Is The NBA Summer League A Good Indicator? The Answer May Surprise.

The NBA’s Las Vegas Summer League is one of the first opportunities for young players and rookies alike to show off their skills and prove they belong at the next level. 

We NBA fans and analysts like to use the Summer League as a sort of barometer to test how well certain hyped players will respond to the increase in competition. Will a young player suddenly fold under the pressure or will they rise to the occasion and perform as expected?

This five game mini-series is supposed to be an indication of the young players’ future performance, but does it actually do its job?

A lot of people say no, the Summer League cannot give any sort of indication because all we see is rookies and other skilled but young players playing against other inexperienced players and castoffs. Those people claim that the lack of competition in turn produces a lack of predictive ability.

But at the same time, no matter what level you play at, success is success. You can only win by so much and produce at so high a level. So some argue that winners are winners at any given level.

To tackle this difficult question, I went back and looked at the statistics for the 2007, 2008, and 2009 NBA Summer Leagues. For each season, I singled out the fifty players who averaged the most points per game and then eliminated those who did not play at all in the regular season.

Next, I added up the combined points, rebounds, assists, steals, and assists (which I will call Summer League Total or SLT) for each player per game because those are the various ways that each player can contribute to their team. After that I did the same thing for each and every one of those players in the following season (I called this RST). Some players did appear on the list twice because they played in multiple Summer Leagues.

Once I compiled all of those numbers, I put them on a scatterplot to see if there was any correlation between RST and SLT. I expected to see a positive trend so that as a player performed better in the Summer League they would also have a better regular season. 

Sadly, I was wrong. There was a slight positive correlation but it was quite weak as the players who barely played in the regular season threw the data off tremendously.

I wasn’t satisfied with that answer, so to correct the data, I prorated the data to form RST per 48 minutes (RST/48) and SLT per 48 minutes (SLT/48). This was done by dividing each players SLT and RST by their minutes per game in each respective season and then multiplying the result by 48.

The results of graphing SLT/48 on the x-axis and RST/48 on the y-axis shows a much stronger correlation that the previous attempt did. 

The graph quite clearly shows that as a player’s SLT/48 increases, so too does his RST/48.

Also, while looking at the data I noticed that all but two of the 112 players had a higher SLT/48 than RST/48. The two exceptions were Detroit’s Will Bynum in 2008-2009 (32.96 SLT/48 and 40.51 RST/48) and Oklahoma City’s Russell Westbrook in 2009-2010 (40 SLT/48 and 42.96 RST/48).

To emphasize this, I graphed SLT/48 as the independent variable and the difference between SLT/48 and RST/48  as the dependent variable. The graph showed a random clump of numbers with all but two of the data points above the x-axis.

So what does all this mean?

Based on the number crunching, the Summer League is indeed indicative. If a player has success in the preseason, he will also have success in the regular season, at least in a per 48 minutes way.

Also, the Summer League provides a ceiling for a player’s performance that year. Whatever the SLT/48 is for a player that season, it is highly unlikely that his RST/48 that season will be higher.

For example, John Wall this season had 23.5 points, 4 rebounds, 7.8 assists, 2.5 steals, and 0.5 blocks in his 32.3 minutes per game this 2010 Summer League. These statistics would give him a SLT/48 of 56.92, which incidentally would have been the third highest such number in the analyzed data—Anthony Randolph in 2009 and Kevin Love in 2008 had better Summer Leagues.

What this says about Wall according to my theory is that he will without a doubt be successful this season. If he plays 35 minutes per game, the sum of his points, assists, rebounds, steals, and blocks would most likely be less than 41. But hey, I’d take 25 points, 10 assists, three rebounds, two steals, and a block from my rookie point guard any day of the week.

So in conclusion, it seems like the Summer League is a good indicator for the regular season after all.

Do you agree?

Note: If you'd like to see the data and graphs, message me your email address and I'll send it along.

You can see more of my work at www.sportshaze.com, where I serve as the Atlanta Hawks analyst.