which likely ends their at-large hopes. There just isn't enough meat on the Musketeer resume. Villanova's own tourney hopes took a hit yesterday with the loss at Providence. If they win their next two (at UConn, vs Marquette) maybe they can get another look, but for now they are under water and continuing to take on losses outside Q1 isn't helping.

Finally, let's see where we stand now with both what we learned from the Selection Committee and Saturday's results included:

Multibid Leagues

SEC: 11

Big 10: 10

Big 12: 8

ACC: 4

Big East: 4

Mountain West: 3

A-10: 2

Big West: 2

WCC: 2