Drives in SportVU: Where Evan Turner is Not Looking For the Pass
Back in the days of yore before SportVU video tracking data was available to fans, like September, I was doing a series on shot creation and getting to the rim. That series, like most of my others, was done with an eye toward creating a sort of Expected Value model for all sorts of different actions on the court.
Now with SportVU data available that vague aspiration is both closer at hand and further away. First, the league has only released a small portion of the data being collected by SportVU. Opting, I suspect, for the bits they think will be of the most general interest not the most analytic benefit for bloggers.
Still it is new data and, generally, speaking more data is a good thing. But, new data presents the analytical challenge of re-sorting through the wheat and the chaff, to find what is actually useful and how it is useful.
Not too surprisingly, my pre-SportVU look, using the suddenly defunct HoopData, found that players tended to fall into two groups in terms of getting shots at the rim — cutters and drivers. The drivers tended to be ‘Shot Creators’ overall in terms of getting unassisted two point shots. There was not, however, a big difference in the number of attempts at the rim based on the two groups.
The new SportVU data brings the Drive part of the equation into better focus. The NBA defines drives as any touch that starts at least twenty feet from the basket and is dribbled within 10 feet of the basket, excluding fast breaks. As of this morning there have been 4,315 drives in the NBA this year. One of the more interesting stats tracked is the inclusion of ‘Team points per drive,’ which includes points coming from assists either in interior passing or kick outs. (Not sure if it includes Kobe assists, which is actually important as I have estimated that the offensive team is 8% to 10% more likely to get the rebound on a shot at the rim than attempts from further out). It also appears to include points from free throws generated by drives (I am inferring this from all if the players with one drive and one point),
Below are some of the aggregates for Drives, via NBA.com:
||Total Player PTS on Drives
||Average Player PTS on Drives
||Average Team PTS on Drives
||Player Pt Pct
Interesting bits being that, Driving is only an efficient play because of the ability to get teammates involved and so far this year only 59% of the points coming from Drives to the hoop are scored by the Driving player.
Even more interesting is the variation between players that have accumulated a decent number of drives so far, where we can start to see their tendencies. Using 30 drives as my minimum, there are 51 that meet that number. Here are a couple of the standouts:
Not looking to pass: Evan Turner with 85.1% of the points from his drives belonging to Evan Turner, get that money Evan.
Not looking to score: Deron Williams with 19.8% of the points from his drives going to his teammates. (Also not that able to score with a field goal percentage of 27.3%)
Not Able to Score: Deron Williams generating only 0.194 points per drive for himself. Play pass until he proves otherwise folks.
Not Able to be Stopped: Evan Turner with 1.26 points per drive, you deserve that money Evan!
Just Stop: Norris Cole, whose drives are generating 0.75 points per drive for the Heat so far this year.
More, please: Kyle Lowry, generating 1.57 points per drive for his team. Also, beats the opportunity cost is low on that team.
Most Likely to Drive: Monta Ball, baby, with 11.4 edging out Jeremy Lin (Get out of the way Omer and Dwight).
And here’s the leaderboard to date based on team points per drive:
||FG% on Drives
||PTS Per 48 Min on Drives
||Player Pts Per Drive
||Team Pts Per Drive
|Kyle Lowry (TOR)
|Evan Turner (PHI)
|Will Bynum (DET)
|Ty Lawson (DEN)
|Chandler Parsons (HOU)
|Paul George (IND)
|Brandon Jennings (DET)
|Eric Gordon (NOP)
|James Harden (HOU)
|Jeff Green (BOS)
|Gordon Hayward (UTA)
Needless to say, these are smaller sample sizes, leads will change. It’s also a little early to tell what these stats mean in terms of added understanding of the game. At some point we will be able to get a better read on who is successful at using drives in their offense and how much that is aiding their teams.