STAT CHECK – Using Goldsberry formula who on last year’s Jazz was best shooter

Today Kirk Goldsberry did an interesting piece at Grantland. The premise was to figure out who is  the best shooter in the NBA, using a metric better than field goal %.   The thesis was De Andre Jordan has a better FG% than LeBron, but it is because of the lack of variance in his shot selection.

Goldsberry plotted every shot of every player and compared what their projected points scored should be compared to the league average based on shot location.

Inspired by Goldsberry work today I built a similar chart  on last year’s Jazz team.  I don’t have access to anywhere near the data that Goldsberry does.   Instead, I broke all shots into 5 locations (rim, 3 to 9 feet, 10 to 15 feet, 16 to 23 and 3 pointers) and compared how players performed based on the league average at those locations.

Looking at the charts below you will see the Expected Pts Total, this is the amount of points a player who shot average from the 5 locations would score.  Next is what the player actually scored.  This is followed by the difference and then finally equalizing for shot attempts we have the Shot Score (Goldsberry term) for every 100 shots.  If a player is positive this is the amount of points he would score above the league average of his shots in 100 shots and negative is the opposite

expected point chart 1

In this next chart I added some other data.  Goldsberry did nothing to incorporate free throw shooting.  I added a section that counts a players performance at the free throw line.

expected point chart 2

There are some flaws in this research.  Players who choose to shot a large amount of shots from inefficient spots on the floor, but shoot those shots at a higher percentage than league average are rewarded.  However, Goldsberry premise was he was trying to discover the league’s best shooter.

Points per shot attempt or pts per scoring opportunity (FGA and FT trips) are much better at evaluating a player’s offensive performance because they reward getting to the line and shooting three’s.

Therefore, in this chart  I have added pts per shot and pts per scoring opportunity

expected point chart 3

What I discovered is Al Jefferson is truly one of the best shooters in the game.  The shots he takes are very inefficient yet he does it well.  His shooting was the best on the team last year.   However, the majority of shots were very inefficient s and therefore his points per shot and points per scoring opportunity was not very good.

In contrast, Hayward has a terrific points per scoring opportunity, but this is based on shooting efficient shots and going to the line.  It turns out Hayward is not an elite shooter.  His shot score per 100 shots was only .99 pts above league average.

Of the current other 3 young Jazz players this shows how far Derrick Favors needs to go to develop his offensive game as he was considerably below the league average.  Alec Burks did not have a good offensive year last year, farther below average than Derrick Favors.

On the other hand, Enes Kanter graded out as the 2nd best shooter on the Jazz roster (excluding Jeremy “small sample size” Evans).  If Kanter continues his development he may be the offensive scorer of this group.

Now this leads to the next research are you better off having good shooters or just shooters who shoot from the “good”(efficient) spots on the floor?