Over the past month and a half, basketball analysts have tried in vain to figure out what has been wrong with the Indiana offense. The flaws have been far more subtle than normal. Basketball geniuses like Zach Lowe even decided, after much research, that the Pacers are simply “not playing basketball as well.”
Indeed, not much has seemed to change in Indiana’s offense from the start of the season. Usage rates for each player have remained remarkably consistent. Only Roy Hibbert has seen his shot attempts decline, and even his decrease in looks hasn’t been an outlier of sorts. The only substantial difference is that the entire team, across the board, seemed to have forgotten how to shoot the ball. After shooting 47% as a team in December, the Pacers shot 45% and 42% in February and March.
Still, ardent fans of the Pacers “felt” that something just seemed different. The offense wasn’t flowing the way it should. George Hill was being too passive. Paul George and Lance Stephenson were forcing the issue. Roy Hibbert was being too Todd MacCulluch-y. But was there something more tangible?
After the Pacers awful loss at home to the Hawks two weeks ago, basketball expert Haralabos Voulgaris tweeted that they went through the whole game without setting a proper screen on offense. Since then, several writers at 8pts9seconds have been wondering if there was more truth to that than anyone wanted to admit. Unfortunately, no real stats (at least stats that are available to the fans and the media) exist regarding in-game screens. It’s borderline amazing the amount of information that the new SportVU cameras have provided us with, and yet, there are still holes in the story.
So last night, I decided to watch the game and solely focus on the screens that both teams set throughout the game. Admittedly, this was a tough task, but it did provide us with some interesting information.
A few ground rules:
- Keeping track of every single on-ball and off-ball screen from a basketball game is probably more difficult than you would imagine – I did my best and used my DVR to the best of my ability – these numbers may be slightly off, but they are at least a good estimate.
- Determining whether a screen was “successful” or not is obviously somewhat subjective in nature. My criteria was whether the player working off the screen saw enough daylight to attempt a shot or start an open drive to the basket.
- While dribble hand-offs are just as good as screens, they were not included in last night’s numbers for the completely arbitrary reason that I decided not to record them.
- Attempted screens that were denied by the ball-handler (for instance, if David West came to set a high screen but George Hill were to drive the other way because his defender was “cheating” over the screen) were not included in tonight’s numbers.
- If a player attempted to make one screen, failed, and then immediately turned to set another screen for the same player in the same area, this was only counted as one screen attempt.
- Since oftentimes three and four screens are set on inbounds plays, none of these screens were counted as they didn’t come within the regular flow of the offense.
- Both teams pulled their starters with about six minutes left in the game – stats were taken for garbage time, but were not deemed relevant enough to be included in this game synopsis.
On to the numbers!
A few quick notes to summarize:
- For the game, Indiana successfully executed 19 of 48 screens, with 12 of those screens coming off the ball. Conversely, Atlanta was 39 of 69, with 27 of those coming via on-ball pick-and-roll type variations.
- Atlanta had 90 offensive possessions while Indiana had 87 on the night, meaning that the Pacers averaged 0.54 screens per possession while Atlanta averaged 0.77.
- Atlanta successfully executed 73% of its on-ball screens while Indy could only muster 37%. As anyone that has ever played basketball can tell you, this probably has more to do with the ball-handler than the actual screener, as it’s the ball handler’s job to run his man into the screen. This definitely matches up with the good old eye-test, as Jeff Teague is a wizard with a screener while Lance Stephenson, Paul George, and George Hill all appear to feel more comfortable with more space or even in isolation-type situations.
- Atlanta’s bigs are screen-setting monsters. In the first half, Pero Antic and Paul Millsap connected on 14 of 20 screens while Indiana as a whole was only good for 11 of 23. For the game, Antic, Millsap, and Elton Brand successfully executed 31 total screens, 50% more than the entire Pacers’ roster.
- While the number of screens per quarter for Atlanta stayed pretty consistent, the Pacers set far more screens from different places in the game-changing 3rd quarter. This is what we will focus on for the remainder of this piece.
A simple look at the box score will tell you that the Pacers made their big second half run when Paul George and George Hill started to get into their groove. Indeed, the Georges combined for 18 points on 7 of 10 shooting during that stretch, and Pacers fans were once again delighted to see their point guard become the aggressor.
What was so interesting, though, was the way Indiana came out in the second half. After setting only one screen in the entire first half, Paul George set six screens in the first five minutes of the third quarter, all from the extended left elbow like so:
On five of those six possessions, Indiana scored relatively easy baskets ranging from David West layups to George Hill cuts down the lane. George also typically caught the ball closer to the basket, a step or two inside the 3-pt line, in close enough range to get to the rim with one or two decisive dribbles. In the final four minutes of the quarter, George Hill set four off-ball screens after setting zero for the first 32 minutes of game action. Hill was in position, after one screen, to score a putback off of an offensive rebound, and opened up Paul George for a 3-pointer on another.
It’s unclear whether this shift in strategy was intentional. It’s even more unclear whether these screens from the Pacers’ backcourt were actually opening up new options, or if their increased involvement in the play was simply a psychological boon to their play. Sometimes, the best remedy for a slump is to find something simple that you can do effectively in order to build some confidence. Both George and Hill seemed to pick up their defense in the 3rd quarter as well, and their whole games just fell into place.
In conclusion, it’s probably unwise to try to deduce too much from tonight’s screening statistics. Again, with the unavailability of screen stats in general, it’s unclear whether the Pacers set more or less screens tonight than normal. Further, I’m not even sure how to evaluate screen totals in general because I have no idea what a “good number” of screens would even be. My gut tells me that Atlanta probably sets more screens on a game-to-game basis than the Pacers because of their offensive system. Teams like San Antonio and Dallas, again, would seem to set a lot more screens than others – but I really have got no idea.
Still, the third quarter stats have to mean something. Paul George and George Hill involved themselves into the offense in different ways by setting screens for other guys and then involving themselves in later.