NCAA Weekly, Vol. 8.2: How the 2023 NCAA Tournament's locations matter for this year's field
seven swans, eight "sub-regionals"
When I was first getting really into the March Madness weeds - 2010 or so - there was this site called Crashing the Dance. There still is such a site, but it’s updated infrequently. It was a tremendous resource before I knew what Torvik or Haslametrics or whatnot were, but its best aspect was a page where it showed you the distance from a team to a potential site they might play at.
This was extraordinarily useful for attempting to figure out what teams might want what pods/regional sites. I can look at a map, because I am an adult, but I cannot look at a map 500 times in a row to figure out everything in a quick fashion. The point is less about that than what an incredible resource it was; it no longer being updated with regularity starting on New Year’s is quite the bummer.
So: I’ve decided to do my own. One of the few drawbacks of Substack is that Google Sheets do not embed properly, so here is the link, but below is a screenshot of what we’re looking at.
What this sheet offers is an overview of the top 50 teams in the Massey Ratings Consensus and their approximation to the eight Round of 64/Round of 32 sites, as well as the regionals (Sweet Sixteen/Elite Eight). I’ve attached this as a free file and made this a free post for any writers or journalists who would like to use it, as I’m not actively aware of anything else publicly available in this domain.
Below this is a dissection of the data itself: what it means, which teams could get a particularly good or bad draw, which regions could be represented by which teams, etc.
The eight sites themselves
We’ll break these down into which days they’re on, starting with
Thursday/Saturday sites
Birmingham is first up alphabetically and is in a weird position. My assumption is that as long as neither offer up a complete collapse down the stretch, at least one and only one of Tennessee/Alabama will be here. I’m of the opinion that Alabama is significantly more likely to end up in Birmingham than Tennesssee, simply because Tennessee has a pair of other backup sites within 400 miles and Alabama does not.
Who will be joining Alabama in Birmingham is an unusual thing. Of the nine teams with at least a 10% chance of a 1 seed, per Torvik, Houston, Arizona, and Texas all fail to reside within 500 miles of one of the first-round sites. That’s not their fault, of course, but it is a fact. Arizona will likely go to a West Coast site. Because it’s much more likely Houston ends up above Texas than vice versa and this is Houston’s closest site by nearly 300 miles, I think Houston ends up in Birmingham alongside Alabama.
Des Moines will be hosting Kansas, as this is the only site Kansas is within 500 miles of and there’s a >90% chance Kansas ends up a 1 or 2 seed. Past that, they’ve got some options. Iowa State is not ruled out because Drake University is technically the host, but I’d imagine it unlikely that a team currently on track for a 4/5 seed is given a straight-up home game for two rounds.
The next-most likely candidates to join Kansas are all on the 3/4 lines. Arkansas, Kansas State, Marquette, and even Missouri are all closest to this site among top-30 options. If it’s not one of them, it’s a wild card who couldn’t find a home elsewhere.
Orlando is a host in a rare year where there may be zero Florida/Deep South teams among the top-flight of the sport. Their sincere hope may be to somehow steal Miami as a 4 seed, and let it be known that I absolutely do not believe in such happening. (They’re 44th on KenPom, three spots ahead of Colorado.) Among the top 28 teams on Torvik’s season-long bracketology, Auburn is the only team within 500 miles. This is a site ripe for a pair of 4 seeds who couldn’t find a home anywhere else; it also stands to be an Alabama backup site if they deem Tennessee is more deserving of Birmingham.
Sacramento will very likely host UCLA because they need to sell tickets. Beyond that, it’s either a wild card (Arizona comes to mind) or hoping that one of Gonzaga/Saint Mary’s (CA) go on a run in the next two months.
Friday/Sunday sites
Albany will be hosting UConn. Beyond that, I have no idea. I think they’re clearly hoping that either Rutgers (176 miles away) or Providence (163) emerge from the pile to be top-4 seeds so they can steal one of those. The next-closest offering is Virginia, who is barely under 500 miles away and has a pair of closer sites to potentially choose from. This has the vibe of UConn plus a random team that got bounced from other, closer options.
Columbus will be the most competitive spot on the landscape this year, as it’s within 500 miles of 10 (!) of the top 30 teams. (This is obviously because Columbus is within 500 miles of approximately 97% of the US population.) Purdue feels like a pretty obvious pick for one of the two top-4 seeds to go here. Beyond that, I assume that the people who sell tickets for this are overwhelmingly rooting for Xavier (under two hours away) to claim a 3 or 4 seed so they can take them in. If Xavier ends up a 5 or 6 seed, this is perhaps the domain of an Iowa State or a Virginia. This would be Tennessee’s third-choice pick if you’re keeping score.
Denver has no teams in the current Massey top 50 within 400 miles of the site, meaning that unless you’re able to get some really invested locals to come through the gates (or the teams they get have fanbases with a lot of disposable income), it may well be the least-attended site for this year’s Tournament. Any two teams on the 3/4 seed lines could end up here and I would not be surprised. This could also be a potential Arizona/Gonzaga landing spot.
Greensboro is tied for Tennessee’s #1 choice, but we’re guessing that Birmingham is relatively unlikely. No other top-15 seeding team is within 400 miles of the site, and considering this is about four hours from downtown Knoxville, it’s not a bad guess to think this is where the Vols eventually end up.
Who joins them there is an interesting question. Ohio State cannot go to Columbus, making Greensboro its next-best destination, but I have some doubts Ohio State ends up among the top 16 teams. North Carolina and Duke would be obvious picks, but neither is on target for the top 16. (They can still go here to sell tickets, but it would be very unfair to their higher-seeded counterparts.) Virginia is even closer than Tennessee, so I guess they have a real shot. Beyond that, this is an option if Xavier and Columbus somehow don’t link up.
A SPECIAL NOTE HERE THAT ALWAYS GOES UNDERREPORTED: Because the First Four is structured so that one 16-seed game is on Tuesday and the other is on Wednesday, it’s an unwritten requirement that at least one 1 seed plays on Thursday and Friday. It’s not as much of an issue this year due to how the top teams appear to be settling out, but keep this in mind when you see a bracket that may have a weird “X team got sent to Y site” situation. The same thing applies for at-large teams; the goal is to send them to their appropriate partners, but sometimes (like last year) you give Indiana/Wyoming 12 seeds because the one 6 seed playing on Thursday (Colorado State) would’ve had a conference conflict with Wyoming.
How the regionals may/may not be decided
There are four regional sites this year, with the appropriate directional denomination in parentheses: Kansas City (Midwest), Las Vegas (West), Louisville (South), and New York (East). Las Vegas and New York are Thursday/Saturday sites, while Louisville and Kansas City will play on Friday/Sunday. There are potential minor rest advantages at play if you’re a team that plays your R64/R32 slate on Thursday/Saturday then you go to a Friday/Sunday regional, but if they exist statistically I’ve yet to see it.
As of the time of writing, Torvik gives the following nine teams a >10% chance of eventually earning a 1 seed:
Houston (68.7%)
Kansas (60.3%)
Tennessee (57.8%)
Purdue (55.9%)
Connecticut (46.4%)
Alabama (41.4%)
Arizona (18.7%)
UCLA (16.3%)
Texas (15.3%)
My personal opinion is that you could eliminate Texas from this list due to their coaching drama, but we’ll leave them in for the sake of completion. In Torvik’s TourneyCast database (dating back to the 2017-18 season), only one team on January 11 with a sub-10% chance of getting a 1 seed ended up with a 1 seed: 2017-18 Xavier. It’s very likely that if you’re going to end up among the top four teams, you’ve already shown something by now to end up near the top of that list.
In order, we’ll break down the preference each side will have for their regional, using the map data I’ve accumulated so far.
Houston’s main preference is simply to make it to the Final Four, which will be held in early April in…Houston, Texas. So that’s a nice thing for them. As for the regionals, both New York and Las Vegas (1,400+ miles away) seem suboptimal, leaving Friday/Sunday options Louisville (South) or Kansas City (Midwest). This gets interesting fast when you consider the other teams at the top of the list here. Kansas City is closer and certainly more (future) Big 12 territory than Louisville, so I imagine that’s their preference.
Kansas, meanwhile, would kick your dog if it meant a second weekend in Kansas City, which the campus is all of 40 miles from. If Kansas is the #1 overall seed, that will be where they go in 100 of 100 simulations. If Kansas is out-ranked by Houston, it gets interesting quickly. Similar to Houston, they’d prefer to avoid New York or Las Vegas…which would leave Louisville (South) as the secondary option.
Tennessee is relatively unique in that they and Purdue are the only teams on this list to be within 1,000 miles of three of the four sites. That aids their flexibility, I suppose. But we all know Tennessee badly wants to go to Louisville, as that’s far closer than any other option. The secondary option is a near-equal split between New York (707 miles away) and Kansas City (731), but if Kansas and Houston both end up as 1 seeds alongside Tennessee, I cannot picture them sending Kansas to New York just so Houston can enjoy Kansas City while Tennessee gets Louisville.
Purdue, like Tennessee, wants Louisville, which is about three hours from their campus. Everything else is secondary, but Kansas City (under 500 miles away!) is not a bad backup option. The only option that’s legitimately very far away is Vegas, but if you’re a Purdue fan and your team is a 1 seed for the first time since 1996, I imagine you would travel happily. Same goes for Tennessee for the record.
Connecticut’s first, second, and third preference is to play in Albany in the first two rounds and NYC in the second weekend. The next-closest option is Louisville, which is nearly 900 miles from their campus.
Alabama will obviously want Louisville as that’s the closest option, but Kansas City is a somewhat compelling backup. I don’t think they would be very happy to be in either of the other two.
Arizona and UCLA both have Las Vegas as their #1 destination.
Texas is unique in that only Houston has a lower closest destination. I suppose their preference would be Kansas City, but it’s not an ideal year for a state of Texas team to be a 1 seed. Great year if you make the Final Four, obviously.