NBA Summer League is pretty early to call “bust” isn’t it? Yeah it sounds crazy and it might be rash, but hang in there with me.
What if I told you there is enough of a correlation between poor summer league performances and a top 10 draft pick being a bust down the road? As crazy as it sounds and yes with a small-ish sample size of summer league play, there is what seems to be a good indication that a call of “draft bust” can be made.
It would take a ballsy front office to make a call, as they have no doubt invested hundreds of hours into scouting. Not only convincing themselves but also the team owner that this is the right player for their franchise. But you know what? Scouting is never a guarantee of hitting a home run with a draft pick, there are far more busts in the top 10 than all-stars.
So why not save face early? Cut your losses and trade that top 10 pick before the rest of the league catches on, and the trade value of the player plummets.
Why not try and get back a future top 10 pick from another team? And try and draft an all-star calibre player again next season?
The NBA’s collective bargaining agreement allows teams to do it. There is a “30 day rule” where in a team that has signed a draft pick can trade that player 30 days post signing their rookie contract. A player can therefore have a poor summer league and can be traded before the regular season begins.
So let’s take a look at my research and what conclusions we can draw from it.
To make the correlation between NBA summer league play and how a draft pick will project, I have used “PER” (Created by John Hollinger, formally of ESPN and now with the Grizzlies and which stands for Player Efficiency Rating) as the metric to base my research on.
I can hear the groans of the hard core basketball nerds out there already.
Yes I have used PER, which is a player rating metric that is about 10 years old now. However, it still does a good-enough job of rolling together what a player does on the court, spitting out one number to rate a player, that we can then use for comparison. It has its flaws, but all metrics do and PER is still widely used by ESPN, TV broadcasts and many statistical websites, so the casual NBA fan may have an awareness of it.
The data I therefore have is both the PER of the draft picks initial year at NBA summer league and the PER of each season a top 10 draft pick has played in the NBA, going back to the 2009 draft.
If a player did not play enough minutes at NBA summer league or in an NBA season to qualify, then I have removed their PER for that summer league or NBA season.
For example: Julius Randle of the Lakers played 14 minutes total in his first NBA game before suffering a season ending injury. His PER for that season does not count.
The PER of a player in general goes up each year they play in the NBA, as they progress from a rookie to a veteran. Peaking somewhere around the 6th-9th season they play in the league.
For my research I have taken the average PER of a player from their first 3 seasons playing in the NBA. This has allowed me to have a greater data set to compare players against, and it still allows for a clear call in regards to a player being a bust or all-star calibre player.
Let’s take a look at the below which is a linear regression (for math nerds) that shows the correlation between summer league play and a player's average PER after 3 years. Again for the math nerds, the correlation coefficient is 0.62 which means it is statistically relevant and there is a meaningful correlation. For everyone else, it means there is some substance behind the numbers.
To help read the below, the blue spots are individual players and the black line is what is known as the regression line. The regression line helps us compare a summer league PER vs the 3 year PER of a NBA player.
eg. A summer league PER of 10 correlates to a 3 year PER average of just under 12
I have highlighted the “bust zone” in red above. If a player falls within this zone they are highly likely to be a bust of a top 10 pick and become barely a rotational player.
The “bust zone” tells us that in nearly all cases a player with a summer league PER of 11 or less projects to have a 3 year PER average of less than 12.
OK, so what do these PER numbers actually mean?
The original breakdown of what a PER value indicates being:
PER > 22.5 – All-Star level player
PER 15 - Average NBA player
PER < 10 – Player won’t stick around in the league long
A player's 3 year PER in nearly all cases is not going to see them with an average of greater than 22.5 (All-Star) as players are rarely all-stars straight away. Their average PER will build up over time as they progress as a player in the league.
I came to the conclusion that a 3 year PER of 12 was the cut-off point for the bust zone, by looking through historic 3 year PER data of all players in the league over the past 8 years.
When a top 10 draft pick has had a PER average of 12 or less after 3 years they became, at best, a rotational player for a team throughout their careers.
Whereas a 3 year PER of 17 or greater will see that player in nearly all cases become an all-star level player at some point in their career.
So I have established the “bust zone”, where a summer league PER of 11 or less projects an NBA player to have a 3 year PER of 12 or less and become a rotational player at best.
When a team drafts a player in the top 10 they are certainly looking for someone who is going to become more than just a rotational player.
How has the “bust zone” projection performed?
Below is a list of top 10 picks with 3 years or more NBA experience that had a summer league PER of 11 or less since the 2009 draft.
You can see 6 of the 8 players fall into the bust zone, 2 players (Dion Waiters & Thomas Robinson) fall just outside of the bust zone with a 3 year PER average just north of 12, but are themselves rotational players at best.
See any All-Star calibre players here? Any potential 3rd best players on a championship team? I’m not.
Sure, Otto Porter is having a nice year thus far. He has moved himself out of the conversation for being a bust, but is he anywhere near an all-star level player?
Dante Exum didn’t play due to injury in his second season, so we can cut him some slack. But are we really seeing an all-star calibre player?
For what a front office would hope for out of a top 10 pick, the “bust zone” has been highly accurate in its projection since the 2009 draft.
Let’s now take a look at players with less than 3 years NBA experience, drafted in the Top 10 and who had a PER of 11 or less at NBA Summer League:
As it stands today, Emmanuel Mudiay and D’Angelo Russell are into their 2nd NBA Seasons. This season Russell has a PER of 16.6 and is tracking nicely after an up and down rookie season, far from a bust but still a long way from being an All-Star calibre player.
Mudiay has serious question marks. After 1.5 seasons he is tracking to have a 3 year PER of under 12 which would certainly see him in the bust zone.
From this year’s draft class we have Phoenix Suns' rookies Marquese Chriss and Dragen Bender. Chriss is so far posting a PER of 9.9 (not even average for a rookie) and Bender a PER of 6.2 which would result in being the 4th worst PER for a top 10 pick since 2009.
The bust zone projection of these 4 players is tracking pretty accurately so far, with D’Angelo Russell the only player who might buck the trend and become something.
To sum it all up, the bust zone projection certainly seems to have merit. Since 2009 no player who has had a poor summer league is shaping up to be an all-star calibre player, the vast majority are bench players and projecting to have careers no better than a reasonable player in the league.
So it begs the question, why not trade a top 10 pick after a poor summer league showing? If you can get a future top 10 pick back in return or perhaps even a top 10 pick from the same draft?
It would take some guts, but cutting losses early and getting equal value back is a pretty good outcome.