The best NBA team of the 2010s? One stands above the rest
The ’98 Yankees. The ’85 Bears. The ’96 Bulls. ’09 Barcelona. What do these clubs have in common? When future generations look back on this era of sport, these teams will be at the forefront. However, if I had to choose a team that dominated their competition to a level that has never been seen in the history of sport, surprisingly it wouldn’t be one of these teams. It wouldn’t even be a team that competed in the 20th century. And it wouldn’t even be a team from the early part of the 21st century. It would be a basketball team from the 2010s. Yes, a team from this era of basketball. With 30 teams. And more three-pointers. More analytics. More superstars. More competition. Despite all this, they were so much better than everyone else to the point where it was no longer a competition. Who was it? Let’s dive in.
My Methodology boils down to one key concept: How can I optimize several metrics to create one centralized, compressed number that accurately values how well a team performed? I started looking for statistics that would encompass every aspect of what makes a good basketball team, and I soon found my answer: Dean Oliver’s Four Factors. His Four Factors are a group of 4 stats on both offense and defense (8 in total) that cover the essential aspects of a basketball team. They are as follows:
Effective Field Goal Percentage (OeFG%): (FGM+0.5*TPM)/(FGA)
Defensive Effective Field Goal Percentage (DeFG%): Opponent’s eFG%
Offensive Turnover Percentage (OTOV%): (Turnovers/Possessions)
Defensive Turnover Percentage (DTOV%): Opponent’s TOV%
Offensive Rebound Percentage (ORB%): (Offensive REBs/Offensive Rebound Opportunities)
Defensive Rebound Percentage (DRB%): (Defensive REBs/Defensive Rebound Opportunities)
Field Goals Attempted per Free Throws Attempted (FTA/FGA): (FTA/FGA)
Defensive Field Goal Attempts per Free Throws Attempts (DFTA/FGA): Opponent’s FTA/FGA
The first pair of stats is essentially a measure of points from the field per possession (it’s that number divided by 2). This helps to determine how efficiently a team can score points. The defensive version of the stat does something similar, but it values how efficiently a team can stop its opponents from scoring points. Ball security is critical to performing in the NBA, and TOV% allows the calculation of how well a team protects the ball on offense. On the other hand, DTOV% calculates how effectively a team forces turnovers on the defensive end. ORB% and DRB% are pretty self-explanatory; they measure the rebounding ability of a team on both ends of the floor. Finally, FTA/FGA measures the ability of a team to get to the line. Free throws are as close to a sure thing as there is in basketball, and so teams that can get to the line the most put themselves at a massive advantage. Similarly, DFGA/FTA measures the ability of a team to stop its opponent from getting to the line and hence minimizes the number of “free points” a team gives away.
Oliver weighted these stats as follows:
eFG%: 40%
TOV%: 25%
REB%: 20%
FTA/FGA: 15%
These proportions make sense, but since I am using two stats under each category, the value of each stat needs to be divided by two. Therefore, the final weights are as follows:
OeFG%: 20%
DeFG%: 20%
OTOV%: 12.5%
DTOV%: 12.5%
ORB%: 10%
DRB%: 10%
FTA/FGA: 7.5%
DFTA/FGA: 7.5%
*Note: For some stats, (DeFG%, OTOV%, DFTA/FGA), a lower number indicates higher performance. Therefore, to make everything consistent, I created a new version of each of the three stats as follows: (100%-PreviousPercentage). This standardized the three stats such that a higher number indicates higher performance.*
Using the above percentages, we can create a new all-encompassing statistic that I will call the “team score.” The formula for the team score will be a weighted average, calculated as follows:
(0.2*OeFG+0.2*DeFG+0.125*OTOV%+0.125*DTOV%+0.1*ORB%+0.1*DRB%+0.075*FTA/FGA+0.075*DFTA/FGA).
Although this statistic takes into account individual stats that create a successful basketball team, it does not consider the most indicative measure of a team’s success: winning.
Taking Wins at face value would not be an adequate way to measure a team’s success. Instead, we will use a calculation known as Pythagorean Wins, which calculates how many games a team should have won, and estimates how “lucky” a team got during a season. The Formula is as follows: GP*(((Pts)^x)/((Pts)^x+(Opp. Pts)^x)). It takes the points a team scores, increases it to a certain power, and then divides by Points to that same power plus points allowed to that power. The exponent used for basketball is estimated to be right around 15, as there is much less chance and randomness involved in the outcome of a basketball game compared to other sports.
While the Pythagorean expectation is a useful way to estimate regular-season success, the playoffs ultimately separate the good from the great. To measure a team’s success in the playoffs, I steered close to tradition. Simply put, I measured games won and win percentage in the playoffs. This effectively weighs how deep into the postseason a team went (games won) and how dominant they were against their opponents(winning percentage).
The final step in all this is to put it all together into one number. My “FinalScore” metric measures a team’s overall dominance in all facets of the game, and is a weighted average calculated as follows:
45%: “TeamScore” metric (Four Factors Weighted Average)
35%: Pythagorean Expected Regular Season Wins
14%: Playoff Wins
6%: Playoff Winning Percentage (As a whole number)
*Note: For the lockout-shortened 2011–12 season, I scaled every team’s Pythagorean win totals up so that they would match an 82, and not 66 game season*
The Top 10, without further ado…
The 2016–17 Warriors were pretty clearly the best team of this decade. Their “final score” of 41.5 was by far the best out of any of the 300 NBA teams that competed over the 10 years of this decade. In fact, out of a sample of 47 teams with a final score value of 32.5 or greater(the elite teams of this decade), the Warriors were a whopping 3.07 standard deviations above the mean. As pictured above, the next closest team, coincidentally the 14–15 Warriors, were only 2.19 standard deviations above the mean. In fact, out of a sample of the 30 teams in the 2016–17 season, the Warriors’ final score was 2.69 standard deviations above the mean. To put this into perspective, only one other team across this decade(14–15 Warriors with ~2.2) even had a final score greater than 2 standard deviations above the mean for their season, much less anything close to 2.7.
The Warriors weren’t particularly off the charts when it came to playing strong defense, forcing turnovers, protecting the ball, getting to the line, preventing their opponents from getting to the line, or grabbing rebounds on either end, but they were exceedingly good at one thing: scoring the ball, and scoring it efficiently. Their eFG% of 56.3% was the best in the league by a longshot, as the next best team (CLE) was only at 54.7%. To put these numbers into perspective, an EFG% of 56.3% indicates that on every possession the Warriors scored an average of 1.13 points from the field(2*0.563). On the other hand, their DeFG% was at 0.486, which means on every possession they were giving up 0.97 points on average. That means that, on every possession, they were outscoring their opponents by about 0.15 points, and so they were scoring an extra point about every 6.5 possessions. That’s remarkable! For reference, the Utah Jazz, a distant second in this metric for 2016–17, outscored their opponents by 0.07 points every possession, and hence took around 15 possessions to score a single point more than their opponents. The Warriors were doing this in less than ½ of the time!
The 16–17 Warriors solidified their dominance through on paper results as well. Their Pythagorean Adjusted Record of 67–15 was tied with the 2015–16 Spurs as the best mark of the decade. Only a mere 5 teams had Pythagorean Win Totals of 64 or above in this decade (the above Spurs and Warriors, the 14–15 Warriors with 65, the 15–16 Warriors with 65, and the 12–13 Thunder with 64). Their un-adjusted win-loss record of 67–15 is tied with the 15–16 Spurs for the second-best mark of the decade behind only the 15–16 Warriors who posted an astonishing 73–9 record.
While, by the numbers, the 16–17 Warriors were historically dominant in the regular season, their play jumped to another level in the playoffs. They handily took home an NBA championship, dispatching of LeBron James and the Cavaliers in 5 games in the Finals. Although winning 16 postseason games is an impressive feat in its own right that only one team each year achieves, the way the Warriors did it was exceedingly impressive. They finished their postseason run at a whopping 16–1, which equates to a winning percentage of about 94.1%, amongst the highest in NBA history. They dominated their opponents, not dropping a game until the tail end of the NBA Finals. Their +13.5 net rating in the playoffs is only behind the ’01 Lakers (+13.7) for the highest mark of this century.
The 16–17 Warriors used their combination of lights-out shooting (Steph Curry and Klay Thompson), lethal scoring (Kevin Durant), lockdown defense (Draymond Green), and a well-balanced supporting cast to ingrain their status as the clear greatest team of the 2010s. Given the unique circumstances of their rise to the top of the basketball world, it is unlikely that we will ever see another team in history reach the heights that they did.
