Chris Paul, one of the NBA’s best assist men. (Wikimedia Commons).
Monday I will be sorting through the league’s top point guards and as a precursor decided to look at the best tools we have to evaluate these floor generals ability to distribute. All together, there are four stats readily available on stats.NBA.com and Basketball-Reference that measure a players ability to assist: the original assist stat (AST), Assist Percentage (AST%), Assist Ratio (AST Ratio), and Assist to Turnover Ratio (AST/TO). Each of them provide some insight into a player’s ability to assist or the way they assist within their team’s system, but each also comes with their share of shortcomings that limit it from being an exhaustive evaluation of a player’s ability to assist. Let’s take a look at the strengths and shortcomings of each particular statistic.
Assist
The main value of the assist stat is that it exists. Without the pure assist being recorded we would have no way to develop stats like AST%, AST/TO, and AST Ratio and would essentially have no way to determine how a player is distributing to his teammates. That being said, the assist stat is limited in the same ways most traditional NBA stats are limited. That is, simply recording a player’s assist total does not take things like the pace the player plays at and the opportunity he has to rack up assists into account. Compare Rajon Rondo and Steve Nash‘s best assist years for an example. Their career-best assist numbers are almost identical — Rondo with 11.7 per game in 2011-12 and Nash with 11.6 in 2006-07 — but those numbers do no factor pace into the equation. Rondo’s Celtics played a slow 90.4 pace that season, while Nash’s Suns, in Mike D’Antoni‘s seven seconds or less system, earned a 95.6 pace rating (per Basketball-Reference). That means Nash typically had an extra five possessions in which he could rack up more assists. So while their assist numbers in those year are almost identical, Rondo was actually a more bountiful assist man in his 11-12 campaign.
The point of all that being: while we need the assist to analyze who the best distributors are, simply looking at assists and assists per game will not give us a full understanding of a player’s ability to distribute.
Assist Percentage
So AST Ratio and AST% were developed to give us a more nuanced understanding of a player’s ability to distribute. AST% is a measure of the percentage of his team’s field goals a player assists on. NBA.com boasts an exact measurement, taking the player’s assists and dividing by the number of field goals made by the team — when the player is on the court — subtracted by the number of field goals the player made (AST/). This stat provides some information the traditional assist stat cannot. Particularly, it tells us how much a player is responsible for his team scoring. If a player has a high AST%, he is responsible for a large portion of his team’s offense (ex. Rajon Rondo’s AST% was 49 percent last season, per NBA.com, meaning he assisted on almost half of Boston’s made field goals).
However there is a particular flaw in AST%; let’s compare Rondo and Chris Paul‘s assist numbers last year as an example. Rondo earned a league high 49.0 AST% last season (of players who played at least 30 games last year) while Paul was a distant second with a 44.3 AST% (per, NBA.com). However, Rondo’s Celtics earned a 99.2 offensive rating when Rondo was on the court, while Paul’s Clippers earned a 112.1 offensive rating when Paul was on the court (per NBA.com). So, while Rondo was assisting on a higher percentage of his team’s scores, Paul’s team was scoring more frequently and the rate at which Rondo and Paul were assisting was probably much closer than their AST% would lead on. Point being, AST% tells us a lot, but does not necessarily tell us the whole story.
Assist Ratio
Assist Ratio can be both a player and a team stat and actually has much more value as a team stat than a player stat. Assist Ratio seeks to measure how often a player or team uses (uses are assists, turnovers, field goal attempts and 44 percent of free throws) are assists. Essentially, a player’s/team’s assists are divided by his/their uses (AST*100 divided by FGA+[.44*FTA]+AST+TOs). As a team stat, you can see how this is valuable (though the inclusion of assists for the denominator seems unnecessary, but I may be missing something). Team assist ratio highlights how often a team assists on their possessions; if a team has a higher assist ratio, they assist on a higher percentage of their possessions. However, for a player, this number simply highlights players who assist at a much higher rate than they shoot. For example, imagine Chris Paul matched up against Jose Calderon, and in the game Paul had 20 points (on ten FGA) and ten assists to one turnover, while Calderon had ten points (on five FGA) and ten assists to one turnover. Even though they assisted at a similar rate, Calderon would have a higher assist ratio because Paul took five more shots than Calderon.
Assist Ratio for a player can still shed light on a player — for instance informing that a player like Calderon will distribute the ball without taking up much shots — however, it does not give us much information when determining who the best assist men are.
A quick note on Assist to Turnover Ratio
At its core AST/TO is trying to doing a very particular thing — determine how often a player turns the ball over when trying to distribute — and is not in the same category as AST, AST%, and AST Ratio. However, it is still a very valuable evaluator and reveals a lot about a player’s ability to play point guard efficiently even though it does not necessarily reveal anything about the rate at which a player distributes the ball.
Instead, it seems like AST% — despite its flaws — is the best evaluator of assist rate that we have at our disposal, though I think their may be something more effective that can be developed (more on that later).