Introduction

In the previous analysis, I designed a classification model to predict international players that would make it to the NBA. In this part of my analysis, I take a more in-depth look to go beyond just predicting NBA success and look at what attributes influence NBA success in international prospects.

There exists a wide variety in the type of international prospects that make it to the NBA. I do a case study of three common type of prospects

Dario Saric

The first type of international prospect that makes it to the NBA are “draft and stash”. These are commonly players who are both players who are young and have been productive in their respective league. Dario Saric is a great example of this. He was initially drafted with the 12th pick in the 2014 NBA draft but did not come to the NBA until 2016. Saric was both a young prospect that flashed great expectation as well as a solid contributor to his team.

One of the biggest question marks about Saric entering the NBA was his outside shooting. The shot chart below reveals the makes and misses for Saric for the entire season. As you can see, Saric was most effective inside the paint. He took the majority of his shot in or around the restricted area and showed impressive ability to get his shot off (evident in the small amount of shots blocked). While he most most effective in the paint, that didn’t stop him from taking threes. While the results weren’t overally impressive, Saric shot chart reveals the multi-dimensionality of his game.

ggplot(dario, aes(x= coord_x, y = coord_y))+
  annotation_custom(court, -745, 745, -110, 1175)+
  geom_point(aes(color= make, alpha = .8), size = 3)+
  xlim(-850, 850) +
  ylim(-150, 1000)

This heat chart gives a broader look into where Saric likes to shoot. Again, most of his shots came in the restricted area. In terms of threes, they came mostly from the right corner and the left portion of the top of the arc

Concerns about Saric outside jump shot seem fair. He was not overally efficient from three; but at the same time he did shy away from taking them. The graph below illustrates the number of three pointers he attempted and made game by game. You can see that Saric had a “shooters mentality”; he took at least one three every game and in some games he took six or seven. But he was very streaky. There were a lot of games where he didn’t make any threes, and some where he was perfect from three.

ggplot(dario_gbg, aes(x=game, y = count, color = event_desc_id)) + geom_path() + labs(color = "")

The other part of Saric’s game that was highly attractive to scouts was his passing ability. He was projected as being a modern day “point forward” in the NBA. The graphs below are the shot charts for baskets off of Saric assists. Most of his passes lead to baskets at the rim, or corner threes, demonstrating the high IQ passing skills that made Saric such a highly touted prospect

ggplot(dario_as, aes(x= coord_x, y = coord_y))+
  annotation_custom(court, -745, 745, -110, 1175)+
  stat_binhex(bins = 7, colour = "gray", alpha = 0.7) +
  scale_fill_gradientn(colours = c("blue","red")) +
  geom_rug(alpha = 0.2) +
  guides(alpha = FALSE, size = FALSE) +
  xlim(-850, 850) +
  ylim(-150, 1000)

Overall, we can see that Saric was both highly productive and efficient. The bar charts below show Saric’s averages relative to the league average for the top 20 players in each category and his average per 30 minutes. We can see that he excels as both a scorer and a rebounder. Put contrary to his scouting report, his number suggests that his assist rate lags behind his scoring and rebounding

ggplot(dset, aes(x= ave, y = stat, fill = ave)) + geom_bar(stat = 'identity') + facet_wrap(~category)

DELANEY, MALCOLM

The second type of prospects who may have be highly touted out of college or development leagues but are effective players in their leagues. A good example of this type of player in dataset in Malcolm Delaney. Delaney is a player who went undrafted out of Virginia Tech but was a highly productive player in international leagues. In the 2015-2016 year, Deleaney actually played the most minutes per game

Today, Delaney is a backup point guard for the hawks who has a consistent 3 point shot, but is by no means deadly. And this is reflective of his play in Euroleague

ggplot(delaney, aes(x= coord_x, y = coord_y))+
  annotation_custom(court, -745, 745, -110, 1175)+
  geom_point(aes(color= make, alpha = .8), size = 3)+
  xlim(-850, 850) +
  ylim(-150, 1000)

Delaney took a large portion of his shots from three, and he was fairly effective (40%). In the NBA he takes much less threes, and his 3 point shot has not translated entirely (he shoots 37%) 

shotchart2

In international play, Delaney was a ball dominant guard who could both score and pass. In the NBA, he doesn’t log heavy minutes or have the ball in his hands as much as he did in Euroleague play. But his sharp shooting and penchant for finding the open man has allowed him to be a solid role player in the NBA.

ggplot(dset, aes(x= ave, y = stat, fill = ave)) + geom_bar(stat = 'identity') + facet_wrap(~category)

Luka Doncic

The last type of prospect is the young potential who flashes immense potential. A good example of this is Luka Doncic, the potential number 1 draft pick in the 2018 NBA draft. During the 2015-2016 season, Doncic was just a 16 year old who sparingly played. Yet was there anything in his play that indicated the massive amount of potential that has him buzzing as the number 1 overall pick this year?

Doncic didn’t play much, but that didn’t stop him from being aggressive. Despite playing 5 minutes agame, Doncic averaged about 2 points a game and took about 2 shots a game. While only a small sample, the shot chart below shows that he shot over 50% more threes than twos

ggplot(luka, aes(x= coord_x, y = coord_y))+
  annotation_custom(court, -745, 745, -110, 1175)+
  geom_point(aes(color= make, alpha = .8), size = 3)+
  xlim(-850, 850) +
  ylim(-150, 1000)

shotchart2

Transforming Doncic’s status to per 30 minutes averages may be extrapolating a bit, but it does reveal that even at a young age, he was an efficient player who excelled at scoring, passing, and rebounding.

ggplot(dset, aes(x= ave, y = stat, fill = ave)) + geom_bar(stat = 'identity') + facet_wrap(~category)

Conclusion

Drafting international players and finding talent in the 2nd round of the draft or among free agents has become an integral part of building a competitive NBA team. I began this project to solve a growing need for NBA front offices to effectively scout and evaluate players in foreign leagues. I now that the implications of this project are limited, but I hope that it sets the foundation for future work!

Thanks for reading