Commentators often mention the elusive “hockey assist” as an important potential statistic that could prove that Player X is a better passer than his regular assist numbers might indicate. But tracking hockey assists isn’t easy. Basketball moves very fast in real time, and defining what should count as a hockey assist is tricky. Imagine a possession in which one player hands the ball to a point guard at the top of the key, clears to the corner and watches as the point guard records an assist several seconds later on a pick-and-roll. Has that initial player contributed anything meaningful?
Good news: 10 NBA teams have purchased a super-sophisticated camera system from STATS LLC that tracks every movement on an NBA court to a precise degree. These are the same cameras, you’ll recall, that told us Tony Parker is the fastest point guard in the NBA. These cameras can track and sort everything, and the STATS folks decided to track hockey assists using a specific definition: A hockey assist, for STATS, occurs when Player X passes to Player Y, and Player Y then records an assist after holding the ball for two or fewer seconds and taking zero dribbles. The goal of the two seconds/no dribbles criteria is to isolate situations in which the initial pass — the hockey assist — has compromised the defense to the degree that the player who then records the “real” assist has little work left to do other than make a relatively simple pass.
So, who has the most hockey assists? Who records more hockey assists than we might expect? Who records fewer than we might expect? Before we answer, it’s important to note the caveats here: The STATS cameras are in only 10 of 30 arenas, and in order to filter out random noise, the STATS study supplied to SI.com tracked only players who have appeared in front of the cameras in at least eight games this season. That rules out some pretty darn good passers, including Chris Paul and Deron Williams.
Without further ado, here’s the hockey assist leaderboard, with total games tracked in parentheses:
1. Derrick Rose, 1.9 per game (10 games)
2. Steve Nash, 1.6 per game (8 games)
2. Raymond Felton, 1.6 per game (11 games)
4. Mike Conley, 1.4 per game (8 games)
4. Tony Parker, 1.4 per game (31 games)
6. Brandon Jennings, 1.3 per game (29 games)
6. Rajon Rondo, 1.3 per game (26 games)
All tied at 1.1 per game: Russell Westbrook (35 games), Darren Collison (12 games), Manu Ginobili (15 games) and Jose Calderon (29 games).
You’ll notice that the hockey-assist rankings don’t line up with the overall assist rankings. Rondo and Calderon drop, while allegedly so-so passers like Westbrook, Conley and Jennings shoot up the rankings. Each of those three players, along with Ginobili, has a higher ratio of hockey assists to regular assists than the average player tracked in this study.
You could read this data any way you like, which makes it both confounding and interesting. It may end up saying just as much about a player’s teammates as it does about a particular player.
Take Conley, for instance: He averages a ho-hum 6.8 assists per game for a team that has ranked near the bottom of the league in assist rate every season he’s been in the league, but he does well here. Why? Well, he has the ball quite a bit, and he runs a lot of pick-and-rolls with Marc Gasol, one of the best passing big men in the league. How would John Wall (just 0.5 hockey assists per game in 19 games) or Calderon (fourth in the league in regular assists) fare in this category had they played a ton of minutes this season with a big-man passer on Gasol’s level?
Jennings ranks just 18th in assists per game and doesn’t have a big man on Gasol’s level, but he finds himself this season as the trigger man for a newly pass-happy Bucks team that assists on a higher percentage of its hoops than all but one team (Boston). Is Jennings improving as a creator, or is he benefiting from a dynamic team context? It’s probably a bit of both.
We might expect Rondo, the league leader in assists, to do better here, especially because Kevin Garnett remains a splendid interior passer. But on an aging team of players who aren’t as good as they used to be at creating their own shots, Rondo’s job is to make the last pass — to set up a shot attempt, rather than start a series of passes leading to an eventual open look. The same might true of Ricky Rubio (just 0.6 hockey assists in 26 games); Kevin Love and Nikola Pekovic have worked mostly as finishers this season, and Minnesota’s wing players aren’t especially threatening passers.
It’s fitting that Rose is the leader by a decent margin, because that stat matches well with the eye test and says quite a bit about both Rose and his teammates. You can picture those hockey assists in your head right away, can’t you? Rose sets up for a pick-and-roll with Joakim Noah, and Noah’s man decides to trap Rose above the three-point arc. Rose threads a bounce pass to Noah at the foul line, and Noah, seeing Carlos Boozer’s man rotating his way, slips a quick-hitter to Boozer for a layup. The sequence happens so often, both because Rose demands so much attention and because Noah and Boozer are clever passers capable of playing either role in the above scenario.
The appearance of two San Antonio guards here is not a surprise; no team whips the ball around as quickly or precisely as do the Spurs.
Lastly, Westbrook does quite well here, considering he averages just 5.4 assists for a team that ranks last in assist rate. And Westbrook actually ranked even higher based on the criteria STATS first used in defining hockey assists, says Brian Kopp, a vice president at STATS. The company started with a broader definition of hockey assists that would have allowed for the final passer to hold the ball for up to four seconds and take two dribbles before dishing the actual assist. The company chose that benchmark after studying how the NBA scores actual assists. When using this definition for hockey assists, Westbrook averaged more such secondary dimes than any player in the STATS sample, Kopp says.
But the company ultimately decided the four seconds/two dribbles standard counted too many mundane passes that did little to break a defense, and thus did not fit well with the concept of hockey assists. The company is open to other criteria sets somewhere in between the initial one and the current one, which is a reminder that we are at the beginning stages in this marriage of advanced stats and video technology. The possibilities are exciting.