The Dallas Mavericks @ AAC

With Preseason basketball and training camps underway, I can’t help but reminisce about the Mavs’ efforts last season — before a new season begins and everyone forgets. Whether you’re a non-fan, avid fan, or an ‘MFFL’ (‘Mavs Fan For Life’), everyone would agree on one thing: the team outperformed expectations in ‘20–’21.

Here are the facts:
1a. The Mavs ended the regular season as a 4th seed, in the heavily talented Western Conference.
1b. We were barely edged out by the Warriors for 3rd seed.
2. Luka Doncic, bonafide All-Star, was dealing with a minor injury the final game of the season, causing him to miss a couple games of the 1st round.
3. 1st round: Mavs defeated the Utah Jazz in 6 games.
4. 2nd round: Mavs defeated the Phoenix Suns in 7 games.
*Suns were last year’s Western Conference Champions
5. Conference Finals: Mavs fall to the Golden State Warriors in 5 games.
*Warriors went on to win the NBA finals
6. Dallas Mavericks finished the season as a Top 4 team!!

How did a team like the Mavericks do so well? And is it possible to keep this up? Well, there are a million variables shuffling and affecting a team’s overall record: player’s talent, height, age, skill level, experience, number of other teams played for, midseason trades, coaching and management, draft assets, capital, state taxing, and the list goes on and on.
But one stat drew my attention and actually bothered me: attendance. Attendance affects sales, but more importantly, crowd noise, which affects player performance, which affects wins and losses. In years past, American Airlines Center (the Dallas Mavs’ home, also known as AAC) attendance was spotty and embarrassing to witness on TV. Attendance could surely be better, bringing up the question:
Do the Mavs really have home-court advantage at AAC?

Let’s dive into some data using SQL.

I began by gathering data from all home games at AAC, along with the stats I thought to be significant.

Columns: date, opp, attendance, w_or_l, pts, pt_diff, reb, ast, stl, blk, fg_per, three_per, fta, ft_per, conf_opp, div_opp

Then, I aggregated the data to find averages and counts of the columns.

Noteworthy stats: the Average Attendance was less than 20,000, and Home Record was 29–12

Overall, the stats were not ugly (perhaps beautiful because the total games played at home = 41, the number of Mavs legend Dirk Nowitzki).
Our win percentage was a hair above .700, and the average points scored were over 100 with an average point differential of +7.
But why 12 losses?

To filter by losses, I added a simple WHERE clause to the above query:

As expected, the data revealed every single statistical category to be lower than average. However, the Average Attendance changed by less than 100 people. This was odd. So I dug deeper.

I decided in order to best capture all the data without a GROUP BY clause, I would need to use a Window function. More specifically, I wanted to rank the attendance and also score the attendance in either the top 50% or bottom 50%.

DESC is used here to view higher attendance first

Top 50% is when NTILE = 1, Bottom 50% is when NTILE = 2

The pattern became obvious. More W’s were found at the top half of the table because attendance was higher in those games. In the Top 50% Attendance games, the Mavericks surrendered 4 of 12 losses, and in the Bottom 50% Attendance games, the Mavericks surrendered 8 of 12 losses. A great benchmark number for Attendance is 20000, as shown below.

Filtered data in subquery so that data with attendance > 20000 displays only

Now, I admit: I did disregard basketball chemistry, new players, traded players, play styles, defensive schemes, real basketball stats, and other strategies behind the game. Yes. But I defer to our coach Jason Kidd and staff for that. But from this fan’s angle, yes, the Mavericks do indeed have home-court advantage. Here is my proposition:

Let’s pack AAC and reach 20,000 fans! Let’s get even MORE “Rowdy, Proud, and Loud!”