The art of the assist part 2: The potential of a pass

In early January I wrote the first part of this series examining the flaws of some statistics and offering simple replacements. It focused on AST:TO; if you're interested you can find that article by clicking here.

Not a day goes by where you don't see an argument on TV or social media about who the best point guard is. It will often turn into "X is a better scorer but Y is a more traditional point guard.” These run of the mill arguments have a lot of focus on raw numbers and scoring efficiency, and most importantly people will point to assist totals as the defining factor of who the best passer is.

But is it really that simple?

In the beginning there was... confusion?

When I embarked on this adventure through the land of basketball analytics I intended on designing a new way of measuring a players value as a distributor. My favorite aspect of basketball has always been passing. It's why most of my favorite players right now and of all time are and were great passers; there's something satisfying about seeing a player threading the needle with pinpoint accuracy or exhibiting their court vision with some ludicrous, flashy passing.

But by no means am I an expert in basketball analytics although I do have a keen interest in them; I simply wanted to try and find a better way of evaluating a player’s ability to create opportunities for teammates.

While trying to think of different ways to measure the ability to create shots for others, I did what all good scientists would do; I looked at what people before me had created.

The more I researched the topic the more writers, bloggers and redditors I found trying to do something similar. From SBNation’s own Tom Ziller to BleacherReport’s Kelly Scaletta and Adam Fromal, there are plenty of people before me who have tried to solve the same problem.

When I decided to calculate assist to passing turnover ratio (AST:PTO) it was mostly out of my frustration of the well-known assist to turnover ratio (AST:TO). It's part of why renowned analyst John Hollinger created pure point rating; the now current Vice President of Basketball Operations for the Memphis Grizzlies was tired of coaches and general managers evaluating players using AST:TO although his reasoning was slightly different.

After all of this research the next logical step for AST:PTO appears to be the inclusion of potential assists.

When is an assist not an assist?

For those who don't already know, potential assists are passes which have lead directly to a shot or free throw attempt, meaning if the player receiving the pass had made the shot it would have counted as an assist.

Again I am focusing on passing turnovers as opposed to all turnovers because a lot of the turnovers accumulated in the traditional AST:TO are irrelevant to measuring a players passing ability (stepping out of bounds and backcourt violations, etc). You can read more about my opinion on that in the first article of this series.

Due to time constraints I didn't have the time to calculate the assist + potential assist to passing turnover ratio (AST+poAST:PTO) for every player in the league. What I have managed to do, however, is calculate it for the league leaders in assists (the top 50 to be exact).

Some of the results were surprising to me and others not so much. I created an infographic to highlight some of the leaders in AST+poAST:PTO which you can see below:

How valuable is this information?

In my opinion there is a point at which statistics can become oversimplification. No single statistic can tell you what you exactly want to know; at the very least it requires appropriate context. It's better to look at all of the information available before rushing to any conclusions.

Some of the weaknesses of this ratio which you should keep in mind are as follows:

• Potential assists do have some merit but how much value can we really put into them? For example Stephen Curry has a total of 594 potential assists. Are we to assume that the only reason these weren't assists is because either the player was fouled (and they turned into free throws) or the player simply missed? Is it not possible that the receiving player missed because it was a bad pass? Perhaps it was too low or too high; we can't know for sure.
• There is a lot of subjectivity when counting potential assists and assists in general. It's also worth mentioning that a high volume of assists doesn't necessarily make player A better than player B, and it doesn't always mean they're a better teammate either. There's something to be said for the argument of players who go searching for assists and only pass the ball when the opportunity presents itself versus players who pass the ball around because it's the right play to make.
• Good passers should also be good ball handlers. The elimination of ball handling turnovers means we're only being told the ratio of passes that lead to an assist or could have lead to an assist if the player had made their shot versus passes that resulted in a turnover of possession. For that reason, Hollinger’s PPR is still a much better option as it reflects ball handling, the ability to create good shot opportunities and the relationship between assists and turnovers.
• It could be argued that some players may be passing the ball excessively due to their role, which would possibly inflate their potential assists.
• There are intangibles at play here that we simply cannot quantify, the degree of difficulty being the most obvious. A playmaker who takes more risks and is more creative might end up with more "bad pass" turnovers.

There are counterarguments to all of this and even more flaws that could be discussed (feel free to kick off the criticism in the comment section below!), but for now we're moving on to the final part.

Strength in numbers!

If you truly want to begin evaluating a players ability to create for others, you have to take into account many aspects: assists, secondary assists, potential assists, pace, various types of turnovers and so much more.

While there are many different formulas out there (this goes back to my earlier comments about Tom Ziller and Kelly Scaletta etc), the fact of the matter is the more I research this topic, the more I'm beginning to think that trying to compress a playmaker’s value into a single numerical representation is a rather conceited approach.

The only way to make a truly informed decision is to use all of the information available to you.

And even then, there is a beauty to this game and a level of creativity that the greats possess which no single number can ever define.

I'm hoping to continue this series in the future, perhaps identifying the "best" statistics and formulas to use in conjunction with my own ratio to better identify the top passers in the game (or at least the most efficient).

Until then we as fans can begin to think about how valuable some of these statistics truly are while Dub Nation can continue to boast about Draymond Green’s passing with some additional evidence to back it all up.