Welcome to the first post-top-16-reveal Bauertology!
Throughout January and early February, bracketology is a little bit of a guessing game in regards to what the selection committee most values. We kept in mind that last year’s bracket saw selection of teams mostly impacted by résumé (quality wins, strength of schedule, etc.) while seeding was more so impacted by metrics (KenPom, BPI, etc.), so we ran with that as one of our premises for bracketing.
But even then, forecasting got a little fuzzy, especially when it came to the bubble. How do we know whether or not the committee is going to value the résumé of a team like North Carolina (solid metrics but lacking in quality) more or less than a team like Rutgers (lots of top-tier wins but with bad metrics and bad losses)?
Well, we finally have some clarity thanks to the yearly top-16 preview. This is a great litmus test for bracketologists, as this reveal gives us a taste of what the committee most cares about when it comes to evaluating teams. Overall, there weren’t too many shocking things about the top 16, but there were a couple key takeaways that are going to help us create brackets going forward.
Here are the main takeaways I had from the preview show:
- A team’s NET ranking doesn’t seem to matter all that much when it comes to seeding. It may have some minor impact, as it likely helped No. 1 Gonzaga take the top spot over No. 9 Auburn, but the reveal showed that the committee isn’t going to value that number as much as other factors. There was a debate about whether or not NET No. 4 Houston would be included in the top 16; they were not, and that’s due to their zero Quad 1 wins, allowing teams lower in NET like Texas (No. 15) and Providence (No. 27) to take a spot instead. Duke was also ranked inside the top 8 despite their NET of No. 12, which is the lowest ever NET ranking for a preview show top-8 team.
- Metrics are going to matter a lot when it comes to seeding. This seems like a bit of continuity from last year, which is helpful. The committee is going to place a lot of value on what KenPom, BPI, Sagarin, KPI, and SOR say. They said the margin of difference between Gonzaga and Auburn for the number one spot was “razor thin.” But in spite of Auburn having arguably a better pure résumé, the committee respected Gonzaga being the No. 1 team across the board in the three predictive metrics (KenPom, BPI, and Sagarin), and went with the Bulldogs at No. 1 overall. Likewise, teams that have struggled to impress the predictive metrics, namely Wisconsin and Providence, were seeded lower; both the Badgers and Friars were on the 4-seed line in spite of the strong wins and win totals that they’ve put together.
- While metrics will matter in seeding, backing those metrics up with a quality résumé is even more important. Apparently you can’t survive on good metrics alone. Take Houston, for example. All three predictive metrics pen the Cougars as a top-10 team, just like NET. But Houston was absent from the top 16, due to their absence of a Quad 1 win on the team sheet. On the flip side, both Kansas and Baylor received favorable spots (No. 4 and No. 5 overall, respectively), as their résumés are teeming with Quad 1 wins and good metrics. Some bracketologists had both Kentucky and Purdue penned above Baylor in the s-curve, but the committee likely went with the Bears for their nine Q1 wins and top-6 metric numbers across the board.
- Not all Quad 1 wins are created equal—the committee is going to place more value on elite victories. When you look at a blind résumé test, all you get is a simple number for Quad 1 wins. That number doesn’t tell you whether a Quad 1 win is a tight home victory over NET No. 29 San Francisco or a dominant victory on the road over NET No. 5 Kansas; they’re the same in the eyes of the blind résumé. Those are not the same in the eyes of the committee. If you can put together elite wins, you’re going to receive a higher seed, even if other résumé factors say you shouldn’t. The best example is Duke, who most bracketologists had penned as a 3 seed, below Villanova and Texas Tech. From a pure numbers standpoint, it makes sense—both Villanova and Texas Tech had more Quad 1 (and Quad 1A) wins than Duke, and neither had a Quad 3 loss like the Blue Devils did. But the committee went with Duke for the final 2 seed, citing their neutral-court wins over Gonzaga and Kentucky (both top-6 teams) as the reason why. The committee won’t just be looking at the win total digit when it comes to Quad 1; they’ll be diving deeper into the quality of those victories.
- Teams that had to deal with injuries to key players may get a bit of a pass. One of the most surprising aspects of the reveal was seeing Illinois as a 3 seed, ahead of Wisconsin. In addition to a head-to-head victory over the Badgers, the committee mentioned Illinois playing without Kofi Cockburn and Andre Curbelo for a significant period of time as part of the reason why. This could be the committee saying that they value Illinois as a top-3 seed when those guys are healthy and on the court, and the losses they took with them absent aren’t as reflective of this team at full strength. I’m not sure I totally like or agree with this point, but at least there’s some justification behind it. Why then, didn’t Houston get more leeway for the losses of Marcus Sasser and Tramon Mark? Perhaps because those were season-ending injuries, and Houston’s “full strength” is now without those two. Anyway, this point shouldn’t be too much of a factor, as there have been very few teams as injury-ravaged as Illinois (maybe Florida or VCU), but it’s something to keep in mind.
- Head-to-head shouldn’t matter all that much. This is also in line with years past. Yes, I just got done telling you that one reason that Illinois was ranked above Wisconsin is because of the Illini’s head-to-head win. This seems like an exception though, only pulled out for usage when two teams are right next to each other on the s-curve. If head-to-head were a primary factor, Kentucky would be ahead of Kansas, since the Wildcats beat the Jayhawks 80-62 in Lawrence, and Texas Tech would be ahead of Baylor, since the Red Raiders swept the Bears this season. But the full résumés are what matter more, and Kansas has a superior résumé to Kentucky; likewise for Baylor and Texas Tech. So we may consider head-to-head when it’s really close between two teams, but otherwise, we don’t have to think about it very much. Again, this has been a consistent year-to-year thing for the committee.
With those findings now in our possession, we can use them to create our most well-informed bracket of the season. That’s why you’re going to see a couple drastic changes, especially near the bubble, from what I was projecting beforehand.
It’s also worth mentioning that wins and losses and résumé numbers have already changed since the top-16 reveal (namely Auburn losing to Florida), so there will be some changes in that grouping too… though not as many as I thought there would be.
Lengthy introduction over! Here’s the Bauertology projection for Monday, Feb. 21. Please ask questions if you have them! Feedback and constructive criticism are always appreciated.
THE BRACKET
THE TRUE SEED LIST, BUBBLE, AND BIDS BY CONFERENCE
