"My rule was I wouldn't recruit a kid if he had grass in front of his house.
That's not my world. My world was a cracked sidewalk." —Al McGuire

Tuesday, March 07, 2023

Scrubbing the Top-16 and Beyond


Alabama won in November, but Houston has moved up to the overall 1-Seed

Photo by Bob Levey | Getty Images

This week we go down the top-4 seed lines going team-by-team with explanations as to why everyone lands where they do. There's a change at the end of the 1-line, a difficult decision at the end of the 2, and reasoning why these teams are where they are. Finally, we look at the last teams in and why they landed where they did, inside and out.


  • 1. Houston: Do wins or losses matter more? Historically, the answer has always been wins. The Cougars have the best record in the country, the best Q1 win percentage, the best Q1+2 win percentage, and are tied for the best RAP (Resume average And Predictive Average) Score. Typically, the Predictive metrics have more influence on seeding and Houston is #1 in all three predictive metrics. They also have more Q1+2 Away wins than any other team fighting for the top line. Houston is tied for the most road wins in the country while also being undefeated on the road. Houston did better in the opportunities they had than either #2 Kansas or #3 Alabama. They do have a Quadrant 3 loss, but considering how well they did in the first two quadrants that 13-1 winning record offsets the one bad loss. Looking at #2, Houston has a better winning percentage (or more wins with an equal undefeated percentage) in Quadrant 1A, Quadrant 1B, Quadrant 2A, and Quadrant 2B than #2 Kansas does. Since the week of the Top-16 Reveal, Houston is the only team that was on the top line that has not taken a single loss. If wins matter more, Houston's record at the top of the resume should put them atop the S-Curve.
  • 2. Kansas: The Jayhawks are the popular pick to be #1 overall and they were our leader until the latest scrub. They have a record 15 Quadrant 1 wins and winning records in every sub-quadrant, which is not the case for #3 Alabama. Against the nation's toughest schedule, Kansas answered the bell repeatedly. Because they were behind Houston at the Reveal and have lost since, it's hard to put them ahead of the Cougars, but their top-of-the-resume work is better than anyone else. Both Alabama and Kansas have taken knocks since the reveal, but Alabama's poor play even in victories pushes Kansas ahead.
  • 3. Alabama: The Tide are tied for the best Resume Average and RAP Score. Those numbers give them clear separation from #4 UCLA and #5 Purdue. Their 9 Q1 wins, win at #1 Houston (the best win any team can claim), and 16-5 Q1+2 record with no losses outside Q1 is what you want in a 1-seed. This is where the divide is between the clear and definite 1-seeds and the teams fighting to join them on the top line.
  • 4. Purdue: The Boilermakers have 9 Q1 wins to UCLA's 7, 16 Q1+2 wins to UCLA's 15, and have done so against a tougher schedule. That's the rationale I think the Selection Committee will use, but it's not the real reason Purdue lands on the 1-line. It's because one of the 1-seeds has to play on Friday to accommodate the play-in game on Wednesday. Houston, Kansas, Alabama, and UCLA, by virtue of playing in Birmingham, Des Moines, and Sacramento, will all be playing on Thursday. That means someone from the 2-line has to come up to be the last 1-seed. Purdue should be the top 2-seed, but because of bracketing, they land as the last 1-seed and the Selection Committee will justify their inclusion in that spot.


  • 5. UCLA: The Bruins have the best RAP Score by far of any remaining teams. There's a strong case for them as the last 1-seed, but it's impossible due to bracketing, so they get the top 2-seed. There is a gap between them and the rest of the 2-line.
  • 6. Texas: The biggest difference between the Longhorns and #7 Baylor are the metrics, where they have the same Resume average of 6.0 but Texas has a better Predictive Average (7.3 to 12.7) and RAP Score (6.65 to 9.35). Both have 11 Q1 wins, 15 Q1+2 wins, and zero losses outside the first quadrant, but Texas has the better winning percentage in those areas.
  • 7. Baylor: Baylor is neck and neck with Texas. While #8 Gonzaga has a better RAP Score and Predictive Average, the 11 to 5 Q1 win disparity, 15 to 10 Q1+2 win disparity, and no losses outside Q1 while Gonzaga has a Q3 loss keeps Baylor ahead.
  • 8. Arizona: This is one of the most difficult decisions so far. The positives for Arizona are the 8-2 Q1 record, the 10.6 RAP Score (ahead of #9 Marquette-15.1 and #10 Kansas State-16.1), and knowing the Selection Committee had them #6 at the Reveal. The 4 losses in Q2 is a problem, but they already had 3 of those at the Reveal and were still well ahead of the teams they are competing with. Wins trump losses in this case. It gets much more interesting if they are scrubbed against Gonzaga, who is metrically superior in every regard, but it feels like Gonzaga was too far back to land this high.

Tyler Kolek has led Marquette to the top of the 3-line
 Photo from Marquette Athletics


  • 9. Marquette: It's close with #10 Kansas State, but Marquette's edge in Q1A (4-3 to 4-5) and having added a Q1A road win (at Creighton) mitigates the advantage at the Top-16 Reveal. Marquette hasn't lost since the week of the Reveal while K-State has two losses. And while Marquette's worst loss is a Q3 loss to bubble team Wisconsin, K-State's worst loss is at Butler, a team below .500 and not even in NIT consideration. This is a place where the Quadrants say one thing, but anyone looking at the actual result (especially considering K-State's 12-point loss at Butler compared to Marquette's 16-point win in the same building) will again mitigate the impact of those losses. Finally, Marquette's outright Big East Championship is an extra feather in the cap compared to the Big 12 3-seed.
  • 10. Kansas State: Kansas State has just 3 road wins all year, but two of them are at Texas and Baylor, teams on the 2-line. Their 9 Q1 wins is the most of any team not already in the field. The drawback is neither their Resume nor Predictive averages are as good as #11 Gonzaga, but the value placed on true road wins at the Reveal keeps the Wildcats ahead.
  • 11. Gonzaga: The drawback for the Zags is starting further back than the competition. They have RAP Score edge not just on every team left in the field, but also the three teams ahead of them. However, since the Reveal the only real win of substance they added was St. Mary's at home. Even if they win the WCC Final, that is a neutral court win and wouldn't seem to tip the scales more than the Alabama and Xavier wins they already have on neutral courts. Their placement here seems secure, but the only reason to move them up is metrics for a Committee that clearly placed value on the true road wins Gonzaga just doesn't have.
  • 12. Tennessee: It's another razor-thin decision between #12 Tennessee and #13 UConn. The Volunteers have the slightly better RAP Score (9.25 to 9.65), the slightly better Q1 record (7-6 to 6-6), and the slightly better best road win (NET 46 Mississippi State to 59 Florida). The real deciding factor was record against the field. Tennessee is 9-6 while UConn is just 5-5. In addition, we know Tennessee was a 3 at the Reveal while UConn was outside the Top-16, so it's a much higher hill to climb for the Huskies. It's possible the Zakai Zeigler injury will knock them down, but they did already beat Arkansas without him, so we aren't going to punish them for that just yet.


  • 13. Connecticut: Metrically, UConn's 9.65 RAP Score is far better than anyone else at this point (Xavier is next best at 16.85). Their 4 Q1A wins are as many or more than anyone else being considered here but they also have just one loss outside Q1, which is fewer than #14 Xavier, #15 Indiana, or #16 St. Mary's. UConn is the last team in contention for a 3-Seed, while everyone behind them seems capped at a 4.
  • 14. Xavier: Since the Reveal, Xavier has added their two best road wins, Q1 results over Providence and Seton Hall. Their 13-6 Q1+2 record is better than #15 Indiana or #16 St. Mary's. They have the highest remaining RAP Score. While the recent news of Zack Freemantle's surgery isn't helpful, they have already shown their ability to succeed without him.
  • 15. Indiana: The Hoosiers are largely held aloft by their sweep of Purdue, though true road wins at Xavier and Illinois are also incredibly valuable. That overall body of work has them ahead of #16 St. Mary's, who has zero Q1 road wins. One word of warning for the Hoosiers, no team with 11 losses has reached the 4-line since 2016 Iowa State. Without a Big 10 title, that would describe Indiana. Sheer volume of losses could see them fall to a 5 behind a team like Virginia or San Diego State.
  • 16. St. Mary's: Winning the WCC Tournament would likely secure this spot. St. Mary's win over San Diego State compares favorably to #17 Virginia's Baylor win, and the Gonzaga win St. Mary's has is much better than anything else on the UVA resume. Metrically, the Gaels also have a better RAP Score than #18 San Diego State or #19 Kentucky, and it leans on their better Predictives, which are more seed dependent.

Jalen Pickett & Penn State are on the right side of the bubble

Photo from Penn State Athletics

The Bubble

  • 40-Utah State: Only 1 Q1 win, but 9-5 v Q1+2 and a stellar 21-5 v Q1-3 while being better in both Resume and Predictive averages than anyone else on the bubble.
  • 41-Mississippi State: The Marquette win is huge and everything else is good enough. With virtually everyone having fatal flaws, the Bulldogs simply don't.
  • 42-Penn State: Two wins over the field in the past week moved them up significantly. 5 Q1 wins, 6 v the field, and metrics that finally match an at-large pedigree. They just avoid Dayton.
  • 43-Nevada: Their opportunities are few, but 4 Q1 wins, 8-7 v the top two quadrants, and 17-9 v the top three separates them from the field, especially with zero Q4 losses.
  • 44-NC State: Their resume is not great. The numbers check out, but 2 Q1 wins and 7-9 v the top two quadrants isn't exciting. Zero losses in Q3/4 helps, but if there are bid thieves they are in danger.
  • 45-Wisconsin: This is the team I just can't get out of the bracket. 6 Q1 wins, 5 against the field's top-40, and 3-2 against other bubble teams. Their resume numbers are good enough, but barely. They could play their way out with an early loss in the Big 10 Tourney.
  • 46-Arizona State: The last team in is only here because of the half-court buzzer beater over Arizona. They have multiple losses in the Q1A, Q1B, Q2A, and Q2B, but their only loss outside that is a true road loss in Q4. This is another team that could play their way out.
  • OUT-Pittsburgh: Rescrubbing, their resume is not very good. 7/10 losses are to non-tourney teams and their Resume Average isn't good enough to prop up two barely Q1A wins. They are close enough that if there are no bid thieves and someone in the final few spots plays their way out, Pitt could still sneak in.
  • OUT-Rutgers: Last year they got in with a similar Resume average and Q1 win total, but most of those wins were Q1A, whereas this year they only have 2 located there. They have been a different team since losing Mawot Mag, and that has to matter.
  • OUT-North Carolina: 1-9 vs Q1 really says it all, but it was more palatable when they were perfect in Q2. They are 1-3 in Q2 since February 1, and that 7-12 Q1+2 record isn't good enough. They need the ACC Tourney title to get in.
  • OUT-Oklahoma State: While their quadrants look similar to West Virginia on the 8-line, the computer numbers aren't enough and 5 of their 9 Q1+2 wins are against teams outside the at-large pool. Ultimately, 15 losses (which they will have without the auto-bid) are too many for a team without much of a margin to start with.

Here's the complete S-Curve as it stands:

No comments: