Nylon Calculus: Who are the NBA’s most productive passers?

Passing is a valuable skill in basketball and does not receive enough credit.  Typically, only the best superstars have the ability to create their own shot at an efficient rate. The majority of players require a pass from a teammate to score. A good pass is a selfless act that makes teammates better. A great example is Steve Nash, one of the newest members of the Hall of Fame, who helped teammates A’mare Stoudemire and Shawn Merion become All-NBA players with his ability to attract defenders and make on time and accurate passes to open teammates.

A more recent example of the importance of passing comes from the Golden State Warriors. Known for both their player movement and ball movement the Warriors’ offense has performed at historic levels. It doesn’t hurt that they have three of the most efficient scorers in the league but their passing plays a large role in their success. The one drawback for the Warriors has been that they have too many turnovers, supposedly due to their high number of passes made. The Warriors routinely rank in the bottom third in the number of turnovers in the league over the last four seasons. The thing is, they make a lot of turnovers but the risk is worth the reward. Their skill level doesn’t hurt either as they still come out way ahead of most offenses.

I want to look deeper than assists or assist to turnover ratios. My goal with this analysis is to find the players from the 2017-18 season that had the most productive passing performance. To start, I define a productive pass as the sum of assists, secondary assists, and potential assists. Productive passes are passes that lead to a shot. Bad passes are taken from the play-by-play data available from stats.nba.com with two distinct types; live ball and dead ball. Dead ball bad passes (BP_Dead) are usually the outcome of an errant pass that sails out of bounds and ends in a turnover. Live ball bad passes (BP_Live) result in a steal for the opposing team and are worse as they generally lead to a fast break for the opponents.

After aggregating the productive passes and the bad passes I created two ratios to analyze productive passing of each player. First is the good-pass-to-bad-pass (gp-to-bp) ratio and, as the name implies, is good passes divided by bad passes. Second, is the assist-to-bad-pass (ast-to-bp) ratio is similar but only includes assists divided by bad passes. The table below shows the top ten 2017-18 players with at least 500 passes and ranked in order of the number of passes thrown multiplied by ast-to-bp ratio.

Spencer Dinwiddie leads the league as the most productive passer with a career-best 6.6 assists per game. Combining Dinwiddie’s relatively high number of passes made and assists with a very low number of bad passes leads to his high ranking. Take his passing productivity along with his win in the Taco Bell Skills Challenge and Dinwiddie had a pretty good year, especially for a guy who has spent a large portion of his career in the G League.

Additionally, I investigated sorting by a player’s good pass to bad pass ratio.  Awarding credit for secondary assists and potential assists did not change the ranking too much. The only significant change among the top ten was that Goran Dragic replaced Kyle Lowry.  Could this be due to Dragic being surrounded by worse shooters? Or could it be because he was not providing accurate and timely passes for those shooters? Controlling for pass accuracy is not possible with the play-by-play data available. Using a more in-depth analysis of film and in-game tracking such as Ben Taylor’s article could provide more information.  This is one of the reasons I chose to rank players on assists-to-bad-pass ratio over good-pass-to-bad-pass ratio.

Another benefit of using only assists is that assists are trackable in play-by-play data. With a little work, we can visualize not only the assist-to-bad-pass ratio but also who each player assisted. The following static visual shows each assist and turnover throughout the season for Spencer Dinwiddie. Notice the density of assists compared with the sparsity of Dinwiddie’s turnovers, especially the bad pass branches. Dinwiddie assisted Allen Crabbe,  Rondae Hollis-Jefferson, and DeMarre Carroll more often than the rest of his teammates. 

Lost ball live ball (LB_Live) turnovers occur when a player gets the ball stolen while in possession. Lost ball dead ball (LB_Dead) turnovers typically occur when a player has possession of the ball and steps out of bounds. Offensive fouls are straightforward and violation turnovers include things such as traveling, palming, and illegal screens. For a player with the ball in his hands as much as Dinwiddie, his turnover counts are relatively low.

While Spencer Dinwiddie had the most productive passing season, other players had productive passing seasons as well.  The following interactive visualization allows you to watch the assist and turnovers for almost any player in the NBA from the 2017-18 season. If you have a few minutes and some nostalgia for retro video games, watch your favorite player’s assist to bad pass ratio occur on a play-by-play basis. Select your player, check your volume,  click the play button, and sit back and enjoy watching your player’s assists and turnovers plot out chronologically. The full-screen version is available