The Big East is a Big Mess

With just under a week of Big East regular season basketball to go, I’m going to look at how the remaining nine games left to play in the conference slate will shake out. Because we at Thompson’s Towel like the computers™, I’ll be simulating the remaining nine games of the Big East slate by using an algorithm and advanced statistics from KenPom. From those simulations, I’ll be talking about how the Big East Tournament seeding will turn out and all the different rules that govern the tournament when there’s a tie (and based on the simulations, there’s going to be lots of ties). Let’s get started!

If you want to read my nerd talk and learn more about KenPom and using it as a methodology to predict games, click on this link for that part of the post. If you just want to see the predictions according to the algorithm for the rest of the season, just keep scrolling to the next paragraph.

Okay, let’s get cooking.

Why am I doing this, and why am I using KenPom? Because I hate myself. Because sites like KenPom are favored by Vegas (hey, money talks) in establishing the betting lines, as well as KenPom’s advanced statistics being some of the most accurate available for college basketball. Is it entirely accurate? Nope! KenPom itself notes that we shouldn’t look too much in any score that’s closer by a margin of eight points, since prediction models can’t get much more accurate than that. However, we’ll still be using KenPom data in the model

I should note a few more things; KenPom’s statistics just love some teams way more than others. For example, KenPom’s ratings rank Creighton (at the time I wrote this) as the 50th best team in the country (and third best in the Big East). Georgetown’s ranked at 78th (good for being the 8th best team in the Big East, ahead only of Providence and DePaul). Why is this? It’s not a conspiracy, but rather because the rankings themselves have to prioritize certain things (otherwise they could never rank anything!), and KenPom in particular prioritizes what it calls the “Four Factors.”

I promised I’d save the nerd talk for somewhere else, but this is actually important to understand. These “Four Factors” are effective field goal percentage, turnover percentage, offensive rebound percentage, and free throws to field goal attempts made. These four things comprise the offensive efficiency calculations. It just so happens that Georgetown isn’t great, comparatively speaking, at these four things, and the KenPom rankings don’t count things like “grit”, “height of coach”,  or whether one of our “guards wears a sweet headband.”

Below is a table of the remaining games, and what the algorithm calculates the final score of each game to be.

You might’ve noticed that the Georgetown-DePaul game would result in a tie. Obviously that doesn’t happen in basketball (we’re not soccer!). This is decidedly non statistical, but I can’t think of any other way to break that tie between Georgetown-DePaul above. Statiscally, they’re pretty much dead even (I told you KenPom doesn’t like the Hoyas, this is why we can’t trust the computers), so I’ll take this as a good ol’ fashioned opportunity to (1) mock DePaul and (2) make my prediction for the game. An empty WinTrust Arena does not make for a home court advantage (which would change the algorithm if we removed home court advantage), Dave Leitao cannot coach, DePaul will almost certainly melt down, and DePaul got their “good game” out of the system when they beat St. John’s. I think the Hoyas will come in and beat DePaul (especially if they use Pickett, whose defense would be an asset and DePaul hasn’t seen either, as he was a DNP last Wednesday).

But to be fair, later in this post if the outcome of the DePaul game affects the standings, I’ll also break down what the results would be if DePaul won. After all, they are still at home, coming off a win against St. John’s while Georgetown has yet to string together three wins in a row in Big East Conference play. It could very well be a trap game. So while for the rest of this post I’ll also develop an alternative bracket and standings where DePaul does win that game, because surprisingly, a game that DePaul plays in has a lot of implications for the standings.

Now, here’s a table of what the final Big East standings would be if all games go according to this table. The games almost certainly won’t shake out as the computers think it will, because real life doesn’t take place in a vacuum. Some people are tired, some people find it in themselves to light up one particular team (as seems to happen against us so often), but KenPom’s advanced stats are the handiest predictive tool we have to take the best possible guess at how these games might go. And as I’ve noted, the games can really swing either way within a margin of eight points, but we’ll roll with these predictions for the sake of understanding all the madness that could unfold with the Big East Tournament bracket.

Note the records in parentheses are the records as we simulated for the last games of the season, and have been factored into the “final” Big East conference record. This is the simulated season, with the Georgetown-DePaul game winner as Georgetown.

Holy moly! There’s 3 separate ties in the standings between 8 different teams, which means that for the Big East Tournament seeding, the tie-breaker procedures will need to be undertaken. It’s pretty complex, and there’s a whole document that governs what goes into this. Let’s work through this, and break each tie.

Villanova and Marquette, Tied for First Place

It’s the battle between the guy with the well-tailored suits and the former Duke player. This is a two-team tie between Villanova and Marquette, sharing the Big East regular season crown. However, for purposes of the Big East Tournament, we need to figure out who gets the one seed.

In a two-team tie, we’re supposed to first look at regular season head-to-head results, but Villanova and Marquette split their series this season. The next step to try and break the tie is to look at look team’s record vs. the team or group of tied teams occupying the highest position in the standings. Since they’re already at the top (and are split), the guide book instructs us to do this:

If an advantage is not determined, proceed to the next team or group of tied teams in the standings for comparison. Continue down through the standings until one team gains an advantage.

2019 Big East Tie-Breaking Procedures

Okay, we have a three way tie for third (which we’ll have to untangle later!). Let’s look at the cumulative records of Villanova and Marquette against Georgetown, Xavier, and Creighton. Villanova went 4-2 against these teams, dropping games back-to-back against Georgetown (woo!) and Xavier. Marquette (using the simulation results for the game against Georgetown in these standings) went 5-1, dropping a game against Creighton just this Sunday. Marquette wins the first place seed in the Big East Tournament and Villanova gets the second seed!

Georgetown, Xavier, and Creighton, Tied for Third Place

Now that we have a multiple team tie for third, things are getting interesting. We have to consider these three teams as a “mini-conference” according to the rules when we’re comparing the head-to-head results. From the rule book:

Teams are viewed as a “mini-conference” when comparing head-to-head results. The team, or teams, with the best record vs. the other teams in the mini-conference gains the advantage and the team, or teams, with the worst record vs. the other teams in the mini-conference is seeded the lowest.

2019 Big East Tie-Breaking Procedures

The first order of business is to see how each team stacks up in this “mini-conference”. Georgetown went 1-3 in this conference, dropping both of their games against Creighton and winning one against Xavier.

Creighton went 3-1 in this mini-conference, dropping one to Xavier. Xavier went 2-2 in this mini-conference, splitting their series with Georgetown and Creighton. As a result of the standings in this mini-conference, Creighton gets the third seed in the BET, Xavier gets the fourth seed, and the Hoyas get the fifth.

Seton Hall, Providence, and Butler, Tied for Seventh Place

Based off of the simulation of the rest of the season, we have another three-way tie for seventh, between Seton Hall, Providence, and Butler. Time to break some more ties! We have to first use the same procedure we used to determine seeds three through five, where we make a “mini-conference” and look at the head-to-head record within.

In this “mini-conference”, Seton Hall has a record of 2-2, splitting their series with Providence and Butler. Providence has a record of 3-1, splitting the series with Seton Hall and sweeping Butler. Butler has a record of 1-3 in this mini-conference, splitting the series with Seton Hall and getting swept by Providence.

This gives the seven seed to Providence, the eight seed to Seton Hall, and the nine seed to Butler.

Now that that’s all figured out…

The seeds for the tournament will look like this below;

And this is how the bracket would shake out;

To quote my guy Billy Mays, “But wait! There’s more!”

Remember that game between DePaul and Georgetown in Chicago where the algorithms’ predictions for the score ended up in a virtual tie? While I gave the game to Georgetown for the sake of carrying on with this post (and it’s a homer pick), what if (shudders) DePaul won that game?

The Results if DePaul Won Wednesday Instead of Georgetown

I’ve adjusted the final standings below, adjusting the final records and records rest of the way if DePaul were to win the game on Wednesday instead of Georgetown.

Note the records in parentheses are the records as we simulated for the last games of the season, and have been factored into the “final” Big East conference record. This is the simulated season, with the Georgetown-DePaul game winner as DePaul in this bizarro world timeline.

Okay, now this is just comical. We now have all ten teams in the conference (if they win the games the algorithms predict they will, and DePaul wins at home against Georgetown) tied for four different spots. If you thought Big East Tournament seeding was a mess before, this is on a whole ‘nother level. Let’s work our way through this mess!

Villanova and Marquette, Tied for First Place

Here, we have to do the same thing as we did before. These teams split the head to head, so now we have to look at how they performed against the two teams tied for third.

Uh, wait, Villanova and Marquette have the same record against these teams. They both went 3-1, with Villanova losing a game to Xavier and Marquette losing a game against Creighton. The rule book tells us to continue down the standings until we find an advantage between the two teams.

This means that now we have to go to the teams tied for fifth to determine who wins first place.

Against St. John’s and Georgetown, Villanova split their series for a record of 2-2. Against the same teams, Marquette also went 2-2, sweeping Georgetown and getting swept by St John’s. Wait, what?

The rulebook tells us to keep going down until we determine an advantage, so now it’s up to finding out Villanova and Marquette’s records against the bottom 4 teams in the conference to determine the top seed in the conference. Yep, you read that right.

Against the mini-conference of Seton Hall, Providence, Butler, and DePaul, Villanova went 8-0, sweeping the series with all the teams! Oh, and Marquette did the same. So this tie still isn’t resolved. To the next level of tiebreaking!

The next step to break the tie is to compare the overall conference road record of the teams. In this category, Villanova goes 5-4, dropping games on the road to Marquette, Xavier, St John’s, and Georgetown (woo!). Marquette 7-2, dropping games on the road to Villanova and St John’s. Therefore, Marquette wins the first place seed and Villanova gets the second place seed! Onto the next tie…

Xavier and Creighton, Tied for Third Place

First, we have to look at the head-to-head record. They split their series, so now we have to look at how they performed against the top team in the conference, which we just found out was Marquette.

Xavier got swept by Marquette, while Creighton split their series with Marquette by stealing a road win just this Sunday. Therefore, Creighton gets the third place seed and Xavier gets fourth.

St John’s and Georgetown, Tied for Fifth Place

Looking at the head-to-head record, the teams split their series, the last game a thrilling Georgetown victory at Madison Square Garden. Looking at how they went against the top team in the conference, which we determined to be Marquette, St John’s swept the Golden Eagles, while the Golden Eagles in turn swept the Hoyas.

St John’s takes the fifth place seed, while Georgetown takes the sixth place seed.

Seton Hall, Providence, Butler, and DePaul, Tied for Seventh Place

Just one more tie to break! We have to determine the records of each team within this “mini-conference”.

Seton Hall went 2-4, splitting their series with Providence and Butler, and getting swept by DePaul.

Providence went 4-2, splitting their series with DePaul and Seton Hall, and sweeping Butler.

Butler went 3-3, splitting their series with Seton Hall, getting swept by Butler, and sweeping DePaul.

DePaul went 3-3, getting swept by Butler, splitting their series with Providence, and sweeping Seton Hall.

The final standings and record (within the mini-conference) of this mini-conference is as follows.

1. Providence, 4-2
2. Butler, 3-3
3. DePaul, 3-3
4. Seton Hall 2-4

According to the rulebook, Providence gets the seven seed for winning the mini-conference, while Seton Hall ends up with the ten seed for having the worst record in the mini-conference. To determine who gets the eight and nine seeds, we have to look at the head-to-head record between Butler and DePaul. Since Butler swept DePaul, Butler gets the eight seed, and DePaul gets the nine seed.

Now that this scenario where DePaul beats Georgetown in Chicago is figured out (with the other simulated results holding firm), here’s the final seeding;

And here’s what the bracket would look like with that seeding;

Keep in mind, that this was all changed as the result of the Georgetown-DePaul game result being flipped. None of the other simulated games were changed. If this isn’t proof that every game matters, I don’t know what to tell you. Because of that one game, Georgetown would get a much tougher side of the bracket.

IN CONCLUSION…

If you’re a fan of anarchy, you should be a fan of Big East basketball. While it’s very unlikely that the winners of the remaining nine games will all be what the algorithm predicted them to be, it’s an interesting example of just how close this conference is, and how significant these last few games in the last week of the season are.

So much of the Big East Tournament will be decided this week, as matchups and seeding will come down to the wire in these games. As the Hoyas look to get off of the bubble and in the tournament, the remaining slate of conference play looms large.