NBA, NBA Draft, philosophy

The Art and Science of “Feel” in Basketball

I like this picture of CP3. Reminds me of a certain meme template.

I’ve always wanted to get cute with an intro and this is the launch of my website, a piece about my basketball philosophy, so that’s exactly what I’m going to do. Let’s start with an interactive activity. Local Draft Twitter legend and all around good guy PD Web designed a titular prospect development “web” which postulates 8 skills that, when combined with a prospect’s tendencies and level of opportunity, comprise that prospect’s basketball “ability” (my words not his). These categories are feel, tools, shooting, point of attack defense (or on-ball defense), off-ball defense, playmaking, dribbling, and role. Within each skill category, a prospect is graded within a sector ranging from A-D, where outer sectors (D) represent a more common level of skill (larger population) and inner sectors (A) represent a rare level of skill (smaller population).

Photo Courtesy of PD Web (@abovethebreak3)

As the introduction to this piece, I’d like you, the reader, to rank these 8 skill categories in an order where #1 is the skill you think is the easiest for a player to improve and #8 is the skill you think is the most difficult for a player to improve. I’ll give you a second to do it, and I encourage you to write them down or put them in a NotePad doc or something. Hold on to the list for later. Like a Memento. Might even be fun to look back at in a year and see if future you agrees with current you.

Tattooing is always an option if you don’t have a pen and paper present.

I’m guessing since you scrolled this far, you’re done. Cool. Here’s my list:

  1. Role
  2. Shooting
  3. On Ball Defense
  4. Playmaking
  5. Off Ball Defense
  6. Feel
  7. Dribbling
  8. Tools

Your list may be identical to mine, or it may be the complete opposite. What’s important is that you conceptualize (and keep in mind while scouting) which skills you value most and which of those skills you think you can teach. Obviously none of these are purely physical or purely mental. In addition, these skills are interdependent – the synergies between them (i.e playmaking and feel) amplify utility of those skills. That said, take a look at my rankings and you might notice a slight trend.

On the whole, I ranked more physical/technical skills (shooting, on ball defense) as easier to learn and skills I consider to be highly mental (playmaking, off ball defense) as more difficult to learn. Some of this has to do with available opportunity to practice a skill. For example, a player can practice the rote mechanics of shooting a basketball in a variety of situations and, eventually, the mechanical action will become autonomous. Up until the highest levels of skill acquisition, shooting can be practiced in an empty gym.

On ball defense is the sum of positioning, strength, agility, and footwork in service of the goal of stopping an opponent from doing what they want to do. This one is something every player will be asked to do to some degree during practice/scrimmages/games/whatever. You can’t escape playing defense because if you’re bad, you’ll get targeted (ask Trae Young or Zach LaVine how that goes), so you’re going to get opportunities to practice in game situations as long as you’re playing.

Dribbling, on the other hand, requires a high level of dexterity and is difficult to practice in game after a certain age. This is because only one player can dribble the ball at a time, and players who don’t meet a threshold of dribbling ability will not be asked to fulfill that role as often in game situations. This limits the number of reps the player can get (and the number of situations they can apply the skill to), thus putting a soft cap on the level they can reach. Tools are largely innate physical skills (able to be honed obviously, but there is a genetic component at play here too).

You’ll notice I didn’t speak about feel – very observant of you! That’s because feel is a daunting concept, rarely ever defined to any satisfying extent. Now that you have some cursory background into what I believe and what I value when it comes to scouting prospects, you can probably guess that prospects I rate highly will tend to excel in the “feel” category and in other skills that synergize well with feel (especially at a young age). However, this doesn’t mean that a player’s feel is innate and unchangeable. In this piece, I aim to lay out my definition for “feel.” I’ll also discuss why I believe feel can be developed and I’ll suggest a few high level suggestions for ways to develop it.

Small disclaimer: this piece is meant to give some scientific background and introduce you to a way of thinking. This piece is not intended to be prescriptive in any way. I can hint at potential methods to utilize these concepts in player development, but I never played basketball at a high enough level to have any real intimate knowledge of developing drills or running practices. Take this knowledge and run with it, and show me what you do with it so I can learn from you.

In addition, I am not proposing we attempt to quantitatively measure the micro-skills discussed in this piece and I will not suggest ways in which you can do so because that would be irresponsible. There are probably good statistical proxies to be discovered, but I’m not an advanced enough statistician to be sure. These are concepts to be considered when deciding how best we can help players become their best selves; ideas to ponder while you watch film and think about basketball. These are not suggestions for additional, potentially invasive, biometrics to be collected at the combine. Alright, I’m off my soap box. Let’s get into it.

What is “feel?”

I think it’s important to tell you where I’m coming from here because I think it informs a lot of my philosophy and why I think the way I do. I’m a lifelong lover of science. My degree is a double major in molecular and cellular biology and psychology. My passion lies in the study of the brain, and I was fortunate enough to do neurodevelopment research in an academic lab setting as an undergraduate. I say all of this to say: even when it comes to my hobbies, I spend a lot of time thinking about how concepts in neuroscience might apply.

Basketball scouting and player development are areas where I feel an understanding of cognitive and developmental neuroscience can be extremely valuable. Basketball is a game of decisions. A game predicated on taking advantage of angles. A game of action and reaction. As mentioned before, “feel” and “basketball IQ” are frequently used in analysis and discussion of the game to describe (more or less) a player’s ability to assess the basketball problem and solve it in a time efficient manner. Perhaps this is lack of better alternative, perhaps it’s a quick stand in to save the lengthy description, but I prefer to try to be explicit about my criteria whenever possible. If I had to describe feel in more concrete terms, I’d describe it as the sum of a player’s pattern recognition, visual processing (especially spatial recognition), and processing speed.

Pattern Recognition (The Awareness)

The first part of solving any basketball problem is perceiving that there is a problem to be solved and being aware of all the participating variables. It sounds easy, but it’s one of the most difficult skills to acquire as a basketball player. Since there are a near infinite number of ways that a basketball game can theoretically play out, no two situations are exactly the same. However, games can be broken down into possessions and possessions can be broken down into actions and schematic decisions on both ends of the court. Pattern recognition is the ability to observe all variables in these “problems” and apply broader templates to them, usually gleaned from past experience, as a heuristic to decrease the amount of time the brain takes to come up with a solution.

Here’s a basketball example: the main character in this scenario is the point guard on offense, and that’s who we’ll be today. Not in a spooky, ethereal, Hereditary Type Shit way or anything, we’re not possessing anyone. This is just a hypothetical.

Ok so we’re running side pick and roll (last time I’ll ever spell this out on my site, it’s P&R from here on out) with four out principles on the right side of the court. Our big screens on the left side of our defender, and we hit him with a hang dribble to the left and then a left to right cross to reject the screen. Nasty. The crowd goes wild. Our defender cheated left on the hang dribble, so he’s beat, and we see the defending big show. Heuristically, we know that our roller is now wide open in the dunker spot, unless the low man on the weak side slides to help. At this point, we have two options and we just have to wait to see what low man does to narrow that down to a cool one option. If he slides, we kick out to the weak side corner. If he doesn’t, we hit the big in the dunker spot for an easy two.

This (overly simple) example illustrates the core tenets of pattern recognition in basketball. We recognize the variables in the problem: the position of all players on the court, the set we’re running, our teammates’ skillsets and tendencies. We’ve seen the start of this film enough times to know what’s coming, and we can watch the middle bits play out (how the defending guard plays the screen, what P&R coverage the opposing team’s big plays, what the low man will do as a result of the advantage we created) for the rest of the information we need. After that, we apply a templated solution (who to hit with the pass for the best look) tested in past experience as a response to the predicted ending. Filling in some of the blanks helps us jump to logical conclusions and maker quicker decisions.

What we just did is called “chunking.” We broke the problem up into smaller sections with memorable distinctions (scheme and the most common reactions to it) and we used past experience to inform what the most likely reaction would be, resulting in our solution taking less time to execute. Your brain is applying hierarchical plans to the sequencing of your behavior every time it solves a motor problem (Lashley, 1951). This is supported by research that suggests that the cerebellum, which is most frequently implicated in executive planning and motor control, has been shown to be active during sequential learning (Jenkins et al, 1994). In addition, the hippocampus, the brain area implicated in memory, has also been associated with the ability to imagine the future (Hassabis, Kumaran, Vann, & Maguire, 2007). This implies that past experience in some way informs the planning and execution of executive motor behavior during sequential tasks; in this case a half court P&R possession.

If your brain is making associations based off past experience and using those associations to plan future action, it’s only natural that sequential patterns can be correlated in memory and recalled together to decrease the time to recall. Indeed, it has also been suggested that the long-term acquisition of skills is naturally characterized by the process of forming ever larger hierarchical units or ‘chunks’ (Miller, 1956). These kinds of heuristics save players precious milliseconds, and in a league where the action comes at you fast, those milliseconds can be the difference between success and failure.

Visual Processing (The Assessment)

However, it’s not as simple as diagnosing a static situation and applying a solution that worked in the past. Basketball is a game where parallel processing – the ability to be aware of and process multiple stimuli that are all actively changing – is key. Specifically, the player needs to be able to visually assess the three dimensional positions of all other players and the ball as they navigate through space. Keeping a mental map of these variables and possessing a high level of innate understanding of the relationship between space and time are key advantages for high level athletes. On an even more granular level, the best players can be (consciously or subconsciously) aware of a variety of physical cues – things like center of gravity, hip angle, foot placement, etc – that give them an advantage creating space on offense or taking it away on defense.

That’s where visual processing comes in. Visual processing is the ability for the human eye to decode information from visual stimuli and convert it to understandable and actionable information for the brain. Accurately and quickly assessing the position of players on the court is especially critical to this process. If visual information processing is inaccurate or too slow, the information gathering step of pattern recognition will become a bottleneck that slows the player’s time to a solution dramatically, resulting in missed opportunities on the court. You’re probably already thinking of a million different ways this occurs on the basketball court, but my favorite example is the aforementioned P&R example, so let’s go back to that.

Say we hadn’t realized our defender cheated left, and therefore we didn’t see him throw his center of gravity away from us, taking him out of the play. That’s a missed opportunity. Or perhaps we successfully beat our defender off the dribble, but we weren’t aware of where help would come from and by the time we get to the rim, we’re surround by trees. Our recognition of these variables is aided by past experience sure, but we still need to constantly visually re-assess what’s going on around us so we can update the heuristic with new information.

Visual processing isn’t just helpful for the recognition and planning stages of movement though. Spatial awareness is a specific genre of visual processing that also incorporates proprioception – otherwise known as our ability to perceive the position and movement of our own body through space. Spatial awareness is key in both the planning and execution stages of movement. An athlete with a high level of spatial awareness can accurately diagnose how big a passing window is and therefore how long it will be open while simultaneously directing his limbs to execute the perfect live dribble pass to take advantage of that window.

Developments in these areas have been implicated in success at the NBA level. In one experiment, a multi-object tracking paradigm was used to assess visual tracking speed in a cohort of 12 NBA players. Results on this assessment were correlated with season-long statistics, and a higher VTS scores were correlated with better performance in AST, TO, AST:TO, and STL statistics (Magine et al., 2014). Many scouts like to use AST:TO and STL% numbers as part of a proxy for feel, and this result seems to corroborate that practice.

Improving proprioception has become a hot topic in player development circles, perhaps peaking when Bleacher Report covered LeBron’s use of “yoga bubbles” for proprioceptive assessment and improvement. Proprioception has understandably been suggested as a focus for injury prevention, which isn’t surprising when you think of how important perception of the self in physical space is for landing technique, footwork, and more. However, improving proprioception has been shown to improve more salient “basketball skills” too.

In an experiment involving 26 amateur basketball players in their early 20s, one group underwent a 12 week proprioception training regimen, and the other group did not. Both groups’ passing skills were assessed using the AAHPERD test of passing skill. The experimental group was found to have improved their passing ability by a statistically significant margin compared to the control group (Nikolaos et al., 2012). The assessment is pretty dated and the sample size is admittedly small, but this is certainly a relationship that should get a closer look.

Processing Speed (The Action)

It’s all well and good to be able to diagnose the situation at hand, to be aware of all the variables in this situation, and to have a solution in mind. The real key that separates players described as having “high basketball IQ” from those who struggle is how fast all this information is being processed. Processing speed is the rate at which the brain encodes information from exterior stimuli, processes that information, and begins to respond.

There are “record scratch” moments in pretty much every basketball game: the jarring “deer in headlights” vibe an offense gets when the perception of an opportunity comes too late, resulting in the opportunity disappearing into thin air. When things are humming though, processing speed shows up in ways that can make basketball look almost effortless. I could go into a variety of basketball examples, but Jake Rosen (@jakeinthepaint) recently wrote a seminal article about processing speed and how it can be assessed via film scouting. I highly encourage you to check it out before you read the rest of this section.

Understanding processing speed requires a little bit of background on molecular neuroscience and the way the brain works. The eyes perceive visual stimulus and encode that information into electrochemical signals that are transmitted to various areas of the brain by neurons – this process is called visual encoding. Neurons transport information through a process of electrical excitement. The neuron equilibrates the flow of sodium and potassium across its cell membrane until it hits a certain voltage and then it fires an electrochemical signal across the axon to the next neuron. This is called the action potential.

Important to note, the rate of neural transmission is independent of the level of stimulation. Whether a neuron is at exactly its action potential or 100mV above it, it’s going to fire exactly the same way. Once the neuron fires, it’s up to the axon to conduct that signal to the next neuron and so on until it eventually hits the brain (if it’s an input) or a muscle (if it’s an output).

So if the strength of the signal doesn’t affect the speed at which it’s propagated, what does? The answer is myelin, a fibrous sheath around the axon that improves its ability to conduct electrical signals. It has been illustrated via magnetic resonance imaging (MRI) that learning complex sensorimotor tasks (like playing basketball) changes the structure of white matter in the brain, rewiring connections and changing the degree of myelination, and therefore the speed of transmission, of these connections (McKenzie et al., 2014). In addition, neuronal circuits that are engaged more during learning may signal to other cells to generate more myelin (Long and Korfas, 2014). More myelin means stronger connections and a more efficient transfer of information along any neuronal chain. This is what processing speed improvement entails, and this is part of why engaging these neural circuits frequently and in a variety of ways helps development of sensorimotor skills.

How do you develop all these subskills?

I will fully admit, I haven’t solved this part yet (or even come particularly close, frankly). This is an area of the game of basketball where nerd Twitter needs to do the fusion dance with coach/trainer Twitter. This is the second time I’ll say it: if you want to run drill ideas by me or you have follow up questions about this piece, literally anything, I’ll do whatever I can to help. DM me. The next frontier is going to be developing drills and other training activities designed with these principles in mind, and I don’t have the tools or the experience to make that happen right now.

With our powers combined…

That said, I do think all three subcategories of feel can be developed. We’ve seen a small scale “experiment” of this kind recently, in-fact. The Ball brothers played their entire youth careers in a free-flowing, fast paced, open offense. The goal, especially at Chino Hills, was to create advantages through a series of fast paced offensive possessions designed to force quick decisions. This meant that fast processors thrived, and those that took longer to process found little success. The result was two Balls drafted in the top 3 of their respective drafts. We don’t talk about LiAngelo.

Both Lonzo and LaMelo were highly touted as decision-makers and creators coming out of college. The NBA is the highest level of professional basketball worldwide, and that means the margins are extremely fine. In a league filled with world best athletes who have been playing basketball for decades, the pace of play is breakneck and that means each decision you make is a quick one, theoretically a half second or less in some offenses. After years in similar systems, the Ball Brothers became experts at making good decisions with little time to do so, and their respective tenures in the league have reflected that so far.

That’s an illustrative example, but it’s small sample size theater and it’s anecdotal at that. The developmental conditions of Big Baller Brand Basketball are not 100% reproducible. However, a review of the academic literature backs up the hypothesis that pattern recognition, visual processing, and processing speed are all skills that can be developed. Let’s do a cursory look into some skill development research.

It has long been assumed that brain plasticity is at its greatest when people are young and that plasticity slowly decreases as we age (Pauwels, Chalavi, and Swinnen, 2018). To some degree, this is true, as white matter generation is at its greatest during early development. However, we’re constantly pruning and rewiring synapses in the brain, and while we may not generate as many new neurons, we can always adjust how the ones we have are connected (Kolb and Gibb, 2011). In fact, it’s been found that practice increases the efficacy of nearby neurons working in the same direction during visual processing tasks, and that this group of local interactions may improve connectivity with remote neurons (Lev et al., 2014).

The molecular mechanisms are interesting and they’re a good basis of proof to help us feel confident that these skills can be improved. The million dollar question then becomes “what is the best way to foster development of these skills in a basketball environment?” To start, I firmly believe that incorporating elements of basketball development as early as possible is beneficial; this is especially true for bimanual skills like ball handling and for spatial awareness dependent skills like court mapping where a larger number of experiences means a larger back catalogue of solutions to try (greater variety of past experiences = a higher chance one of those is applicable to the situation at hand). At the end of the day, practicing these skills more leads to a higher level of myelination and the gains in myelination are larger when the player is newer to the skill (Lakhani et al., 2016). Make those big gains happen when brain plasticity is at its fastest!

My focus in developing drills would be to simulate a variety of situations by imposing custom rules. Changing what’s allowed or disallowed by each player, or even altering the speed of play (and therefore how long each player has to make a decision) can be done in a number of ways, and any change will produce novel situations for the players to apply their skills. Think about your drills like a round of the minigame All Fired Up from Mario Party 3. The patterns and the speed of the fire change, but the goal of the game remains the same.

When I played soccer throughout my youth, we would frequently train by playing small sided games on a compressed field with rules about possession and distribution of the ball. A rough basketball example might be a 3 on 3 possession game where the court is restricted to about 25% of its usual size and players are allowed a second or less to decide on their next pass. The team in possession of the ball aims to keep it, and the defending players’ goal is to steal the ball. Success is measured by the amount of time the offensive team can hold onto the ball without taking longer than a second to make a decision.

A really good example of a drill I’ve already seen used by Trae Young and his trainer is the video below, where Trae is thrown some absolutely awful passes and his instruction is to gather, rise, and shoot no matter how the ball arrives.

While this one is more shooting-specific, the idea behind the drill is what I appreciate most. The goal is to get Trae used to arranging his footwork no matter how he receives the pass. The drill simulates a realistic concern for real life basketball games: sometimes you’re going to get some awful fucking passes! This drill also engages the previously discussed neurological systems – visual processing (flight pattern of the ball as he receives it) and executive planning of motor control (queuing up the action of the gather so he can rise and shoot when the ball arrives) and the conditions around the action are unpredictable and constantly changing. This allows Trae to build onto his heuristic for how passes can arrive and what he needs to do to convert shots afterwards, and encourages the molecular development of the systems surrounding these processes. The result is (hopefully) that Trae feels comfortable receiving any nature of pass and will have the awareness and muscle memory to convert regardless of pass quality because he’s seen this before.

Acknowledgements

I wanted to take out a small section of this to thank some people, because I couldn’t have put this together (or frankly even come up with the idea) without all of the wonderful discussions I’ve had with so many folks on Twitter as well as some of the foundational work that has really helped me shape my views on the next frontiers in basketball scouting and player development.

I wanted to give specific thanks to Ben Taylor, whose book “Thinking Basketball” marked a turning point in my basketball viewership. Reading that book was the first time I had ever seen cognitive science used in service of sports analysis, and it fundamentally changed the way I think about the game of basketball. Without Ben, I don’t think I would have ever considered applying my scientific background to sports.

I also wanted to thank all my friends in the draft twitter community. There are too many of you to name, but draft twitter is one of the most curious, insightful and intellectually rigorous groups of sports analysts in the world, and I’m convinced that my conversations with you all make me a smarter person every day. Extra special shout out to PD Web, Ross Homan, Ben Pfeifer, Max Carlin, Jake Rosen, and Spencer Pearlman. The philosophical work you guys have done on the draft is monumental and you’ve all inspired me to think more critically about scouting and player development.

Last but certainly not least, I wanted to thank my good pals at Premium Hoops. Mark Schindler, Nate Georgy, Cody Houdek, Scott Levine, Corban Ford, and Joe Hulbert are some of the most brilliant hoops minds I know and have all been extremely supportive of me while I wrote this thing. Many hours spent podcasting and countless more shooting the shit with these guys has honed my basketball mind, and their tireless vigor in search of the greater basketball truth inspires me daily. Please, if you don’t already, consider following Premium Hoops on Twitter.

In addition, you can find more of my work over at PremiumHoops.org, including a brand new basketball philosophy podcast called “Sense and Scalability” that I’m co-hosting with Cody and Scott. We’re planning to discuss one theoretical and/or philosophical basketball topic a week – could be anything from the Wrong Initiator Conundrum to ideal roster construction order to the next revolutions in P&R coverage. You can listen to the first episode on Spotify here, and it’s coming on Apple Music soon!

Sources

PDF copies of all the papers I cited can be found at this link. I was going to put together a bibliography, but I’m not THAT much of a nerd.

12 thoughts on “The Art and Science of “Feel” in Basketball”

  1. This is fantastic.

    I was a kid into basketball when Larry Bird entered his career and got to watch his entire peak. He is who I think of when I think of feel. And what folks who didn’t see him play games miss when they evaluate him. His spatial awareness and speed of thinking are can be compared to just tiny amount of other players. Have you watched enough of him? He would be a perfect example to illustrate this.

    Look at the whole development of his famous steal against Detroit. He starts on the floor after his layup attempt is blocked. The Celtics think they have possession, the ref says no and he ends up being the only Celtic in Defensive position and he takes one and then jumps to another. I’m not sure who else in NBA history would have been able to do that. Watch from here

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s