Nylon Calculus: AAU basketball state of the union

HYATTSVILLE - JULY 11:Old Dominion University basketball coach Jeff Jones, left, talks with Team Takeover Director and Camp Manager Keith Stevens at the Ty Lawson and Victor Oladipo All American Camp held at DeMatha High School on July 11, 2014. (Photo by Toni L. Sandys/The Washington Post via Getty Images)
HYATTSVILLE - JULY 11:Old Dominion University basketball coach Jeff Jones, left, talks with Team Takeover Director and Camp Manager Keith Stevens at the Ty Lawson and Victor Oladipo All American Camp held at DeMatha High School on July 11, 2014. (Photo by Toni L. Sandys/The Washington Post via Getty Images) /
facebooktwitterreddit

Premier AAU basketball has been around for years, with the top players in high school uniting for showdown games that rival college game attendance at this point. In fact, LeBron could not even get into the game between LaMelo Ball and Zion Williamson last year in Las Vegas. As we work on trying to project NBA prospects earlier and earlier we run into the problem of increasingly unreliable data samples. AAU stats have been largely un-mined to this point, for a number of reasons, but there are some useful insights to be found.

Collin Sexton, Trae Young, and Michael Porter Jr. are the top-three scoring leaders from the 2016 Nike EYBL season.  With the draft fast approaching, all three players rightfully will be part of the lottery pick discussion. The use of AAU data can start to fill in the gaps and understand how players develop, what can or maybe more importantly can’t be developed, and provide more history for players. There are only so many college games in a season, providing a limited sample size.

Of course, when it comes to the data, nothing is perfect. Adidas Uprising and Under Armour Association (UAA) have box scores and play-by-play, while the EYBL only has box scores. Much like college play-by-play data, substitutions contain errors but we have seen concerted efforts to improve their data, particularly from UAA. At this point, EYBL is the worst now, not even charting minutes for this year which has real implications when predicting future performance. A structured system with data being used properly would create a tangible value for players and the programs.

So what can we do with the AAU data we have now?

This graph simply highlights Nassir Little’s rank in each of the basic box score statistics, relative to the rest of the players on the Adidas Circuit. Looking at more advanced stats is easily the next small step, especially with the play-by-play data provided by Adidas. However, starting the conversation here and interpreting this is important.

Little dropped 28 points to lead the McDonald’s All-American Game this year. Interestingly, according to the graph above, Little is good at several things but not great at anything. The higher percentile on field goal makes than attempts may have alluded to his potential to explode for a big night under the golden arches.
Little dropped 28 points to lead the McDonald’s All-American Game this year. Interestingly, according to the graph above, Little is good at several things but not great at anything. The higher percentile on field goal makes than attempts may have alluded to his potential to explode for a big night under the golden arches. /

Above is an assist chart from one of Game Elite 2019’s games. Ashton Hagans led the Adidas Circuit last summer in assists. The addition of play-by-play data allows us to start utilizing more advanced statistics and creating network analysis to see how players across a team assist their teammates. The use from an advanced scouting perspective is letting coaches know if the passing lanes are blocked with Hagan’s as the ball-handler, especially the gap for a pass to Richey, the most frequent connection. This kind of basic strategy can be implemented more into AAU game planning. Of course, the small number of games matched with quite a bit of player movement over the summer requires a very careful use of analyzing the data.

The major step is finding if there is a level of play differential across all three leagues. This matters greatly in projections and the ability to compare players across leagues. The first step may be as simple as projecting each league to the next level and finding if there are different weights and correlations between each league. If so, player rankings can start to be formed but EYBL charting minutes and potentially even play-by-play would be a major victory to unlock more of the advanced statistics that are better predictors.

The ultimate reason we need to be exploring AAU statistics is Darius Bazley and the many who may follow. Bazley committed to Syracuse but opted instead to head to the G-League. As more high school players head straight to professional basketball AAU data will become increasingly important.

Once more historical data is accumulated, we can start to calculate what transfers at the next level be it the G-League or college basketball, the same way draft models are built. For example, free throw shooting in college can generally be a better indicator of 3-point shooting in the NBA than college 3-point shooting. Darius Bazley struggled behind the arc for the Nike Mean Streets shooting only 28.1 percent. However, he shoots a fairly decent 76.1 percent from the line. That being said, other attributes such as rebounding may or may not be as predictive because the competition is smaller and not as strong especially for a long but wiry frame like Bazley.

Next: Rankings the best NBA nicknames of all time

The onus is on us to have conversations and properly educate coaches whether it be AAU, high school, college, or the NBA. AAU season is almost in full swing with an opportunity to continue the analytics conversation and properly educate programs. Analytics is a tool, not the answer, after all.