Nylon Calculus: Could the Golden State Warriors be the most efficient offense ever?
By Matt Femrite
Whether you looked at NBA.com, Basketball-Reference, or Nylon Calculus, the Golden State Warriors led the league in offensive efficiency last season. The question for the 2016-17 season isn’t whether they’ll lead the league again — that seems obvious given the addition of Kevin Durant — it’s whether or not they will be the greatest offense in league history.
To weight the likelihood of the Warriors becoming the most efficient offense ever, I looked at where last season’s squad left off. Using Basketball-Reference’s formula that calculates offensive rating for teams since 1974, the 2015-16 Warriors scored 114.5 points per 100 possessions. That ranked 12th out of 1,133 teams, but simply looking at raw offensive rating props up squads from the mid-1980s to mid-1990s, when efficiency was higher than today, and penalizes the most prolific ones from the late-1990s to early-2000s, when offense was largely stagnant except for a few teams ahead of their time.
A simple way to adjust for league averages during different eras is to take a team’s efficiency and either subtract or divide by what the league average was for that season. For this post, I chose the latter method, updating what Andrew Lynch created in 2014: ORtg+ and DRtg+. If we divide the Warriors’ 114.5 offensive rating with last season’s average of 106.4, we get an ORtg+ of 107.61.
As we see in the table below, this bumped them up from 12th to third but they still fall between the most devastating teams led by Steve Nash.
Had the Warriors scored 114.5 points per 100 possessions in a scoring environment similar to 1998 (105.0 average offensive rating) or the lockout-shortened 2012 season (104.6), they would’ve went down as the most efficient offense ever by ORtg+. To surpass the 2004 Mavericks under last season’s average, they would’ve had to score 116 points per 100 possessions. That’s never happened before, but 1.5 more points per 100 possessions wouldn’t be the most ridiculous bump to expect when looking at the current roster.
Below is a breakdown of the raw offensive rating the Warriors would need next season for an ORtg+ of 109, depending on the league average. For example, if we assume league-average offensive efficiency next season goes up to 107 for the first time since the 2011 lockout, the Warriors would have to score 116.6 points per 100 possessions to get to that 109 ORtg+ needed to pass the 2004 Mavericks.
Golden State has the firepower to build off of last season but a lot needs to go right, just like when they won 73 games and got the finest individual season ever by Offensive Box Plus-Minus from Stephen Curry. If we take some offensive statistics from the 10 most efficient squads and adjust for league averages, like above with ORtg+, we can see which ingredients the Warriors had last season and should already possess, and where they could improve compared to other powerhouses.
The table below looks at overall shooting efficiency, trips to the free throw line, shooting from beyond the arc, how often team’s shot from three, how often they generated second chance opportunities, and how well they took care of the ball:
By eFG+, last season’s Warriors were the best shooting team of the last 45 seasons, recording the best mark since the 1970-71 Bucks (113.4) and 1948-49 Rochester Royals (113.8). Golden State’s raw effective field goal percentage of 56.3 percent was the highest ever. With Curry, Klay Thompson, and Draymond Green on the floor at the same time, their percentage jumped to 60.5 in a 1,971 minute sample, according to NBAwowy, and it was 65.3 when plugging Andre Iguodala into that combo for 551 minutes. In 127 minutes with their full death lineup featuring Harrison Barnes, their effective field goal percentage was 72.3. However, the league also experienced its highest average at 50.2.
A big reason for those marks by Golden State was their shooting range. They were the best three-point shooting team from the modern-day arc on significant volume, which shouldn’t change with Durant on board. Using a 1,000 three-point attempt filter that included 536 teams, only the 1996-97 Charlotte Hornets, who had Glen Rice and Dell Curry teeing off from a shortened line, were more accurate on their three-pointers. Below are the top 10 teams by 3P%+. The 2015 Warriors also made an appearance, ranking seventh in 3P%+ and 34th in ORtg+.
Before acquiring Durant, Golden State was already historically great at the most important skill to a great offense. When charting ORtg+ with eFG+, the r^2 was .64, meaning the two measures often go hand-in-hand. We’d have to go all the way down to the 79th-ranked team by ORtg+ to find the first below-average shooting team, the 2002-03 Warriors, who made up for it with the fourth-highest ORB+ since 1974 and 85th in FTr+ since 1947.
The Warriors were iffy in those two stats last season. But neither should be a big deal before the playoffs, and their most destructive lineup combos were better than their overall rate. As a team, they grabbed 23.5 percent of available offensive rebounds, 19th in the league, but that went up to 24.5 with Curry, Thompson, Iguodala, and Draymond Green all on the floor and 27 percent with their full death lineup. They could receive contributions from Zaza Pachulia (13.7 OREB% last season), Kevon Looney, and continue capitalizing on defenses scrambling before shot attempts.
Read More: Nylon Calculus: Best and worst contracts for the 2016-17 NBA season
Regardless, the trade of a player who can crash the boards for an extra one to space the floor is fine. When charting ORtg+ and ORB+, there is no correlation between the two since 1974, and the results were the same when looking at just the last five seasons, last decade, and looking at only the dark ages of offense from 1998 to 2004. However, there was a notable correlation during the ‘80s and early-90s, when offensive rebounding percentages were around 33 percent, compared to 23 percent last season. Depending on what years from the ‘80s were charted, the r^2 went as high as .205, but since then second-chance opportunities haven’t been a necessity for the most efficient offenses.
Neither have free throw attempts, though the Utah Jazz under Jerry Sloan were one exception (as mentioned here). Golden State ranked 25th in free throw rate in each of the last two seasons, but Durant and Pachulia should help, especially the former in the postseason. Looking at foul drawn rates, both drew two to three times as many team fouls per 36 minutes as Barnes and Bogut last year and they’re reliable free throw shooters.
Below is a table comparing the foul drawing (excluding offensive fouls) of Warriors acquired this summer who played last season, and those who were traded or signed elsewhere.
The real nitpicking of the Warriors’ offense starts with their turnovers, an underrated part of why the early-2000s Mavericks were so good. With Nash, Dirk Nowitzki, and Michael Finley carrying the playmaking load, the 2002-03 and 2003-04 squads had the two lowest turnover rates since 1974, either by TOV%+ or just raw TOV%. Key players like Nick Van Exel and Antawn Jamison also had seasons of high-scoring, low-turnover rates both with Dallas and elsewhere.
The Warriors are not the Mavericks. Their TOV%+ last season was the highest mark among the top 10 offenses and was only the eighth time a top-100 team coughed the ball up at that rate or higher (OKC accounted for three of those teams). Their worst enemy is Fancy Play Syndrome. On one hand, their passing is part of why they’ve been so explosive the last two seasons, using it to hunt for a great shot and/or a knockout blow. On the other hand, the range of passes multiple players could throw led to several unforced turnovers.
Durant should ease the playmaking load on Curry and Green, either with the ball or using his gravity without it, and it helps that Klay Thompson doesn’t think twice before launching threes. They also signed David West, who’s provided efficient scoring with low turnover rates throughout his career.
It’s a long season, though, and the margin of error for the Warriors is wide. There will be turnovers from sheer boredom and there will be a few in garbage time, which is another obstacle. With garbage time and rest could come less time for Curry, Thompson, Iguodala, Durant, and Green to prop up the team’s overall efficiency.
Below is a look at how the 2015-16 Warriors compared to the other top nine offenses in minute distribution for their top 12 players, formatted by each row.
The other teams largely came from an era with more minutes given to their core, though this chart shows that the Seven Seconds or Less Suns took this to the extreme, a decent reason why the Houston Rockets could also finish very well in offensive efficiency under Mike D’Antoni. Meanwhile, Steve Kerr had a longer leash for fringy-rotation players, in part because of injuries and blowouts, whether it’s because of their defense or their scoring barrages. Now, the Warriors need to find replacements for LeAndro Barbosa, Brandon Rush, and Festus Ezeli. That’s not the worst problem to have given what’s at the top of the roster, but bench-heavy lineups could lead to scoring droughts.
The season itself is a beast, giving teams enough time to figure themselves out but it leaves room for unfortunate luck. For now, the Golden State Warriors have a shot at another massive win total, but keep tabs on their offensive efficiency. They’re coming off the greatest shooting season of the modern NBA and the third-best total offense by ORtg+. With Durant in the mix, the next six months could produce the most efficient team in league history.