Dallas Mavericks Machine Learning Scout

Embed from Getty Images

There is no easy way to put it. Dallas stink right now.

Sitting at the bottom of the league with a 1-10 record. It's very un-maverick like to see them struggling so badly. Even in the past few season with a so-so roster, Dallas has extracted everything they possibly could from their teams and have always been in the hunt for the 8th seed.

Dallas prides itself on not bottoming out and putting together a decent product for their fans.

So what gives this year?

Well they are managing their cap space diligently and currently have the 3rd lowest pay roll in the league. It's pretty simple in the NBA, if you don't pay for talent in this league you are going to stink.

Dallas will have about $42m in cap space next season and about $73m in 2019/2020 when Wes Matthews comes off the books. Dallas have ZERO guaranteed contracts for 2020/2021.

We could theorize what Dallas might do over the next few seasons, but it's pretty clear they are looking for a reset and maintaining cap flexibility to go after some big fish over the next few seasons. Pairing them with Dennis Smith Jnr and whatever top 10 pick that comes their way at the next draft.

As I said earlier, Dallas is not a traditional tanking team. Rick Carlisle isn't made that way.

So I thought I would dive into my automatic scouting algorithm (You can read about it here: http://www.zigzaganalytics.com/home/automatic-nba-scouting-with-machine-learning) and see how we would can help Dallas out.

In short, the algorithm looks at over 50 statistics and spits out a ranking of what it sees as the biggest reasons for winning and losing.

With any algorithm, it loves data and lots of it. We have only played 11 games this season, so the accuracy isn't as good as after 82 games with a bigger sample size. However it has certainly found some trends.

As I said in my original automatic scouting post, the algorithm will rank the biggest reasons for winning and losing. Pointing the coaching staff and scouts in the right direction, for them to then dig into the detail further.

Well in this case, I have done just that. Digging into the automatic Dallas scout a little further and matching the findings with advanced stats and video.

Below, I have put together the findings of the automatic scout and my deeper analysis on an easy to read 1 page report.

The algorithm found the best biggest reason for the Mavericks struggles has been their mid-range shooting game, and the shots they are getting through isolation play rather than ball movement.

Looking through film, the first 3 minutes of the Mavericks last game vs the Timberwolves is a microcosm of their mid-range woes this season and provides some evidence for what the algorithm is seeing.