First impressions of SportVU

sportvu

Ever wanted to know how many hockey assists a basketball player racks up, or how many total points a team produces off one player’s passes?

Those statistics come from what all 30 teams have invested in, called the STATS LLC’s SportVU Player Tracking system. It features six cameras installed in catwalks of 29 arenas (the Clippers and Lakers share one, remember), tracking the movement of every player. From there, a program takes in the data and delivers statistics on things like running distance and the assist opportunities I mentioned before, among other things.

And now, those stats are available on NBA.com’s statistic page.

I took a look at all the statistics available, expecting to have no idea how to interpret what they even mean, but the nine categories of available stats (speed & distance, touches/possession, passing, defensive impact, rebounding opportunities, drives, catch & shoot, pull up, and shooting efficiency) were very user-friendly. Most statistics were understandable, but right now standing pat on all the data seems like the right move. Sample sizes for every category are so tiny right now. We’ll have to wait a month or two before analyzing what these statistics say about players we value so highly, ones we value so little, and if SportVU changes our views on those players.

More statistics should be added over time, but what’s available right now by SportVU has been more than enough to keep me glued to my laptop. It helps answer the questions we’ve asked over so many years, but it also left me with new ones. Here are my first impressions from each category and the statistics inside each of them that stood out to me:

Speed and Distance

Definition (from NBA.com): Statistics that measure the distance covered and the average speed of all movements (sprinting, jogging, standing, walking, backwards and forwards) by a player while on the court.

Notable stats

  • Minutes per game
  • Distance Traveled (total miles)
  • Average Speed (mph)
  • Distance Traveled per game (miles)

When filters let viewers sort average speed, etc. by halves and quarters, the minutes per game will be cool to look at. Just how much less distance does a player travel as the game progresses? With that said, I’m still not sure how much weight can be put into stats like average and top speed even when sample sizes increase to an amount worth evaluating.

Distance traveled per game brings the talk about a player’s “mileage” to a literal sense. Just how much does LeBron travel (in distance, not extra steps with the ball) per game? In future seasons, will this further affect how much key players play as the season winds down?

I’ve always been curious just how much players run every game though, and now total miles traveled is a stat. Going into Friday night’s games, Goran Dragic topped at 3.1 miles. Which teams will run more total distance: slower-paced teams because of running more offense in the half court or faster-paced teams who play more possessions overall?

Touches/Possession

Definition: The number of times a player touches and possesses the ball, where those touches occur on the court, how long the player possessed the ball and the number of points, assists and turnovers that occur with the ball in his possession.

Notable stats

  • Touches per game
  • Time of possession with the ball (minutes)
  • Close Touches per game
  • Elbow Touches per game
  • Points per touch

I’d imagine touches per game will be dominated by point guards with star wings and maybe a few sweet passing big men who can get an offense rolling with touches at the elbow. It wouldn’t be surprising either if Al Jefferson and Dwight Howard, two focal points of their team’s offense, will be in the top 25 as well.

Going into Friday’s games, Derrick Favors had 101 touches. He was between Chris Paul (107) and Ricky Rubio (100) for most touches per game. Favors played five more minutes than the two point guards, but I still can’t figure out why he’s in the top three.

But that’s why watching the actual basketball game helps.

I like locations for touches as well, like the close ones. It could be more specific, such as what kind of close touch it is like a post-up, a cut to the rim, or catching a pass off a pick and roll. It’s no surprise on Friday that Marc Gasol led the league in elbow touches per game with 21 as Memphis has run their offense through him at the elbow for some time now.

Points per touch and points per half court touch leave me a little confused right now. Does it only include when the player passes and the guy he passes to shoots?

Passing

Definition: The total number of passes a player makes and the scoring opportunities that come from those passes, whether they lead directly to a teammate scoring a basket or free throw, or if they set up an assist for another teammate.

Notable stats

  • Passes per game
  • FT assists per game
  • Secondary assists per game
  • Assist opportunities per game
  • Points created by assist per game

This is the most exciting category to see. There’s the hockey assist, points created by assists, and free throws created by assists, and assist opportunities which include players who missed their shots.

From 1997 to 2005, Stephon Marbury averaged 8.3 assists per game, which never fails to be surprising when considering how much of a scoring guard he really was. With the era we’re in that features several great point guards (and different types of great point guards), these new statistics may help decipher who really is a creator for others, such as Steve Nash, others who have more selfish assists, like Marbury had once upon a time, those between the two ends of the passing spectrum.

Also, we’ll get to find out just how versatile LeBron James is.

Defensive Impact

Definition: Statistics measuring the impact a player has on defense, including blocks, steals and protecting the rim, which measures the opponent’s field goal percentage at the rim while it is being defended. Rim protection is defined as the defender being within five feet of the basket and within five feet of the offensive player attempting the shot.

Notable stats

  • Opponents’ field goals made at rim per game
  • Opponents’ field goal attempts at rim per game
  • Opponents’ field goal percentage at rim

These should show the difference between big men like Carlos Boozer and Roy Hibbert and guards like Steve Blake and Avery Bradley. Like everything else in these new statistics, we need a month or two before putting some weight into how many attempts a player allows at the rim and just how effective they prevent those attempts from being points.

Rebounding Opportunities

Definition: The number of times player was within the vicinity (3.5 ft) of a rebound. Measures the number of rebounds a player recovers compared to the number of rebounding chances available as well as whether or not the rebound was contested by an opponent or deferred to a teammate.

Notable stats

  • % of rebounds per chance
  • Rebound chances per game
  • Contested rebounds per game
  • Uncontested rebounds per game
  • Contested rebound %

Another exciting category, much like assists in a way. Are some players’ rebounding numbers more inflated than others? Now we get to see just how many rebounding opportunities a player has and how many times they take advantage of those. It can also be telling of how often a player crashes the boards.

And then it’s divided into two more categories: contested and uncontested. In a way, this reminds me of misleading 3-point percentages. It’s easier to grab an uncontested rebound, much like it’s easier to make an open 3-point attempt (unless you’re too wide open). Some three-point shooters, like Ray Allen, can make both kinds. Others, like Rajon Rondo who shot 35 percent from three last November, have much better success when wide open.

We’ll see if a similar case involves rebounding. There’s already been an advance in rebounding statistics, such as the percentage of rebounds a player grabs compared to total shots taken while said player is on the floor, but this adds another wrinkle.

And by wrinkle, I mean a really cool stat.

Drives

Definition: Any touch that starts at least 20 feet of the hoop and is dribbled within 10 feet of the hoop and excludes fast breaks. Measures the total number of drives as well as the points, assists and shooting percentages on drives to the basket.

Notable stats

  • Field goal % on drives
  • Total player points on drives
  • Drives per game
  • Player points per game on drives
  • Team points per game on drives

Just how scary is a player when he starts around the arc and gets into the paint? Like total points from assist opportunities, now we’ll get to see how many points are produced when a player drives into the lane.

After watching the Magic-Wolves matchup Wednesday night, it’s no surprise to see Jameer Nelson being one of the top players in drives per game (as of Friday) but also in the middle of the road when it comes to team points from those drives. It confirms what I saw from Nelson as he would slash into the lane but look for opportunities for others to score once he got there. He’s really only like 5’9”, so looking to pass instead of score would make sense, only there were rarely any good looks generated from his drives.

Meanwhile, James Harden, Ricky Rubio, and John Wall, among others, have generated more team points than Nelson which is understandable. They’re all taller and either finish stronger or have uncanny court vision.

We’re only a few days into the season though, so the sample size for everything, including drives, is microscopic.

Catch and Shoot

Definition: Any jump shot outside of 10 feet where a player possessed the ball for 2 seconds or less and took no dribbles.

Notable stats

  • Catch and shoot points per game
  • Catch and shoot field goals made per game
  • Catch and shoot field goals attempted per game
  • Catch and shoot field goal %
  • Repeat last three for 3-point shots
  • Catch and shoot effective field goal percentage

Basketball-Reference in particular already has ways to show who scores off of passes versus off their own dribble, thanks to assist percentages off made field goals. Danny Green’s made threes last season, for example, were assisted 95.7 percent of the time while Chris Paul’s were at 41.5.

But now we get to see how many total points come off catch and shoot opportunities without the help of a calculator, as well as effective field goal percentages and just how many attempts are taken per game.

Pull Up

Definition: Any jump shot outside 10 feet where a player took 1 or more dribbles before shooting.

Notable stats

  • Pull up points per game
  • Pull up field goals made per game
  • Pull up field goal attempts per game
  • Pull up field goal percentage
  • Repeat last three for 3-point shots
  • Pull up effective field goal percentage

It looks like this category would go hand in hand with Drives as to how often a player pulls up from the perimeter versus going all the way to the paint for a bucket.

With the statistics available, we’ll now get to see just how good a player shoots off the dribble from three and how often they take shots from there. It can also help differentiate players who prefer to either get all the way to the basket or shoot a three from those who spread their shots from everywhere on the court. The analytic minds prefer the former, so statistical categories like the Pull Up and Drives help assist what’s already available to look at.

Shooting Efficiency

Want to kill a few birds with one stone or get a few different stats in one click? This is basically putting a bunch of offensive stats together: Drives points and %, catch and shoot points and %, and pull up shot points and %, and effective field goal percentage.

It’s an exciting time for basketball analytics. What was available just over a decade now, like shot locations and shooting percentages from areas on the court, has been expanded into tracking every movement of every player on the court. Who knows what will be available in another decade. Will a stat from SportVU like points created by assist per game become so standard by 2023 that we’ll be curious about a new stat that will trump it?

It feels like this is only the beginning.

About these ads

Tagged: ,

One thought on “First impressions of SportVU

  1. […] We’re now one month into the NBA season, but also one month into sorting through STATS LLC’s SportVU Player Tracking system where stats from hockey assists to a player’s speed can be tracked. (Shameless promotion: I wrote about my first impressions of it here.) […]

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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 1,318 other followers

%d bloggers like this: