Pages Navigation Menu

Shooting Data from SportVU: The Power of An Assist

US Presswire

US Presswire

A couple of weeks ago I sent out a ‘Deep Tweet’ of the Day:

I have only had two, so I guess I am not that deep. In any case, everybody knows, or at least assumes, that passing the rock is a valuable skill for basketball players. But analytic analysis has often struggled to quantify the value of the one passing stat we have; the assist. Now with the new SportVU data from video tracking cameras in the rafters we a couple of more tools to do so. The video tracking system released at the beginning of the season by the has the ability to capture a nearly infinite amount data depending on the way it’s parsed. One of the categories of data released out of the SportVU data are passing statistics. This post isn’t about those stats.

But, since two of the other categories of data the NBA released in player tracking data were Catch and Shoot data and Pull Up shot data I decided to compare the two in order to get an idea of the difference in degree of difficulty between shooting off the bounce and getting an assist. I downloaded the data through November 17th, from NBA.com for both Catch and Shoot tracking and Pull Up tracking.  Then I aggregated the player data by team.

In addition to the significant difference in expected eFG% between the two shot types I found a couple of other interesting patterns.  There have been slightly more Catch and Shoot attempts than Pull Ups, with about 7,400 Catch and Shoot attempts and just under 7,000 Pull Up shots.  There is a huge difference in the number of threes taken via the Catch and the Pull Up, with 63% of Catch and Shoot attempts being three points and only 19% of Pull Up shots being three pointers. Below is the table for Catch and Shoot totals by team:

Team Catch & Shoot Total PTS Total C & S Att PP C&S Shot 3P eFG% 2P eFG% Pct C&S for 3
ATL 330 299 1.10 55.8% 53.5% 67%
BKN 222 219 1.01 54.3% 44.7% 58%
BOS 218 242 0.90 46.9% 42.4% 60%
CHA 205 246 0.83 47.4% 34.5% 55%
CHI 176 195 0.90 52.9% 36.6% 52%
CLE 301 290 1.04 57% 46.2% 54%
DAL 318 268 1.19 65.6% 42.9% 72%
DEN 219 209 1.05 57% 45% 62%
DET 140 176 0.80 41.6% 32.5% 78%
GSW 307 227 1.35 76% 52.1% 69%
HOU 276 265 1.04 53.7% 42.9% 86%
IND 297 246 1.21 62.2% 57.6% 60%
LAC 281 260 1.08 57.7% 48.4% 65%
LAL 407 370 1.10 67% 37.2% 63%
MEM 243 262 0.93 48.8% 44.4% 48%
MIA 311 241 1.29 71.4% 49.3% 69%
MIL 182 177 1.03 63.9% 33.3% 61%
MIN 267 267 1.00 55.3% 40.4% 67%
NOP 281 250 1.12 75% 44.1% 39%
NYK 251 253 0.99 51.8% 46.3% 68%
OKC 153 188 0.81 38.5% 44% 60%
ORL 262 260 1.01 59.6% 40.3% 52%
PHI 248 257 0.96 55% 34.8% 66%
PHX 217 214 1.01 52.5% 45.8% 73%
POR 318 280 1.14 66.1% 42.9% 60%
SAC 227 240 0.95 49% 45.1% 63%
SAS 252 266 0.95 56.8% 30.9% 64%
TOR 178 194 0.92 48.9% 39% 70%
UTA 209 261 0.80 44.3% 32.7% 61%
WAS 300 272 1.10 63.1% 35.9% 72%
Grand Total 7,596 7,394 1.03 56.9% 42.3% 63%

Portland and Miami so far have been the kings of Catch and Shoot efficiency, or lucky. While Utah and Detroit have been the least efficient off of the Catch and Shoot, with Oklahoma City surprisingly close behind. Overall the Catch and Shoot three has had an eFG% of 56.9% and the two point Catch and Shoot has had an eFG% of 42.3%. Finally here is the same data for Pull Up shots:

Team Pull Up Shots Total PTS Pull Up Total Att PP Pull Up Shot Pull Up 3P eFG% Pull Up 2 eFG% Pct of Pull Ups for 3
ATL 143 179 0.8 46.60% 38.7% 16%
BKN 149 228 0.65 33.90% 33.5% 14%
BOS 186 227 0.82 51.40% 39.6% 15%
CHA 138 209 0.66 36.00% 32.6% 12%
CHI 79 138 0.57 32.60% 27.8% 17%
CLE 239 322 0.74 44.10% 35.8% 16%
DAL 195 251 0.78 35.90% 40% 18%
DEN 165 216 0.76 56.70% 33.9% 21%
DET 178 233 0.76 48.00% 35% 21%
GSW 274 296 0.93 55.60% 42.8% 27%
HOU 112 176 0.64 33.30% 31.1% 31%
IND 164 224 0.73 35.70% 36.8% 19%
LAC 253 305 0.83 42.50% 41.2% 22%
LAL 205 289 0.71 49.10% 31.6% 20%
MEM 112 153 0.73 37.50% 36.5% 10%
MIA 153 179 0.85 40.40% 43.1% 15%
MIL 191 235 0.81 62.00% 36% 20%
MIN 194 267 0.73 48.50% 32.2% 24%
NOP 172 220 0.78 39.50% 39% 17%
NYK 206 264 0.78 47.10% 38% 19%
OKC 242 247 0.98 60.00% 46.9% 22%
ORL 165 195 0.85 48.80% 40.6% 21%
PHI 186 249 0.75 41.40% 36.6% 23%
PHX 186 210 0.89 58.70% 41.5% 22%
POR 294 321 0.92 59.10% 42.4% 21%
SAC 163 202 0.81 50.00% 38.2% 22%
SAS 176 193 0.91 55.60% 44% 14%
TOR 212 284 0.75 51.70% 33.6% 20%
UTA 145 221 0.66 30.90% 32.6% 15%
WAS 127 229 0.55 30.90% 27.7% 15%
Grand Total 5,404 6,962 0.78 46.80% 37.1% 19%

Really I could have set the conditional formatting to make every team’s Points Per Pull Up Shot red except for OKC’s, who score more efficiently on Pull Ups.  In fact OKC is the only team shooting more efficiently on the Pull Up shot than the Catch and Shoot with a .17 higher PPS on Pull Ups.  Meanwhile nobody on Washington should even think about pulling up for a jump shot with .55 points per shot. Here is the differential for each team:

Team PP Pull Up Shot PP C&S Shot Difference
ATL 0.8 1.10 0.30
BKN 0.65 1.01 0.36
BOS 0.82 0.90 0.08
CHA 0.66 0.83 0.17
CHI 0.57 0.90 0.33
CLE 0.74 1.04 0.30
DAL 0.78 1.19 0.41
DEN 0.76 1.05 0.29
DET 0.76 0.80 0.04
GSW 0.93 1.35 0.42
HOU 0.64 1.04 0.40
IND 0.73 1.21 0.48
LAC 0.83 1.08 0.25
LAL 0.71 1.10 0.39
MEM 0.73 0.93 0.20
MIA 0.85 1.29 0.44
MIL 0.81 1.03 0.22
MIN 0.73 1.00 0.27
NOP 0.78 1.12 0.34
NYK 0.78 0.99 0.21
OKC 0.98 0.81 -0.17
ORL 0.85 1.01 0.16
PHI 0.75 0.96 0.21
PHX 0.89 1.01 0.12
POR 0.92 1.14 0.22
SAC 0.81 0.95 0.14
SAS 0.91 0.95 0.04
TOR 0.75 0.92 0.17
UTA 0.66 0.80 0.14
WAS 0.55 1.10 0.55
Grand Total 0.78 1.03 0.25

I suspect the team data has some noise in the differentials at this point in the season, but in addition to OKC and Washington, Miami and Indiana look like teams that you want to force into pull up shots. Just keep LeBron James out of the paint and stay home on the shooters, how hard can that be? In terms of Effective Field Goal percentage for Pull Ups threes have an eFG% of 46.8% and two point Pull Ups only have an eFG% of 37.1%. So, bringing it all together, here are the differences on assisted Catch and Shoot eFG% and Pull Ups.

Catch and Shoot Pull Up  Shot Making Diff
Three Pointers 56.9% 46.8% 21.6%
Two Points 42.3% 37.1% 14.0%

And, of course, assists are potentially getting players shots in more efficient locations than  they would be able to off the dribble considering how few unassisted threes are shot in the NBA. In a future post I plan to look some of the individual players’ shooting stats in the Catch and Shoot and Pull Ups.

  • Larry

    Really interesting post. Is there anyway to compare with defenders close? Seems like you would need more data than the NBA provides. Possibly higher percentage guarded for pull ups. Would be interesting to compare guarded pull ups vs. guarded catch and shoot for end of shot clock decision making.

    • Andrew Johnson

      Good question, I am pretty sure the data could be pulled out of the tracking data, but you’re right the NBA doesn’t yet provide that.

      Likely catch and shoot situations are generating more open looks. But also harder mechanically to balance and align shot pulling up off the dribble.

%d bloggers like this: