Synergy Sports Technology, data provider for every NBA team and most NCAA teams, is a unique stats source, offering data categorizing every offensive possession by play types. Those play types are actions such as post-ups or isolation possessions.
Here at Nylon Calculus, we’ve used these play types previously to help define offensive roles for NBA players, as well as define team offensive play styles. There are two different versions of this Synergy play type data, the limited amount of data you can see for free on NBA.com/Stats*, and data from Synergy which requires a subscription. Synergy releases a lot of great information, but the explanations of their data aren’t the most comprehensive, especially for what’s available publicly on the NBA’s site.
I’m someone who has looked through hundreds of clips from Synergy’s database and taken lots of mental and written notes about what fits into which category. Based on that knowledge, I’m here to try to explain what’s what. Here are each of those Synergy play types and what they mean:
These are possessions finished by the ball-handler in the pick-and-roll. This includes pull-ups, floaters, and shots at the rim by that player. It also includes possessions where the bal- handler shoots before even dribbling off of the screen, as well as when he denies the ball screen and dribbles away from the pick.
These are surprisingly much lower scoring possessions that you might guess, but the offense generated from the pass outs to cuts, rolls, and spot ups are fairly high scoring. The lower average scoring efficiency of these shots is likely from the number that end up being pull-up mid-range shots.
The public data available on NBA.com only looks at scoring, not passing to others from the pick-and-roll. That data is also available on Synergy, but isn’t part of this overall play type number.
NBA Average PPP: 0.850
NCAA Average PPP: 0.768
Pick-and-roll roll man
These are the slips, rolls, and pops from screeners in the pick-and-roll. This is a tricky top-line stat to make judgements based off of due to the variation that exists within it. When analyzing players, I make an effort to look more at the efficiencies at each of those three specific actions. If a player has mediocre roll man numbers but elite popping data, there’s more value than initially meets the eye.
NBA Average PPP: 1.037
NCAA Average PPP: 0.987
Transition possessions are about the defense not being set, and don’t have anything to do with the time left on the shot clock. That means there’s no time cutoff that makes a possession a halfcourt possession rather than a transition possession. On a more granular level we can look deeper into the role a player had within a transition possession. That can be as a leak-out man, the ball-handler, left/right wing, or a trailer.
NBA Average PPP: 1.103
NCAA Average PPP: 1.040
These possessions are generated by a player running off of a screen, whether it be a pin-down, flare screen, elevator screens, or any other of the plethora of screen variations before they receive the ball. That player catches the ball coming off of a screen and either shoots immediately, dribbles into a pull up, dribbles into a floater, or dribbles and takes a shot at the rim. Occasions where a player curls off of a screen toward the basket are also counted. However, UCLA screens and flex screens do not fall into this category. Those would be logged as cuts.
NBA Average PPP: 0.943
NCAA Average PPP: 0.883
Spot-up possessions are similar to off-screen possessions, but there’s no screen being used before the player catches the ball. Players spotting up don’t need to be stationary, but they can’t be running off of screens before catching the ball. Players just standing in the corner before catching-and-shooting, or guys relocating to the 3-point line or fading to the corner and getting the ball on a kick out are all spotting up.
These possessions aren’t just catching and shoot. They can be catching-and-shooting, but attacking a close-out by dribbling into a pull-up, dribbling into a floater, or driving to the rim are also included.
NBA Average PPP: 0.999
NCAA Average PPP: 0.941
I don’t think I need to explain isolation, but I will say one thing: If one of these other actions occurs and is broken, it may end up being logged as an isolation possession. For example, if a pick-and-roll ball handler dribbles off of the screen and needs to retreat dribble twice, then attacks after that substantial delay, it’s an iso possession. A spot up possession where the player catches then does several jabs or tries to size up his defender is now an isolation possession.
NBA Average PPP: 0.875
NCAA Average PPP: 0.787
Handoffs are the dribble handoffs or flip/pitch plays we’ll see. They may come from the passer being stationary or the passer dribbling at the receiver and then handing the ball off. This is an action that isn’t used much by most teams. The Celtics are one of a few teams that run a lot of dribble handoffs.
NBA Average PPP: 0.916
NCAA Average PPP: 0.843
This category includes backdoor cuts and dump-offs as “basket cuts”. UCLA cuts and flex cuts also fall into this category as “screen cuts”. “Flash cuts” are the third subgroup within the cut category. These include times a player, without a screen, cuts out or toward the ball to receive it (like for a V cut).
NBA Average PPP: 1.246
NCAA Average PPP: 1.121
Putbacks are the tip ins and quick shots after offensive rebounds. Very rarely this will also includes long rebounds that result in a quick shot. Due to most shots being right at the rim, these are generally very high PPP opportunities.
NBA Average PPP: 1.084
NCAA Average PPP: 1.084
These are all of the traditional post-ups we’re accustomed to. This category counts back-to-the-basket and face-up post possessions.
NBA Average PPP: 0.885
NCAA Average PPP: 0.814
This is a potpourri of possessions that don’t fit into any of the other categories. This category includes plays such as:
- Possessions where the ball goes off a leg or is deflected and is picked up by another player and shot.
- Players being fouled in the backcourt
- Illegal screen offensive fouls
- Errant passes out of bounds or that are intercepted from players not currently in one of the other actions (most times they’re just standing at the top of the key), even if it’s to a cutter, player coming off of a screen, etc. On a playing coming off of a screen, this is miscellaneous rather than an off-screen possession because the turnover was on the passer, not the receiver.
- Possessions where the player dribbles into a pull up 3-point shot in the halfcourt.
- Inbounds passes that go directly out of bounds
- Another strange example: A possession where a player had the ball in transition, dribbled into the paint and lost his dribble on a spin move, recovered the ball and threw it out of bounds. It’d be a transition possession but by the time he threw it away the defense had recovered and been set.
NBA Average PPP: 0.530
NCAA Average PPP: 0.543
Full Points Per Possession Chart
Yes, pick-and-roll ball-handler possessions are extremely inefficient. They’re less efficient than isolation! There’s a reason the Warriors and Celtics take less of these shots than just about any team. Those are bad shots. The offense you can generate on rolls/pops/slips, cuts, and spot-up looks generated from the pick-and-roll are productive, which makes the best use of a pick-and-roll a primary action that sets up something else. However, sometimes the ball-handler shooting is the only option available so having an efficient pick-and-roll scorer can help make or break an offense.
Post-ups are also bad possessions relative to the rest of these actions. Obviously having Shaq on the block is a different story, but for most post players, the offense they’re giving you isn’t great compared to what you can get from players utilizing cuts, handoffs, and guys running off of screens. The best use of post-ups is to pass the ball there, run off-ball screens and/or split cuts, try to get shots for the cutters and off-screen shooters there, and after those options don’t work, then attacking in the post becomes Plan B.
Isolation is another area that isn’t efficient. It does generate pass-out opportunities the same way post-ups and pick-and-rolls can. However, these shots generally aren’t as efficient.
Frequently asked questions
- What is a possession? A possession is only logged if it ends with a shot, turnover, or shooting foul. If the Lakers run a post-up and that player kicks the ball out to a spot-up shooter for a shot, that will be a spot-up possession. In Synergy’s granular data, that pass-out will be marked, but the top line possession that’s logged is a spot-up possession in that scenario.
- Do Points Per Possession (PPP) stats include turnovers? Yes, they include turnovers and fouls shots made.
- How many spot-ip possessions are derived from passing out of the pick-and-roll, isolation, and from post-ups? Only around 50 percent of all spot-ups. These possessions come from all over: spot-up shooters passing to other spot-up shooters, off-screen shooters passing to spot-up shooters, cutters passing to… you get the idea.
- Why would an illegal screen be miscellaneous and not count as an off-screen possession? Because the foul was on the screener, who wasn’t the player coming off of a screen. If a player is running off of a pin down and pushes his man off for an offensive foul, that would count as an off-screen possession.
- If you have any questions, @ me on Twitter (@T1m_NBA) and I’ll respond as soon as I can.
Other random Synergy notes
- The Synergy pick-and-roll ball-handler turnover data can be misleading, as we saw with Lonzo Ball. The turnovers from trying to score and trying to pass will count against their pick-and-roll scoring data and not account for their passing possessions. To get a true turnover percentage for pick-and-roll players, you’ll need to take: PnR BH TOs / (PnR BH Possessions + PnR BH Passing Possessions). Those passing possessions are listed on Synergy, but not in the public data on NBA.com.
- Who decides what each play is categorized as? It’s not actually a manual process. Synergy has to process clips from NCAA men’s and women’s games, the NBA, the WNBA, junior college, Olympic teams, and teams overseas as well. Their technology analyzes the video that’s uploaded and logs each possession, and with stunning accuracy. I’ve probably gone through more than a couple thousands clips by this point. In that time I’ve seen less than five that have been categorized inaccurately (and they were all from NBA games from 5+ years ago).
Other tools and data
Utilizing this Synergy play type data I’ve been able to build several other tools and stats that you may find interesting:
- To see any NBA team’s 2016-17 breakdown of their offense by Synergy play type, check out my spreadsheet here. You can see the projected PPP for each team based on that breakdown and their actual PPP
- To see which NBA player your pickup basketball style best matches, click here
- To see which NBA player each draft prospect from this year’s draft matches, click here
- I calculated a Points Over Expectation metric, which shows the points a player creates over an average player would. This is based on Synergy play type data and encompasses offense and defense. To see the offensive, defensive, and overall data for that, click here
- To see the rest of my work, go to my digital portfolio here
*Note: NBA.com/stats currently has Miscellaneous and Putback data reversed