WNUR Bracketology 2020 – Week #6: Trust No One
March is upon us! Judgment day nears. I am fearful. Take a look at the (slightly stale) projections and then I will explain.
The Usual Disclaimers
- This is what I think the committee would do given what we know about each team. I do not necessarily agree with the committee’s projected evaluations of these teams.
- My bracketology does not aim to predict what will happen; it is a simulation of what would happen if the season ended
todayafter Tuesday March 3rd - Asterisks in the graphic denote conference champions (determined by highest remaining seed in conference tournament. For conferences still in regular season play, it’s determined by fewest losses in conference play, with NET as a tiebreaker).
- Projections made based on data entering Wednesday, March 4th
- Please feel free to @ me on Twitter but keep all of this in mind when you do.
The Projections
The Column
With Selection Sunday now just 10 days away, I am dealing with three bracketology crises at the moment.
Crisis #1
Crisis #1 is at the 2-seed line. The top 5 teams are very clearly Kansas, Baylor, Gonzaga, San Diego State, and Dayton, in that order. Then, there’s a question of which teams are #6, #7, and #8. Unfortunately, every candidate for those three spots seems to be actively avoiding them at the moment. Three of Maryland, Duke, Florida State, Creighton, and Villanova were primed to round out the 2-line in this week’s projections, and then every single one of them lost over the weekend.
Seton Hall surged ahead of all of those teams. Then, last night, the Pirates lost at home to Villanova, but not before I had already made this week’s projections. I’m completely lost.
crisis #2
Crisis #2 has to do with the bubble. The last two spots this week came down to four teams: Rutgers, Purdue, Cincinnati, and UCLA. All four of these teams are special cases because their resumes contain both clinchers and deal-breakers.
Rutgers is 32nd in NET, 38th in SOS, 18-11, and has four wins against really good teams. They have to go in! But they still have just one win away from home this year, and that was at Nebraska (194th in NET). I can’t let them in!
Purdue’s record is 16-14, there’s no way I can let them in. But I’ve never kept a team that’s top 25 in KenPom, Sagarin, and BPI out of the tournament. How am I supposed to snub them?
Just about everyone in the Bracket Matrix says the Bearcats are in. But they have just two Q1 wins and FOUR quad 3 losses. I’ve been hung up on that for weeks. Am I just being too stubborn about these bad losses? Maybe I am, maybe I’ll let them in. But am I really just going to abandon my gut, my principles, and everything I’ve been saying about them for the past month? I don’t know.
People are really high on UCLA. The Bruins, my beloved Bruins. The team I was raised on; the team I have always loved; the team I have cheered for; the team the team I have cried for; the team of my family; the team that runs in my blood. They have 6 Q1 wins, they’re 18-11, they’re one of the hottest teams in the country. But their metrics are atrocious and they’re 76th in NET. No one has ever made the tournament with an RPI or a NET that low. There’s no way they’re in right now. Right? Or am I overcompensating for my bias?
I can’t trust myself with these teams. I’m running circles in my head, fighting wars in my mind. I’m supposed to be confident now that it’s March. And yet every time I look at my spreadsheet it’s like:
crisis #3
Crisis #3 has to do with building the bracket itself. Two years ago, I registered my best-ever performance in the Bracket Matrix not because my final projections correctly named 66 of the 68 teams, but because out of those 66 teams I had seeded 65 of them either correctly or off by one seed line. My only big miss was Butler, who ended up as a 10-seed.
But I wasn’t that far off on Butler. As a matter of fact, I had them ranked 30th overall (making them an 8-seed), just three spots away from what the committee ranked them, 33rd.
But Jake, you ask, I just did the math in my head. If Butler was 33rd overall, that would make them the best 9-seed. How did they end up as a 10-seed?
Astute observation! Things go a bit haywire in the bracket when too many teams from the same conference finish close to one another in the overall rankings. That’s because regular-season rematches are not allowed in the first round, and teams who played each other twice are not allowed to play each other in the first two rounds.
In other words, teams from the same conference usually cannot be placed in the same pod. Pods are (almost always) made up of four teams: the two teams playing each other in the round of 64 plus the two teams they could potentially meet in the round of 32.
The 1-seed, 8-seed, 9-seed, and 16-seed in a region are one pod. The 4-, 5-, 12- and 13- seeds in that region make up another pod. Same goes for 3/6/11/14 and 2/7/10/15.
There are 16 pods: those four seed-clusters in each of the four regions of the bracket.
This means that, for example, the 1-seed, 8-seed, 9-seed, and 16-seed in the East region should all be from different conferences. If Duke was the 1-seed in the East and NC State was an 8-seed, you would put NC State in a different region from Duke so they don’t meet up in the second round.
But remember, there are only 4 regions. Which means you can’t have more than four teams from the same conference in the same seed cluster, otherwise you’ll have a region where two teams from the same conference get placed in the same pod.
Two years ago, the committee ran into this problem with the Big East. Villanova was a 1-seed in the East. Xavier was a 1-seed in the West. Seton Hall was an 8-seed, and they got placed in the Midwest. Creighton was also an 8-seed, so they had to be placed in the South.
Next up was Butler, #33 overall. Butler was supposed to be a 9-seed, but there was a problem. If Butler were made the 9-seed in the South, they would play Creighton in the first round, which isn’t allowed. If Butler were made the 9-seed in the Midwest, they would play Seton Hall in the first round, which isn’t allowed. If Butler were made the 9-seed in the East, they might play Villanova in the second round, which isn’t allowed. And If Butler were made the 9-seed in the West, they might play Xavier in the second round, which isn’t allowed.
That left the committee with no choice but to knock Butler down to a 10-seed.
Situations like these dramatically shrink bracketologists’ margin for error. Think about it: I ranked Seton Hall #27 overall, making them a 7-seed, taking them out of the 1-8-9-16 pod, which meant Butler could stay in the 1-8-9-16 pod and not have to move down. Instead, Seton Hall was #29 overall, making them an 8-seed, which meant Butler dropped to a 10-seed.
Simply by misranking a team by just two spots, I ended missing an entirely different team by two seed lines. Putting Seton Hall above Nevada and Rhode Island cost me 7 points in the Bracket Matrix. I would have been the #1 bracketologist in the world for 2018.
I don’t like those situations. I didn’t have to deal with any last year. But this year, the Big Ten is problem. This week, EVERY SINGLE FIVE-SEED IS A BIG TEN TEAM. Here’s what that did:
The First Four matchups are supposed to be Wichita State vs Richmond for an 11-seed and Rutgers vs Purdue for a 12-seed. But Rutgers and Purdue can’t play each other because they’re from the same conference, so instead I switched the First Four matchups so that Wichita State would play Rutgers for an 11-seed and Richmond would play Purdue for a 12-seed.
But what if Purdue beats Richmond? Purdue would be a 12-seed, so they would have to play a 5-seed in the first round. But all four 5-seeds are Big Ten teams. So Purdue vs Richmond also has to be for an 11-seed. But now we have an extra 11-seed and a missing 12-seed. So Texas moves down from an 11-seed to a 12-seed.
But now we have another problem. Michigan State and Maryland are both 3-seeds. That means we can only have two more Big Ten teams as 6-seeds, 11-seeds, or 14-seeds. But Indiana is an 11-seed, and we just made it so that both Rutgers and Purdue are also playing for 11-seeds. And we can’t move Rutgers or Purdue down to a 12-seed because all four 5-seeds are Big Ten teams, that’s why we’re in this mess in the first place. So that means Indiana gets moved up from an 11-seed to a 10-seed, and Providence moves down from the 10-line to the 11-line.
In summary, Indiana moves up one line to a 10-seed, Providence moves down one line to an 11-seed, Rutgers and Purdue both move up one line to an 11-seed, and Texas moves down one line to a 12-seed.
Almost none of those things happen if, say, Maryland’s actually a 2-seed or Michigan State’s actually 4-seed. Or if UCLA and Cincinnati make the tournament instead of Rutgers and Purdue. One mistake can cost me dearly.
The Madness has already begun.