April 17, 2016

Are the 2015-16 Warriors The Best NBA Team Ever?

Jack Beckwith + Dillon Chen

The NBA’s 2015-16 regular season concluded last Wednesday night with a whole lot of drama. We had Kobe Bryant closing out what has been an illustrious career in the most Kobe way possible – 60 points on enough field goal attempts to make even Wilt Chamberlain blush. But we digress.

The story of that night, and seemingly every night this NBA season, was the Golden State Warriors. With a 125-104 drubbing of the Memphis Grizzlies, Curry & Co. surpassed the infamous 1995-96 Bulls for most wins in a regular season. It was just one more record for a team that’s already broke an absurd amount of them this season. We’ve certainly heard about Steph Curry’s individual greatness over the course of the year, and how he is literally breaking the game of basketball.

But how good are this year’s Warriors relative to the NBA’s legendary teams? #NBATwitter has been filled with arguments about who would win in a hypothetical seven-game series between Air Jordan’s Bulls and the current iteration of the Warriors. And while the conversation’s been pretty entertaining, it hasn’t really settled any questions. In the end, it comes down to the conundrum that is comparing teams across different eras. After all, there’s no telling what MJ’s career PPG would be if he had matched up with James Harden for a few nights every season.

For us here at the DataFace, we couldn’t let this debate just go in circles. So we decided to take a closer look at five-decades’ worth of NBA team data, stretching back to the NBA’s inaugural season in 1949-50. Our goal was to remove any of the punditry from the discussion (#factsonly) and let the data show where the Warriors stand compared to other great teams in NBA history.

The Warriors are Indisputably the Best Shooting Team of the 3-Point Era

We’re starting off with the basics here. In recent years, the NBA has become a 3-point shooters’ league and this year’s Warriors are perhaps the best manifestation of that trend.

They’ve put up 31.6 3-point attempts per game this season, second most by a team since the 3-point line was instituted in the 1979-80 season. But it’s not just the rate at which the Warriors are jacking up 3-pointers that’s impressive. <em>It’s the rate at which they’re making them</em>. Check this out, a scatterplot of 2-point % vs. 3-point % for every team since the introduction of the 3-point shot.

It’s clear that the Warriors are hoisting up so many 3-pointers for good reason. They now have the second-best 3-point shooting percentage of any team in the modern era at 41.6%. They trail only the 1996-97 Hornets, who shot a ridiculous 42.8% from downtown (and coincidentally had sharp-shooting Dell Curry, Steph’s dad, on its roster). The Warriors’ 3-point field goal percentage is even more incredible considering the 1996-97 season was the last year that the NBA experimented with a uniform 22-foot 3-point line1. One can only imagine imagine the type of numbers the Splash Brothers would have put up if the 3-point line remained at that distance.

No team this season even came close to the Warriors in 3-point shooting; the San Antonio Spurs, the league’s second-best 3-point shooting team, were way back at 38.1% from 3-point range. What’s more, the Warriors outpaced the league average by an ungodly 6.5%. That’s the largest disparity between the league leader and league average since those ‘96-’97 Hornets.

But don’t think the Warriors are a one-trick, 3-point shooting pony. Of the 1014 team seasons logged since 1979-80, the Warriors are 17th all-time in 2-point shooting percentage. In other words, they’re in the top 2% of teams in terms of accuracy inside the arc. That’s particularly impressive given the fact that they don’t have a dominant big man to dump it down to in the low post (sorry Bogut).

To get an overall gauge of the Warriors’ shooting proficiency, we use a metric called effective field goal percentage (eFG%). It’s essentially a weighted average of a team’s shooting percentage from 2-point and 3-point range, adjusting for the fact that 3-pointers are worth 1.5 times a 2-pointer. The idea here is that 3-pointers should be weighted in proportion to the additional benefit from making them. It’s calculated as follows2: (FG + 0.5*3PT)/FGA

This season’s Warriors just surpassed the 2013-14 Miami Heat--a season in which Lebron was living on the low block--by a full 0.9% in eFG%. In the realm of shooting percentage, that’s a massive leap. If you thought the Warriors were a good shooting team last year, consider that this squad has increased its eFG% from the 2014-2015 regular season by a full 2.3%.

In comparison to the 72-10 Bulls, we find that the Warriors are the better shooting team. Those Bulls, led by Michael Jordan, are tied for 85th all-time in terms of eFG%. While this still ranks well within the top 10% of all teams, it’s nothing spectacular. We tried controlling for the league’s defensive proficiency in a given season, reasoning that part of the discrepancy might originate from the fact that the Bulls were facing tougher defenses in the ‘90s. If anything, we found the defense played across the league in that 1995-96 season wasn’t even as good as the defense the Warriors faced this year3. On pure shooting alone, it’s this year’s Warriors over those ‘95-96 Bulls in a landslide.

The Warriors are More Efficient Than the ‘95-’96 Bulls

Of course, an NBA team’s greatness is composed of more than just its ability to put the ball in the basket. It has to be able to pass, rebound, and play a little defense too. To measure the overall strength of a team’s players, we created a metric called weight team efficiency (WTE). It was calculated using Player Efficiency Ratings (PER)4, a statistic developed by basketball stats guru John Hollinger. PER is a measure of a player’s per minute contribution while on the floor in terms of stats like field goals, rebounds, steals, and blocks. It’s calibrated so that the average player PER is 15 across the league for a given season. To compute WTE, we summed the PER for every player on a given team, weighted by the number of minutes they played during the season. In essence, WTE provides a look at how much a team’s average player is producing while on the floor. Thus, the higher the WTE, the more skilled the team.

Just like PER, the average team in any given season scores a 15 in weighted team efficiency. In other words, teams that score above 15 in WTE have an above average amount of talent on their rosters as measured by PER. Only 18 teams in history have logged a WTE above 17, two of which are this season’s Warriors (17.39) and this season’s Spurs (17.10). Teams that top 17 in WTE tend to have serious postseason success; seven of the 16 teams that did so in previous seasons ended up winning an NBA title. Ten of the 16 went to the NBA Finals. With this in mind, we believe our “weighted team efficiency” metric is a decent approximation of historical greatness for an NBA team.

So where do the 2015-16 Warriors rank according to WTE? They’re in the very upper echelon, standing 5th all-time among teams dating back to the 1955-56 season. Interestingly enough, the only teams they trail are two Steve Nash-led squads from the mid-2000s, and two MJ-led Bulls teams from the ‘90s. Note, however, that neither of the Bulls teams that outperformed the Warriors in WTE are the infamous ‘95-’96 Bulls squad. Indeed, our WTE metric gives a slight edge to the ‘96-’97 Bulls and the ‘90-’91 Bulls over that ‘95-’96 team.

When we compare the ‘15-’16 Warriors to the ‘95-’96 Bulls head-to-head in terms of WTE, we see two very different teams.

The Bulls had three players in the ‘95-’96 season--Michael Jordan, Scottie Pippen, and Toni Kukoc--who all exceeded 20 in PER. Only Steph Curry was able to accomplish such a feat for the Warriors this season. However, among players that averaged at least 8 minutes per game, the Warriors had six players (four starters and two bench players) with a PER above the league average of 15. The Bulls, on the other hand, had only four players above 15, with Steve Kerr eeking out a PER of 15.2. Said another way, 11 of the Bulls’ 15 players that ‘95-’96 season were below average players according to PER. Though their WTE scores are close, this year’s Warriors clearly received more balanced contribution from across the roster than the top heavy Bulls.

Another difference we find is that Steph Curry played substantially fewer minutes this year than Michael Jordan, who was the league MVP in ‘95-’96. In fact, in ‘15-’16 Curry5 has played the third fewest minutes/game (34.2) among all NBA MVPs going back to 1950. He trails only Bill Walton in ‘77-’78 (33.3) and...himself in ‘14-’15 (32.7). In contrast, Michael Jordan played an average of 37.7 minutes/game during the ‘95-’96 campaign, 3.5 minutes more than Curry this year. The fewest minutes/game Jordan ever played in a MVP season was 37.0 in ‘90-’91.

This means, in essence, that the Warriors WTE is deflated by the fact that their best player spends an abnormally low number of minutes on the floor each game. It's not as if Curry is an aging star who needs the extra rest; Jordan was four years older than Curry is now in that '95-'96 season. Curry’s limited action this season is likely a result of the Warriors’ ability to jump out to commanding leads early in games, allowing them to rest their star for much of the 4th quarter.

The Warriors Build Up Leads Very Early in Games

In order to test this hypothesis, we looked at a team’s average scoring margin, or the amount by which they typically outscore their opponent. The '95-'96 Bulls and '15-'16 Warriors both rank in the top 3 all-time in average margin of victory (i.e. scoring margin at the end of the game). The '15-'16 Warriors outscored opponents by 10.76 points per game, just slightly ahead of the '15-'16 Spurs for tops in the league this season. The Bulls, on the other hand, outscored opponents by 12.24 points per game, nearly 1.5 points more than the Warriors and best all-time. Both teams were clearly a dominant force in their respective seasons; over the entire course of the game, the '95-'96 Bulls more so.

Yet, when we dig deeper into the scoring margin at the end of each quarter, an interesting trend emerges. The Warriors average scoring margin at the end of the 1st quarter, 2nd quarter, and 3rd quarter is actually larger than that of the '95-'96 Bulls. It’s only the Bulls’ high scoring margin in the 4th quarter that propels their average margin of victory above that of this season's Warriors.

This validates our hypothesis surrounding Curry’s low minutes: on average, the ‘15-’16 Warriors accumulated a more substantial lead earlier in games than the ‘95-’96 Bulls did. This allowed them to rest Curry and other key players for longer in the 4th quarter. We would argue that, given that the 4th quarter often amounted to “garbage minutes” for each team, average scoring margin heading into the 4th quarter is a better benchmark for dominance than scoring margin over the entire game. And by this measure, the Warriors were more dominant than the ‘95-’96 Bulls.

Conclusion

Obviously, this post can’t settle the debate about the Warriors’ place in history once and for all. Given the overwhelming number of Jordan stans and the existence of Twitter, we’re not sure it ever will be. At the very least, however, we’ve shown that the Warriors are pretty darn good. They’re unparalleled in shooting ability; they’re among the most efficient teams in history; and they’re arguably more dominant than the ‘95-’96 Bulls were. Still, the legacy of any team is ultimately cemented by its playoff performances and its number of rings. As the playoffs get under way, we’ll see if the Warriors can convert their impressive statistical achievements into back-to-back championships.


Footnotes:

  1. In 1996-97, the NBA moved the 3-point line up in hopes of counteracting a decrease in scoring across the league. Pretty funny to think about now given the amount offense we see in the league today. ↩︎
  2.  FG = Field Goals Made, 3PT = 3-Pointers Made, FGA = Field Goal Attempts ↩︎
  3.  To obtain a proxy for the league’s defensive proficiency in a given year, we created a metric called Weighted Defensive Ability (WAB) derived from Dean Oliver’s "Four Factors" which include Defensive eFG%, Defensive Turnover%, Defensive Rebound%, and Defensive Free Throw Rate (Opponent FT/Opponent FGA). For each season we found an average of each of the four statistics and then assigned weights of 40%, 25%, 20%, and 15% to Defensive eFG%, Defensive Turnover%, Defensive Rebound%, and Defensive Free Throw Rate respectively (as prescribed by Dean Oliver). To derive a final value for WAB we subtracted out the sum of Weighted Defensive Turnover% and Weighted Defensive Rebound% from the sum of Weighted Defensive eFG% and Weighted Defensive Free Throw Rate. A low WAB translates into a better overall level of defense in the NBA for that year. The WAB for the 1995-1996 season was 0.061 compared to 0.047 in the 2015-2016 season, meaning the overall level of defense was better in 2015-2016. ↩︎
  4. Please see Basketball-Reference for an explanation on how PER is calculated. ↩︎
  5. We are taking a minor leap of faith by assuming that Steph Curry will be the league MVP this year. ↩︎