.png)
As All Star weekend is approaching in the 2020-2021 NBA Season, NBA twitter is currently storming with comments about the starters, snubs, and selection criteria. Recently, I started to wonder whether the players selected for the All star game were actually the 24 best players in the league, and if they are, does having more all stars on your team guarantee a better winning percentage.
The problem is that we use All-Star selections as a criteria for determining how good a player is without considering how unfair that is players that barely miss the cut. Anybody that is somewhere near the 25-30 ranked player in the league will have equal or similar impacts on their team's winning percentage as the All-Star players who barely make the team, yet some players will be held in high regard while others remain to be under appreciated.
So I decided to look back at the regular season and playoff successes of each team in different seasons and compare that to how many all stars they had for that year. I decided to look at the last 5 full NBA seasons(2015-2019) in order to gather my data.
Results
From the data, we can see that there is a positive correlation between the number of All-Stars a team has and how successful they are. The mean number of wins increases as the number of all stars goes up, while at the same time the range of the data decreases for each successive number of all stars per team. However, the extent to which this is true is not as great as most people would believe. There have been numerous teams with 0 all star players that have done just as well or even outperformed teams with 1 or 2 all stars in the regular season. Nevertheless, a clear pattern emerges from the regular season data for the success of a team and the number of all stars they have.
Interestingly, when we look at the playoff data, the strength of the correlation significantly decreases. The number of all-star selections plays a less important role in predicting how many game a given team will win in the postseason.
Interpreting the Data
So why is the correlation so different for the 2 data sets? The answer lies in the selection bias for the game itself. While fans play a significant role in determining the all star players, 7/12 players for each conference are selected by the head coaches of the 30 teams. These coaches have historically been biased towards players that are on teams with winning records, which means that some very talented players will never be selected just because they are in the middle of the pack and struggling for a playoff spot. In contrast, some players will get the all star nod despite simply being an above average starter at their position and just happen to be surrounded by a group of talented players.
Once teams get to the playoffs, having more all star players only helps to a certain degree. As teams start to get deeper into the playoffs, factors such as depth, and coaching become more and more important, as most of the teams have similar talent levels. It becomes easier to game plan against teams that only have 1 or 2 guys who can dominate the game, meaning that teams with poor depth often get eliminated in the first or second round. Well coached and balanced teams like the Indiana Pacers and Utah Jazz always make the playoffs and pose a serious threat to teams with more stars even though they have usually had 0 or 1 all star selections per year during this period.
Even teams with 3 or 4 All-Stars aren't guaranteed to win, such as the 2015 Hawks or 2016 Warriors. In contrast to this, we have seen multiple teams with only 1 all star player get their teams all the way to the conference finals or NBA finals, including the 2016 Cavs, 2017 Spurs, 2018 Rockets, 2018 Cavs, 2018 Celtics, and 2020 Heat. It would be very hard to find someone who didn't think any of those teams were talented, yet the number of All Star selections they had did not fully express the potential of the team, which was only later shown in the postseason.
It is also important to consider the differences between borderline All-Star players and the Super Star franchise players who get selected year after year. Players like Kyle Korver, Jeff Teague D'angelo Russell, Pascal Siakam, and Draymond Green are valuable to a team, but are not consistent players that you can put a tier above other notable starters and role player in the league. These guys aren't nearly as impactful as superstars like Lebron James, Kevin Durant, Steph Curry, and Giannis Antetokounmpo, let alone perennial all stars like Paul George, Russell Westbrook, and Kyrie Irving.
In conclusion, All-Star selections are a better representation of team's regular season success than they are of an individual player's actual talent. Teams with championship aspirations will need superstars, not just all stars, to have a realistic chance at achieving their goal. While the number of all stars does have a positive correlation with the success of a team, after a certain point the contributing role players and coaching staff become vastly more important in determining their odds of winning a game or a playoff series. Remember, players like Jrue Holiday, Jimmy Butler, Bradley Beal, and Chris Paul have all been "snubbed" from All Star games for a variety of reasons in the last 5 years, but any team in the league would be willing to trade for them in a heartbeat due to their amazing skillset and basketball talent. Every team in the league has players that don't get selected as All-Stars but are still talented and can contribute to a deep playoff run. The truth is that the worth of a player is made in the moments where it really matters, not in an exhibition game.
Comments