The Diff is your weekly WFNY look into the amazing world of sports statistics. For a complete log of articles, click this link. Last week, I looked at the Cleveland Indians offseason with their 2014 salary situation. This week, I’m talking about advanced shooting stats in the NBA.
The NBA is on the verge of something humongous. And it has to do with big, gigantic data. MLB had its signature data moment with the 2003 publication of Michael Lewis’ “Moneyball.” Now an over-used economic reference, “Moneyball” was actually about past events and what the Oakland A’s were already doing with proprietary data. In the NBA, the ground-breaking data just was released en masse in its most complete form. And there’s no telling what the consequences of SportVU could be.
For as good of SportVU primer out there, you’ll want to read Grantland’s Zach Lowe’s article from back in March. Lowe had unprecedented access to the Toronto Raptors analytics department as they were one of 15 test teams using the $100,000 camera system from STATS Inc. last season. This season, the cameras became mandatory in all 30 NBA arenas. And the hyped-up NBA.com/stats became the public home of the data, where fans can track assist opportunities, distance run and all types of statistics never before heard of outside of closed NBA offices.
TIME Magazine’s Sean Gregory had an article the week of the SportVU public unveiling and even had some quotes from Cleveland Cavaliers executive David Griffin. Amid Gregory’s anecdotes of how he enjoyed seeing the distance stats and more, there was this telling quote from Griffin about the new data possibilities: “You can almost come up with a way to quantify anything you can dream of.”
So how can this matter long-term? Over in baseball, one tidbit within “Moneyball” was defensive-independent pitching statistics (DIPS). Beyond The Box Score’s Lewie Pollis wrote recently about how these stats theoretically could be a reason for decreasing offensive statistics. With knowledge of how pitchers and defenses directly interact, teams can self-optimize, lowering overall opportunities for opposing offenses and thus, league-wide scoring. In the context of the new-ness of DIPS theory, here’s what Pollis wrote about how it might have changed the league as we know it.
“As teams gradually start to consider DIPS theory in evaluating players, greater proportions of personnel choices are made in concert with clubs’ comparative advantages. … It is impossible to point to a single date when this line of thought took hold around the league, but the paradigmatic shift in how teams think was undoubtedly recent enough that we could still be working our way to a new equilibrium.”
In the NBA, there are no limits over how big data could affect the game. There’s no way to potentially know. Even last season, ESPN The Magazine’s Jordan Brenner wrote in the “Analytics Issue” about how the Portland Trailblazers utilize 24-year-old Ben Falk to provide occasional stats to head coach Terry Stotts and players. Sports Illustrated’s Pete Thamel also wrote about another basketball stats guru weapon in Butler’s Drew Cannon, who jumped over to the pros in Boston with coaching prodigy Brad Stevens.
For one final thought, here’s what Lowe wrote about the comparison of the SportVU revolution and “Moneyball.” This is what sets this NBA movement apart and what could cause an even more dramatic shift in the game over the next several years:
“But let’s not exaggerate: This isn’t Moneyball, with people at each other’s throats and folks threatening to quit their jobs. It’s not even close to that, actually, and that’s in part because the SportVU data do something most smart NBA people have been doing for a long time: combine video (the ‘eye test’) with advanced statistics.”
Enough with the theoretical talk, on with the data. The rationale for my investigation this week was the struggling Cleveland Cavaliers offense. Last season, the Cavs were a bad team – but a bad team with a terrible defense and fairly mediocre offense.
As I wrote in a summer edition of The Diff about facts we kept forgetting, the Cavs actually had the NBA’s seventh-best offense in a 38-game stretch last year. One of the major reasons for that stretch was the team’s No. 2 ranked turnover ratio.
The Cavs, with thanks to since-departed players Luke Walton, Shaun Livingston and Wayne Ellington, were great at taking care of the ball. They managed to have the NBA’s seventh-most efficient offense over nearly three months despite only ranking No. 21 in efficiency field goal percentage. So the Cavs were fine offensively, but weren’t necessarily all that good at shooting the basketball.
One would think the two points are a bit more correlated than that, but basketball is a complex game. Skip forward to 2013-14 and, after an offseason of practicing virtually no offense, the Cavs have the NBA’s No. 29 offense, No. 28 efficiency goal percentage and No. 16 turnover ratio through eight games. It’s certainly a small sample size and a huge portion of that drop is the play of All-Star Kyrie Irving, but it’s a major concern.
Piggy-backing on the analysis I did a few weeks back related to Kirk Goldsberry’s ShotScore statistic, I wanted to see what SportVU could tell us about the Cavs’ early shooting woes. This new data offers all kinds of new avenues for analysis, so I wanted to see what the average fan – and even statistically-oriented me – might be missing from just the typical box score data.
SportVU shows three new shooting statistics:
Drives – “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.” (Unfortunately, this data on SportVU did not have total field goal attempt numbers, so this was not included for my analysis today.)
Catch and Shoot – “Any jump shot outside of 10 feet where a player possessed the ball for 2 seconds or less and took no dribbles.”
Pull Up Shots – “Any jump shot outside 10 feet where a player took 1 or more dribbles before shooting.”
So my intention was to analyze available SportVU data on a wide lens, providing much-needed context for the new numbers and then looking at the struggling Cavaliers. This took a tremendous amount of data mining since the SportVU data isn’t terribly well organized by team on the stats site. Nonetheless, I persevered and the chart below shows NBA-wide shooting data through games played on Monday, Nov. 11:
Some of these numbers might be surprising, some might not. Again, it’s a fairly limited sample size of just 10,256 NBA team minutes and 17,548 field goal attempts over a two-week span. By the end of one or two seasons, this chart will really portray a more accurate distribution of NBA shooting.
It’s still kind of fascinating to see that pull-up field goal attempts are (as expected) very inefficient. And that catch-and-shoots are 75% of three-point field goal attempts. Those were my first two clear observations. Now, with this context in hand, it’s time for the team-by-team data.
The data about the Cavs shouldn’t be too surprising. They are one of just six teams to be below average in all three shooting types: pull-ups, catch-and-shoot and others. The other five teams are Charlotte, Memphis, Minnesota, Sacramento and Utah. Perhaps by nature of employing shoot-first guards like Kyrie Irving, Jarrett Jack and Dion Waiters, the Cavs have one of the higher pull-up shot attempt percentages in the league. They rank fourth with 35.3% of their field goal attempts coming in the least-efficient variety. That alone doesn’t necessarily mean the Cavs are destined to be a poor offensive team. Heck, two of the three most pull-up happy teams in the league are the Golden State Warriors and Los Angeles Clippers, both off to hot starts. The issue for the Cavs is they aren’t shooting well from this type or any other type. Let’s go now to the Cavs player-by-player data for even more new analysis:
The Cavaliers chart above is sorted by field goal attempts thus far in the 2013-14 season. Begrudgingly, you see a lot of red (i.e. below average) numbers on the chart. Only three players have above-average efficiency field goal percentages overall thus far: CJ Miles, Anderson Varejao and Alonzo Gee.
Kyrie Irving’s struggles have been pointed out a lot in terms of the team’s slow start. It’s fascinating to see how he’s succeeded in pull-ups (over half of his attempts), but not in other field goal types. Perhaps he’s struggling with the adjustment to being an off-ball player. That could happen with the addition of the next guy.
Jarrett Jack, meanwhile, has had the opposite shooting data than Irving: He’s been terrible with pull-ups, which constitute more than three-quarters of his attempts. He has just 32 points on 56 pull-ups, yet has 32 points on his 16 other attempts in the other two types.
Other noteworthy items are the success of CJ Miles and Dion Waiters in catch-and-shoot attempts. These are the players that should be shooting well here, so that’s great to see for the team. An issue of course, is Andrew Bynum’s efficiency in post-ups, not a defined type, but something we likely see in the others category.
Again, this is all new data at our fingertips. Being able to scientifically track every movement and moment on the basketball court provides a smorgasbord of data for folks like me and the executives in NBA front offices. It’ll be fascinating to see how these ratios change throughout the season and how the complete availability of these numbers might be a factor in that change.
Jacob Rosen is a long-time contributor to WaitingForNextYear. He's also a writer online at SportsAnalyticsBlog and Nylon Calculus . An Akron native, Jacob is a current MBA student at the University of Oregon's Warsaw Sports Marketing Center. You can follow him on Twitter @WFNYJacob or e-mail him at udjrosen(at)gmail(dot)com.