With the Big East Tournament kicking off today in Madison Square Garden, it’s a great time to look back at how the end of season predictions we generated using an algorithm for the last nine games of the Big East conference slate fared, and then to again use our algorithm to predict the outcomes of the Big East Tournament. Let’s get started!
Below is a table of the nine games that were remaining in the Big East’s regular season at the time of when we published that first piece and the results of the games as simulated through our algorithm next to the actual results.
| Date and Time | Road Team | Final Score (as simulated) | Actual Score |
|---|---|---|---|
| 3/5/19 at 6 p.m. | Xavier @ Butler | 70-66, Butler | 71-66, Butler |
| 3/6/19 at 6:30 p.m. | Marquette @ Seton Hall | 74-71, Marquette | 73-64, Seton Hall |
| 3/6/19 at 8 p.m. | Providence @ Creighton | 74-67, Creighton | 76-70, Creighton |
| 3/6/19 at 8:30 p.m. | Georgetown @DePaul | 80-80 (wait, what) | 101-69, DePaul |
| 3/9/19 at 12 p.m. | Villanova @ Seton Hall | 70-68, Villanova | 79-75, Seton Hall |
| 3/9/19 at 12 p.m. | Butler @ Providence | 67-66, Providence | 83-70, Providence |
| 3/9/19 at 2:30 p.m. | Georgetown @ Marquette | 83-73, Marquette | 86-84, Georgetown |
| 3/9/19 at 5 p.m. | St. John's @ Xavier | 74-72, Xavier | 81-68, Xavier |
| 3/9/19 at 8 p.m. | DePaul @ Creighton | 82-73, Creighton | 91-78, Creighton |
As we can see, the algorithm correctly predicted the final outcomes (the winners) of five games; those matchups were Xavier @ Butler, Providence @ Creighton, Butler @ Providence, St John’s @ Xavier, and DePaul @ Creighton. It got the prediction of who would win wrong three times; those games were Marquette @ Seton Hall, Villanova @ Seton Hall, and Georgetown @ Marquette. For the Georgetown @ DePaul game (do we have to talk about this?), the score generated was a tie at 80-80, so we can’t really judge it either way. I got too cocky and believed we’d win that game, but instead I was traumatized by that guy on the bench (a walk-on!) doing jumping jacks on the court after being subbed in for the last minute.
One important thing to keep in mind about predicting these basketball games is that Ken Pomeroy himself says that if the final predicted score has a margin of eight points of less between the teams, it’s pretty unpredictable since things can swing either way.
I also find it pretty amusing that the only games that the algorithm got the winner of wrong was games that Seton Hall or Georgetown played in. Marquette’s also in that list, but it’s worth pointing out considering the looming Big East Tournament matchup between the Pirates and the Hoyas.
Overall, the algorithm went 5-3-1, which is fairly respectable. If you bet on the winners of the games, you’d win more money than you’d lose, which seems pretty successful to me. The scores were mostly off (our Xavier @ Butler prediction was pretty much as good as it got- we peaked early), but as for winners, it did a decent job of predicting who would win.
Now, let’s look at the Big East Tournament and predict the matchups! We already have the initial teams for the Wednesday games and the Thursday games.
If you’re interested, I explain in this article how the algorithm is set up. Long story short, I use KenPom for advanced statistics, and then put that into an algorithm. For this go-around, I removed the effects of home court advantage (since it’ll be a neutral site). Below is the table of the simulated games (using the simulated winners to advance the bracket/tournament simulation)
| Date and Time | Teams | Final Score (as simulated) |
|---|---|---|
| 3/13/19 at 7 p.m. | Butler vs. Providence | 68-67, Butler |
| 3/13/19 at 9:30 p.m. | DePaul vs St John's | 81-78, St John's |
| 3/14/19 at 12 p.m. | Butler vs. Villanova | 70-66, Villanova |
| 3/14/19 at 2:30 p.m. | Creighton vs. Xavier | 75-73, Creighton |
| 3/14/19 at 7 p.m. | St John's vs. Marquette | 79-73, Marquette |
| 3/14/19 at 9:30 p.m. | Georgetown vs. Seton Hall | 79-76, Seton Hall |
| 3/15/19 at 6:30 p.m. | Villanova vs. Creighton | 73-70, Villanova |
| 3/15/19 at 9 p.m. | Marquette vs. Seton Hall | 76-71, Marquette |
| 3/16/19 at 6:30 p.m. | Villanova vs. Marquette | 71-70, Villanova |
And here it is in bracket form, to make it more digestible. When a team’s in italics, it means that was the simulated winner of their respective game before the new one.
Stop me if you’ve heard this before: Villanova wins the Big East Tournament. Our beloved Hoyas are defeated in their first game in the BET by Myles Powell and company.
I don’t want to lose to these people.
So let’s simulate the rest of the BET, and see how the Hoyas stack up against each team they would face (with all other simulated results holding the same) win or lose. For example, even if the Hoyas lose against Seton Hall, for purposes of this exercise, I’d advance them in the bracket and see how they would face up against the next opponent they faced, which according to the simulation would be Marquette.
| Date and Time | Teams | Final Score (as simulated) |
|---|---|---|
| 3/13/19 at 7 p.m. | Butler vs. Providence | 68-67, Butler |
| 3/13/19 at 9:30 p.m. | DePaul vs St John's | 81-78, St John's |
| 3/14/19 at 12 p.m. | Butler vs. Villanova | 70-66, Villanova |
| 3/14/19 at 2:30 p.m. | Creighton vs. Xavier | 75-73, Creighton |
| 3/14/19 at 7 p.m. | St John's vs. Marquette | 79-73, Marquette |
| 3/14/19 at 9:30 p.m. | Georgetown vs. Seton Hall | 79-76, Seton Hall |
| 3/15/19 at 6:30 p.m. | Villanova vs. Creighton | 73-70, Villanova |
| 3/15/19 at 9 p.m. | Marquette vs. Georgetown | 82-75, Marquette |
| 3/16/19 at 6:30 p.m. | Villanova vs. Georgetown | 78-71, Villanova |
As you can see, it’s a very tough road to win the Big East Tournament for the Hoyas, as they would (according to the simulation) face the top three seeds in the Big East, escalating from third seed to the first seed over the course of three days. To add to that, they’re underdogs in all of their games.
Of course, the Big East Tournament is going to be wildly unpredictable, with nine teams out of ten (sorry DePaul) having some potential to make some noise. Who knows what will happen in the Garden over the next four days? I feel pretty confident in saying that the tournament will not happen as the predictions above say it will- there’s fatigue, some players have bad nights, and for a more scientific reason why, most of the games simulated are decided by three points or less. As Ken Pomeroy himself says, the projections made using this formula can really swing either way when a game’s margin is less than eight points.
Here’s to a crazy next few nights in the Garden.
