Many of you reading have probably heard people talking about advanced statistics. Maybe you’ve watched Moneyball or read any of the thousands of articles talking about NBA advanced stats and why they’re either ruining or saving the league(like this one). But, many are you reading this are probably unfamiliar with exactly what I mean by “advanced stats”. So, I will break down some commonly used ones and explain how they’re calculated, what they’re useful for, and some drawbacks of each.
NBA Advanced Stat #1: TS%
TS% is extremely common in more analytically minded NBA circles. The name is actually very misleading as True Shooting(“TS”) isn’t actually a percentage of anything and is just a unitless number. The actual calculation for TS% is (PTS) / 2(FGA + 0.44 FTA). Now, why is this used? Well, FG% is the commonly used metric you might find in boxscores in newspapers or on ESPN, etc. FG% is deeply flawed because it doesn’t count the difficulty of shots. Free Throws, 2-pointers, and 3-pointers are all counted the same. However, in real life, shots are worth different point values and have varying degrees of difficulty, meaning they should be counted differently. TS% is an attempt to remedy this by counting shots at different rates(ex. FTA are only worth 1/2 other FGA) and including points scored in the calculation.
This makes it useful for comparing the efficiency of different players. As you may guess, efficiency is very important in basketball. Teams can only take a certain amount of shots in a given game. As a result, the players who convert higher amounts of their attempts into actual points are thus more valuable. However, TS% is not without its flaws. The first is that it is a stat without context. All ten of the players on top of the TS% leaderboard are both big men and roleplayers. This is because big men have naturally easier shots closer to the basket as compared to long-range shooters. Furthermore, roleplayers take fewer shots and thus have a lower load to shoulder as compared to stars who shoot a high volume.
This might wrongly give you the impression that Daniel Gafford(70.8 TS%) at #1 on the list is better than Nikola Jokic(65.1 TS%) at #14 on the list. However, Jokic shoots significantly more shots per game than Gafford and faces more difficult defense than Gafford does. Jokic’s high efficiency even in the face of major defensive attention and high volume is what makes him a perennial MVP candidate. TS% is ultimately most useful for comparing players of similar positions and general tier of players.
NBA Advanced Stat #2: Win Shares
Win shares are highly controversial as they are the NBA’s answer to baseball’s WAR. The actual formula is both long and very complicated and its creator explains it here. However, the basic metric of Win Shares is how many wins a player is responsible for in their career. Now, this is a difficult errand, as you try to attribute team success to a single player, which is impossible. However, Win Shares are fairly accurate in terms of calculating a player’s value over a long time. That doesn’t mean Win Shares are without their flaws.
Unlike the previously discussed TS%, Win Shares are a context-dependent stat, meaning a few things. Firstly, great players on bad teams are shortchanged due to the amount of losses their terrible teammates rack up. Secondly, there are only a certain amount of win shares to go around in a season. This penalizes great players who competed in an era with many other great players(ex. everyone who played against Michael Jordan or the Dynasty Warriors). Thirdly, Win Shares are a volume stat. This means that good players who play a lot for a very long time naturally rack up many win shares.
You might be surprised to learn that Larry Bird, widely considered one of the top 5-10 players ever, is only #29 on the win shares leaderboard. He is ranked behind players like Chris Paul, Artis Gilmore, Dan Issel, and Reggie Miller, who are all very good or even great players, but are widely thought not to be as good as Bird. The reason for this is that Bird’s career was cut relatively short at only 897 games and he dealt with injuries throughout his career. The others all played forever(at least 1220 games) and were relatively healthy throughout their careers, allowing them more time to rack up win shares. Win Shares are ultimately very useful, but the context of a player’s career must considered when using them.
NBA Advanced Stat #3: Team Net Rating
The final stat I will cover in this article is Net Rating. Net Rating measures point differential across 100 possessions, either on a team or individual level. This means that if a team allows 100 more points than they score across 100 possessions, they have a net rating of -100. Individual Net Rating is confusingly very different despite being similarly named as it is a player’s offensive rating minus their defensive rating instead of just +/-, but that’s a subject for another day. Net Rating is useful for determining the best teams in the NBA and for determining who’s a contender vs. pretender. However, as you might have realized by now, like all stats, it has its flaws.
Like most advanced stats, it depends a lot on context. For example, Tom Thibodeau’s teams consistently have high net ratings and usually underperform those net ratings. This is because Thibodeau’s coaching philosophy is that players should play the maximum amount they physically can. This often means that Thibodeau’s teams have starters on the court even in blowouts where the other team has sent out their bench. On the flip side, teams with Kawhi Leonard on them tend to have artificially low net ratings. Leonard rests a lot of games per season and is off the court immediately as soon as the game is a blowout one way or another. Thus, Leonard’s teams’ net rating is often counting a lot of time he could have played, but chose not to for one reason or another.
Final Word
This is only barely scratching the surface of advanced statistics. There are many ways to analyze basketball and there have been decades of research on how to correctly do it. The main thing that you have to remember is that context is incredibly important for all stats. No single number will tell you everything about a player or team. It helps to start painting a picture, but context is all mighty. You always, always have to consider the extenuating factors that change stats. After all, there’s a reason they still play the games.
