Same as it Ever Was: The NBA's Most Consistent Players

While it is often said that "the best ability is availability", I'm not sure that is always the case. Sure, it's great to show up to work, but what we actually want to know is who REALLY shows up to work. To tackle this, we have enlisted the help of the indispensable Basketball Reference, which tracks a game score for every game a player plays. Game Score uses data from the box score to come up with a single number value to illustrate how well a player performed in the given game. The scale is similar to that of points scored, with 40 or above considered an outstanding performance, 10 being an average performance, and so on. To illustrate the idea that the same player is not always the same player, let's use Cam Thomas as an example. Cam has a high this season of 30.8 on the game score scale and has 10 games, out of 51 played, with a value above 20. He also, however, has 21 games with a game score below 10, including two games in the negatives! To illustrate what a negative game score looks like, take a look at his January 2nd game against the New Orleans Pelicans, where he shot 0-11 from the floor and finished with 0 points, 2 rebounds, 0 assists, 1 steal, and 1 turnover in 20 minutes of game time. Not great! On the other end of the spectrum, we have his November 6th performance against the Milwaukee Bucks, where he scored 45 points on 33 shot attempts and also chipped in 4 rebounds, 3 assists (a ton for a player of his ilk), 1 steal, and just 1 turnover in 40 minutes on the court. 

So, now that you have a better understanding of how game score works, let's lay out the process we will use to determine which players are the least and most consistent. First, I have taken all of the players' game scores for the season and come up with the average. Let's use Stephen Curry as our example. Steph's average game score for the season is 18.87, which we will use to create 4 buckets of player performance. We will call anything above 125% of that value (23.58) a 'great game', anything between 75% (14.15) and 125% of that value as a 'good game', anything between 50% (9.43) and 75% as a 'fair game', and anything below 50% as a 'poor game'. We will then compare each of his games against these numbers and we find that Steph plays a 'Great Game' 31% of the time, a 'Good Game' 42% of the time, a fair game 8% of the time, and a poor game 19% of the time.

We have now covered how game score works and how we are calculating how well a player performed in each game. The last item to touch on before we get into the data is how we should compare a team's first option to a role player who plays 12 minutes per game. To do this, we will use Offensive Load, which is a metric that attempts to qualify how involved a player is in the game. On the high end of Offensive Load, we would have a player like Luka Doncic, whereas on the low end, you would find someone like P.J. Tucker. So, the first thing we will do is group players into buckets based upon their Offensive Load and then compare players only to other players within their own group. Group One (Primary) will be players with an Offensive Load above 40 (74 players), Group Two (Secondary) will have a load between 30 and 40 (81 players), Group Three (Tertiary) will be between 20 and 30 (136 players), and Group Four (Role) will be those below 20 (57 players). Now that we've set the stage with our approach to analyzing game-by-game performance, let's dive into the superlatives and see who truly stands out.

Mr. Consistent

Primary

Shai Gilgeous-Alexander - Shai is the model of consistency and was actually the reason for undertaking this study. He "plays well", which means he has a game score above 75% of his yearly average in 86% of the games he has played this season. On top of that, he has only played "poorly" in 3% of his games this year, which is just absolutely bonkers.

Honorable mention goes to Luka, KD, LeBron who all play well in more than 80% of their games. Special shout-out to LeBron James for still being the model of consistency, even at age 39.

Secondary

Chet Holmgren - Chet is the only non-primary player to crack the 80% mark in games where he "played well".

Honorable mention to Anthony Davis, Bam Adebayo, Kristaps Porzingis, and Evan Mobley. Mobley, and especially Anthony Davis, come under heat for sometimes sitting on the outskirts of their team's offense, but these guys bring it almost every night.

Tertiary

Clint Capela - Nothing flashy here, just gets his 10, 10, and a block and gets out of there. Something to keep in mind should the Hawks decide to move on from Capela this summer.

Honorable mention goes to Aaron Gordon, Norman Powell, Jarrett Allen, and Jalen Johnson. Look out, league, Jalen is coming!

Role

Walker Kessler - I'm sensing a theme here. The six most consistent role players are all low-usage defensive stalwarts.

Honorable mention to Daniel Gafford, Mitchell Robinson, Nick Richards, Isaiah Hartenstein, and Rudy Gobert.

When it's Good, it's Good

When Jalen Green has it going, my word does he have it going. The only problem here is that while he has a "great game" 33% of the time, he has a poor game nearly just as often. He has 11 games, out of 66, with a game score above 21, which is Devin Booker's average game score. So while he is capable of reaching top 10 heights, he's just as capable of putting up a dud. Jalen has 14 games with a game score below 5, which is just about what Trendon Watford's average game score is this season.

Some honorable mentions in this category go to Reggie Jackson, Keyonte George, Cam Thomas and Scoot Henderson. Aside from Reggie, the theme here seems to be that young, score first, players struggle with consistency. If these guys can turn some of their very worst performances into merely decent games, then we could be looking at some future all-stars.

Data Dump

To view the entire dataset for this project, please visit the consistency player traits page.

GmSc - Game Score; the formula is > PTS + 0.4 * FG - 0.7 * FGA - 0.4*(FTA - FT) + 0.7 * ORB + 0.3 * DRB + STL + 0.7 * AST + 0.7 * BLK - 0.4 * PF - TOV. Game Score was created by John Hollinger to give a rough measure of a player's productivity for a single game. The scale is similar to that of points scored, (40 is an outstanding performance, 10 is an average performance, etc.).