RealGM

Question.

When determining who is a better passer / better for their team's ball movement, what do you usually look at?

I'm assuming some combination of usage % and assist %.

Just using straight assist % wouldn't work - because what if 2 players have the same assist % but one of them has way higher usage %...then obviously the the guy with lower usage % is better in terms of passing / helping his team's ball movement.

Would you just take assist % ÷ usage % which gives you each player's assist to usage ratio and then just rank them that way?

Can look at AST/TO, too

Could look at passing rating at 82games.com and team assists and efficiency on / off.

Some passing ratings at bbindex.com

NBA.com players tracking passing

When I’m evaluating a players passing through a statistical lense I look at

AST/100 / IA AST/100
TOV/100
AST%
cTOV%
O Load
Box OC
Passer rating

And in the pbp era looking at things like where the assists are (are they rim assists 3pt assists mid range asts etc) and in the optimal tracking era Hockey assists potential assists etc are also very nice to look at.

Thread bump, but it seems a little quiet in here.

BackPicks has the Box Creation and Passer Rating metrics. I think both of those are behind his Patreon pay wall, but they make for interesting reads. Related to Passer Rating, Ben did a Nylon Calculus article related to that Passer Rating stuff which isn't behind a pay wall, and it serves as a study of improvement in passing.

There is also the stuff at NBA.com's Player Tracking.

There, you can see stuff like passes made/received, assists, secondary assists, potential assists, assist points created and so on. It allows you to get a better idea of WHAT someone is doing when they pass, how many points their passing is creating and how much their team might be leaving on the table with assists versus potential assists and all that.

Assists by shot location are publicly available.

Sometimes you see by teammate passer / receiver.

Teams could probably see assists and turnovers by court location of passer & receiver and could evaluate relative passing efficiencies. The most serious researchers could get to pass length and efficiencies.

Lately I find the giving (or not giving) of assists likely subjective. I mean a blow-by, three dribbles, and a fade away from an interior defender is most of the work, right?

I you find him open and shoots the three ... sure. If by the time he gets the ball, an immediate shot is going to be contested ... then I wouldn't hand out the assist.

It's a pass yes ... an assist no. He needs to start with an advantage, I think, for you to get extra credit.

how many points their passing is creating

Be careful what you wish for.

A few years back when Spencer Dinwiddie was the starting PG for Brooklyn someone I think on Bleacher Report wrote an article saying he was the best passer in the league. Why? Because at that time he was throwing for assists at the highest rate of assist per pass (something like 1 assist for every 6 passes) in the league.

His conclusion? That Dinwiddie was throwing for the most part great passes that were more often being converted to points - even though other PGs at the time were throwing for more assists per minute (but at rates of something like 1 assist for every 7-8 passes).

But what the author of the article did not in fact take into account was who his teammates were. Two of his teammates were Joe Harris and Allen Crabbe, two definitive spot-up shooters, players who often shot off the catch, who rarely took a dribble once they got a pass. You have teammates like that and you are going to compile assist numbers, even if they are not the best shooters, simply because they shoot often after catching a pass, without dribbling.

You would think playing alongside high scorers as a PG you would get a lot of assists. But you play PG alongside high scorers like James Harden, Lebron James, or Kevin Durant and you won't, because they rarely shoot of the catch.