Are the Mavs on Track for Their Best Season Since the Title Year?

Projecting the win total from a one-third season sample

Photo by Sam Hodde / Getty Images

Besides the awesome community that’s building here on digginbasketball, probably my second favorite thing about running my own website is how a simple tweet can evolve into a full-on data deep dive—and having the platform to share those results with you all.

That tweet in question was:

It sparked some discussion—and got me thinking about how this season's Mavericks stack up against their past teams.

Since the offseason, I’ve maintained that this is the best, most talented, and deepest Mavericks roster in the Dončić era. Barring significant injuries, there should be no excuses for this team not to surpass its best win total since Luka was drafted. That mark, set in the 2021-22 season, is 52 wins. Even when adjusted for the shortened seasons in 2019-20 and 2020-21, it’s still the benchmark—and it feels very achievable this year. For context, that 52-win season, led by Jason Kidd in his first year as head coach, surprised everyone by culminating in a trip to the Western Conference Finals.

But 52 wins isn’t an insurmountable bar, especially given the Mavericks’ current trajectory. They’re on a 65.4 percent win pace, translating to 54 wins. According to Cleaning the Glass, which factors in point differential (Points scored per 100 possessions minus points allowed per 100 possessions), the Mavericks are projected for an even more impressive 58.4 wins. Early clutch collapses have left the Mavericks' actual win total lagging behind their expected win total since the start of the season.

The Cleaning the Glass projection, combined with the tweet, inspired me to crunch some numbers of my own, looking to the past to better understand where this team might end up by season’s end. We’re now at the 26-game mark—or one-third of the season—which feels like a sample large enough to have some fun with the data.

Building a simple model for win total prediction

Here's the process I used to build a simple win projection model of my own:

• I parsed Cleaning the Glass data from the 2003-04 season onward, which includes wins, losses, win percentage, offensive and defensive ratings, and rankings for all NBA teams.

• Then, I adjusted the win totals to an 82-game scale to account for shortened seasons due to the lockout (2011-12 season) and COVID-19 (2019-20 and 2020-21 seasons).

• I calculated offensive and defensive ratings relative to the league average for specific years, to measure how much a team stood out on both sides of the ball compared to the competition.

The chart below plots all teams from every season between 2003-04 and last season, with Point Differential on the x-axis and Wins (adjusted to an 82-game scale) on the y-axis.

How to interpret this chart?

Each blue dot represents one team in a particular season (for example, you can see the best regular-season Mavericks team, the 06-07 Dallas squad, in the upper right corner). You’ll also notice a strong correlation (the red line) between point differential and wins. Teams above the red line are those that exceeded expectations based on point differential, often excelling in the clutch. Conversely, teams below the line typically underperformed in the clutch and won fewer games than expected based on their point differential.

Simple model