In my previous article I looked at Balanced Playmakers and how they contribute to their teams offensive production. It was especially interesting to look at how Steph Currey, J.J. Redick, and Giannis Antetokounmpo could all put their own spin on this archetype. Now we will turn our focus to Spot-Up Wings and how they contribute on the offensive end despite spending much less time with the ball in their hands.

Ready and Waiting

As you would imagine players in this archetype spend almost half of their time spotting up behind the arc ready to launch a catch-and-shoot three. The less obvious piece of information here is how much these players score in transition. At 22% this archetype uses possessions in transition more than any other archetype. This can almost certainly be attributed to how often teams are looking to hit a trailing shooter for three instead of forcing a shot at the rim in transition. On average these players spend no more than 8% of their possessions on any other play type indicating that these players have a very concentrated skill set.

Spot-Up Wings On The Rise

There has been a steady and significant increase in Spot-Up Wings over the last few years. As the league moves more and more towards pace-and-space these types of players will be in high-demand. What also aids in this rise is the league putting more emphasis on scoring in transition before teams are able to set up their half-court defense. I would not be surprised to see this number continue to rise as more teams than ever before are running their offense through one or two designated playmakers and surrounding them with as many switchable shooters as possible.

Putting Faces to the Name

Now that we understand the DNA of a Spot-Up Wing, we can take a closer look at three different players who fall into this group. As you will see, even within the archetypes themselves players can differentiate themselves by how they choose to use their offensive possessions.

Danny Green is exactly the type of player teams would like to run out there as a Spot-Up Wing. He spends almost 70% of his possessions on spot-ups and in transition at league leading efficiency. This is a perfect use of his skill set as these are his most efficient play types and fits perfectly next to ball dominant playmakers like Kyle Lowry and Kawhi Leonard. Any players looking for a long career should look no further than Danny Green’s career especially if they lack the playmaking skills of the elite offensive weapons.

Albeit on a small sample, Luol Deng is a good example of what we might call a “tweener”. The biggest difference between Spot-Up Wings and Spot-Up Bigs is the former uses a higher number of possessions in transition while the latter spends more time posting up. In addition to the expected frequency of spot-ups (36%), Deng spends 27% of his possessions in transition and an additional 14% on post-ups. You will also notice he adds absolutely zero value through playmaking regardless if it is in a ball dominant fashion or leveraging off-ball movement.

Where our first two players added at lease some variety to their offensive usage, Bertans sees no need when you can knock down shots with relative ease. At almost 60% of his possessions being used by spot-ups, three times more than any other play type, Bertans is the epitome of a specialist. Seeing that he finished in the 92nd percentile in spot-ups last year with 1.21 points per possession, the Spurs had no issues leaving him on the perimeter to provide spacing for Derozan and Aldridge to work inside. He did rank relatively high in other play types on small samples which may indicate that if he were asked to do more within the offense he could deliver with at least some efficiency.

Once again we can see that even within an archetype players can carve out their own way to impact the game. While Spot-Up wings only provide value in limited ways, they are more valuable than ever in today’s NBA. We are seeing this impact most in the draft where fewer big men are hearing their name called and teams are constantly looking to draft the next sharpshooter. As the league continues to put an emphasis on outside shooting more and more Spot-Up Wings should find their way into meaningful minutes.

Beyond The Arch: A Closer Look at Spot-Up Wings was originally published in Playing Numbers on Medium, where people are continuing the conversation by highlighting and responding to this story.

In my previous article I took a closer look at how the number of On-Ball Playmakers is on the rise and I detailed three examples of the type of player that fit this mold. In this article I will do the same for Balanced Playmakers. While this archetype is also filled with some of the leagues very best playmakers you will notice how differently they achieve these results.

Playmaking Through Movement

Balanced Playmakers attack using off-ball movement more than any other type of player with almost 20% of their offense coming from Hand-Offs and Screens. While these players are not known for scoring via isolation they do utilize the pick-and-roll much like their On-Ball counterparts. That being said when not using off-ball screens or hand-offs they do spend a lot of time, around 27% of their possessions, spotting up behind the three-point line.

Slight Increase in Balanced Playmaking

Similar to On-Ball Playmaking we have seen an increase over the past couple of years however on a much smaller scale. Playmaking of any kind is as important as ever but not just any player can be used in this way. This can be attributed to the importance of shooting. Off-ball screens and hand-offs in many cases end in shots from three or the long mid-range. Add in the fact that these players spend a little more than a fourth of their possessions as a stationary threat from beyond the arc and it is clear being at least a decent long-distance shooter is a prerequisite for being a Balanced Playmaker.

Putting Faces to the Name

Now that we understand the DNA of a Balanced Playmaker, we can take a closer look at three different players who fall into this group. As you will see, even within the archetypes themselves players can differentiate themselves by how they choose to use their offensive possessions.

Steph Curry is most certainly the most effective Balanced Playmaker the game has ever seen. He not only attacks with volume and efficiency using the pick-and-roll but he is even more efficient coming off screens. Even when his teammates are the focal point of the offense he scores 1.23 points per possession spotting up behind the arc putting him in the 94th percentile. Steph used possessions in nine different ways last season and ranked in the 80th percentile or better in all but two. This is why opposing coaches lose sleep over game planning for the Warriors every season.

J.J. Redick is exactly the type of player I believe is underappreciated when using the traditional positions of basketball. Almost anyone will call him a shooting guard, and even if we consider him one of the best shooting guards of all time we are still missing the point. Most people associate spotting up with J.J.’s offensive game but he spends twice as many possessions on hand-offs with 85th percentile efficiency. In addition to hand-offs and spot-ups, J.J. ranks well above league average efficiency coming off screens, in transition, and surprisingly as a pick-and-roll ball handler. I believe it is about time we stop trying to put Redick in a shooter’s only box and recognize him for the playmaker that he is.

I imagine if my model could talk it would tell me Giannis was the most difficult player to put into one of the six groups. Because of his unique size and skill set Giannis is used differently than any other player we have seen. Giannis is constantly looking to push the ball in transition with 27% of his possessions used this way. What is unique about Antetokounmpo is how often he attacks via the post-up. Only four playmakers post up more often than Giannis which can be attributed to how often he as a mismatched defender on him. At 6’11 and 240+ pounds, with the ability to run the offense with the ball in his hands, there are only a handful of people in the league that can slow him down in the slightest. Giannis spent 5% or more of his possessions on eight different play types ranking no lower than the 56th percentile in efficiency. The scariest thing about Giannis is not that he can attack in so many ways but that he is only 25 years old and will only get better from here on out.

Conclusion

While sometimes not as flashy as their on-ball counterparts, Balanced Playmakers still represent some of the biggest names in the NBA. What is so interesting about these players is how differently these playmakers can contribute on offense. It is not necessarily one or even two different play types but rather a handful of actions these players can utilize if the situations calls for it. This this type of versatility and effectiveness I would not be surprised to see more and more players trying to imitate.

Beyond The Arch: A Closer Look at Balanced Playmakers was originally published in Playing Numbers on Medium, where people are continuing the conversation by highlighting and responding to this story.

Plenty of hyped rookies ended up sitting out their initial draft year due to injury. The likes of Detroit Pistons star forward Blake Griffin, and a pair of Philadelphia 76er All-Stars in forward Ben Simmons and center Joel Embiid sat out their draft year and ended up being a strong force with their squads. As we all know, Zion Williamson has been the most hyped rookie in recent memory. It all started with his high school clips in South Carolina. Someone with such athleticism, size, and natural talent has been a joy to watch flourish from high school years up until the NBA preseason.

In his small sample size of four games in the 2019 NBA preseason, Zion Williamson averaged 23.3 PPG, 6.5 RPG, 2.5 APG, along with 1.5 steals per game (per FOX sports.com). There were flashes of greatness, long strides down the court in a successful attempt of a Lebron-esque chase down block. Along with those flashes, came flashes of amateurism. Breakdowns on defense, poor passing decisions, and poor shot selection, given his 25% 3-pt shooting. He earned his #1 overall pick selection, but we have yet to see him on the court this regular season.

According to Reuters.com, the meniscus injury is commonly recovered from, showing almost no regression prior to the injury. The timetable of a couple months makes us think that Zion should be ready to make a return this month, December of 2019, but the basketball world has heard nothing of the sorts. Where is Zion Williamson? As a competitor of his stature, he is more than ready to make his return, and the New Orleans Pelicans know that. With a franchise-altering piece such as Zion, we are beginning to realize just how hesitant the Pelican’s front office is when it pertains to their star rookie’s return.

Let’s take a look at some of the Pelican’s team statistics through December 17th and see if Zion’s presence is needed on this roster.

According to the stats above, Zion’s rebound ability is something that the Pelicans really need on their roster. Zion’s offensive rating could really help bolster the lackluster Pelican offense, which is 19th in the league when it comes to scoring the ball. His defense could use some help, but this is a common issue among rookies, especially the ones with such outstanding scoring ability like Zion Williamson.

My verdict is to let Zion come back on a minutes restriction, maybe at 20 minutes a game for a few weeks, or even until the all-star break. Zion’s health will definitely have to be monitored closely, and the medical staff will have to be on full alert. Let me know below what decision you think the Pelican’s front office should make when it comes to Zion’s return to the NBA.

To Start or Not to Start, That is the Question (For the Pelicans) was originally published in Playing Numbers on Medium, where people are continuing the conversation by highlighting and responding to this story.

Isolation and Pick-and-Roll Artists

On-Ball Playmakers are defined by heavily relying on the pick-and-roll to generate points as well as taking their defender one-on-one in isolation. These types of players also seem adept at scoring in transition, in many cases leading the break as the primary ball-handler. While some exceptions exist, like Lebron James and Kawhi Leonard, these players typically do not post-up preferring face-up when attacking their defender.

On-Ball Playmakers on the Rise

The number of On-Ball Playmakers has risen over the last couple of years. This may have a lot to do with the increased focus on the pick-and-roll across the league. Not only are teams looking to draft playmakers year in and year out, veteran NBA players work rigorously in the off-season to add these skills to their game.

Putting Faces to the Name

Now that we understand the DNA of an On-Ball Playmaker, we can take a closer look at three different players who fall into this group. As you will see, even within the archetypes themselves players can differentiate themselves by how they choose to use their offensive possessions.

James Harden in some ways has created his own archetype. Harden operates via isolation more than any other player, and it is really not that close. In fact, Chris Paul was the only other player last year to use over 30% of their possessions by taking their defender one-on-one. With another 23% of his possessions as a pick-and-roll ball-handler, James Harden has the ball in his hands on 74% of the possessions in which he takes a shot or commits a turn-over. If we need a poster boy for a ball dominant playmaker, The Beard is our guy.

Kemba Walker dominates offensively with the ball in his hands differently than James Harden. While only 9% of his possessions were used in isolation last year, almost 50% where spent as a pick-and-roll ball handler. This is most certainly a sound use of his offensive time as he ranks in the 91st percentile in points-per-possession (PPP) as a conductor of the pick-and-roll. Like Harden, Walker ranks better than the 80th percentile in both Isolation and PnR points per possession.

The great thing about these archetypes is they allow us to compare players who otherwise would be in different traditional positions. While Kemba and Harden are both Point Guards by traditional standards, Lebron is very much a Forward even if we call him a “Point Forward” with how versatile he is on the court. Lebron spent 30% of his possessions last year in the pick-and-roll with another 19% coming via isolation. This is much more balanced than our two previous case studies but still represents almost 50% of James’ possessions spent attacking with the ball in his hands. In this new world we can compare Lebron to Harden and Walker despite being the size of an NFL linebacker.

Conclusion

As you can see On-Ball Playmakers come in many varieties. Many of the leagues best players fall into this category which should not be surprising. NBA teams have a lot to gain by putting the ball in the hands of their best players. What we will see in the next article is that you can be a star in this league without operating with the ball on a string in isolation and the pick-and-roll almost 50% of the time. In my next article I will take a closer look at Balanced Playmakers.

Beyond The Arch: A Closer Look at On-Ball Playmakers was originally published in Playing Numbers on Medium, where people are continuing the conversation by highlighting and responding to this story.

The Mission and Data

Using play type data from NBA.com I set out to find common themes in how players were used on the offensive side of the ball. Using these similarities I would group players based on how they were used within their offense. I knew that I would need to embark on a data science journey to make this happen but first I needed some data. Fortunately NBA.com has play type data going back to the 2015–2016 season for all qualified players. Below is a list of the play types available as well as the top players in Pick & Roll Ball Handling possessions per game for 2018.

Once I had scraped, manually by hand into CSVs, all of the seasons available or each play type, excluding the current 2019 season, I loaded the data into Power BI which is my tool of choice when joining, cleaning, and visualizing data. While I had many elements to choose from I focused solely on the percentage of total possessions each player used on a given play type. This allowed me to group similar players regardless to total possessions used or efficiency. The end result, which I would use as an input for my model, looked like this:

The Methodology and Model

When I first set out to complete this project I phoned a good friend Ken Jee for some advice. I explained to him what I was trying to do and he pointed me towards K Means Clustering. K Means is a type of unsupervised machine learning model that takes unlabeled data and groups (clusters) them together based on the features you provide. I will save the fancy data science explanations for the experts so here is a nice Medium article explaining this type of model in more detail if you are interested.

I first loaded in the packages I needed to build the model for my player archetypes. Some of these I ended up not using for the model itself but the main items I needed below were sklearn KMeans and pandas.

Next I needed to load in the source data and preview the first few rows to make sure everything was in order.

With my data locked and loaded I now could start to build my KMeans model. The first step, and possibly most important, is deciding how many groups or clusters to model. The most effective way to do this is using the elbow method. A detailed explanation of this can be found here but essentially we are testing out the data with several numbers of clusters to see where we start getting diminishing returns. In the chart this point will look like an “elbow”. Below you can see the code and result for my data.

Many people would look at the graph above and declare the elbow value as 3. They would not necessarily be wrong however knowing basketball as well as I do, I knew that there are many more types of offensive players than three. There seems to be another slight decrease in value at 6 and so this is the number of clusters I went with. After seeing the data I know that by going from three to six the model broke down three overarching groups into two slightly more nuanced groups each. All that was left to do now was fit the model and add the new cluster values to my existing data.

The Results

When I started to look into the data I was giddy with excitement. The players grouped together made sense in my head based on my understanding of the game. While the labels themselves were just a number from 0 to 5, I knew exactly what each of these meant in the grand scheme of things. Below is a table showing the five archetypes created in my model and the average frequency of each play type for the players in that group.

I am not 100% married to these labels, and may tweak them as we go, but you can see some clear lines drawn in the sand. On-Ball Playmakers are defined by an increased number of possessions in isolation and as a pick & roll ball handler. Balanced Playmakers on the other hand mix in off-screen and hand-off actions to keep defenses on their toes. Spot up specialists are defined by a high number of possessions used by spotting up behind the arc while their teammates create. I use the terms Wing and Big loosely for these players because the only difference is Wings also create frequently in transition while Bigs operate frequently as the roll-man in the pick & roll. Post Up Bigs on average post up more than all other archetypes combined and spend a good amount of time in the pick & roll. Finally, Rim Runners very much rely on picking up the scraps on offense mostly generating buckets through setting screens and rolling to the basket, cutting to the basket when their man helps at the rim, and cleaning up teammates misses via the put back.

With my new set of archetypes in hand I can take a closer look at how each player is used within their offense and how each group of players creates value for their teams. Beyond the Arch is going to be a series of articles where I try to leverage these new offensive archetypes into a better understanding of the game of basketball. Through this process I will try to answer questions like :

• Which archetype is most valuable?
• What players excel within each archetype?
• Can players roles change over their career? How about from game to game?
• Are certain lineup combinations more effective?
• Can we use archetypes to predict how well a free-agent would fit into a new system?

The number of questions I have is infinite but now at least we have a really cool way to explore them. I hope you enjoyed this and I would love any questions or comments you might have.

Beyond The Arch: Introducing a New Way to Understand the Game was originally published in Playing Numbers on Medium, where people are continuing the conversation by highlighting and responding to this story.