The Importance of Shot Selection

by Baltej Parmar (Twitter: @BaltejNBA)

Entering the 2018–19 NBA preseason, the Milwaukee Bucks over/under was set at 47.5 wins by Las Vegas. They just won 44 games in the previous season under the coaching of Jason Kidd. A small improvement was expected from internal growth and the change in coaching from Jason Kidd to Mike Budenhulzer. However, the major jump that ended up taking place should have been possible to foresee by the end of the preseason.

As the preseason was ending, multiple columnists and writers notes the change of the Buck’s offense under Budenholzer. Still, most viewed the Bucks as a middle-of-the-pack team in the East that was likely to end up in the high 40s or maybe break the 50 win mark. Their shot selection drastically changed, but the talent was not highly regarded. Let’s dive into the Bucks offense from a statistical perspective and see what the actual changes were.

Note: I will be ignoring free throws for this current exercise. All data is collected from

The Bucks’ Shot Selection

In the 2017–18 season, the Bucks were the 9th-best offense in the league with an offensive rating of 109.8. They achieved the mark while attempting 64.3% of their shots from either the Restricted Area or behind the three-point line. This is what I would deem “smart shots,” as they are the two most efficient areas to shoot from (aside from the free throw line). The Bucks’ 64.3% attempt rate ranked 18th in the league that year. We can look at the percentage of makes for these shots compared to league average in the table below.

*Includes Non-Restricted Area Paint shots & Midrange. The reason for this is the percentage of these two types of shots going in are virtually the same

Overall, Milwaukee’s talent was not bad in terms of shot making. They were below average from three but above average from two-point range. Let’s compare the table to the table we get for the 2018–19 Bucks.

Looking at the table above, the Bucks were below average in three separate parts of the floor (though they did improve in the Restricted Area), yet a jump took place in their offensive efficiency. Even though the Bucks did not have great shot-makers on their team, they were still able to become one of the best offenses in the league by simply restructuring their shot selection.

As mentioned earlier in the article, 64.3% of Milwaukee’s shots were “smart shots” in the 17–18 season. In the 18–19 preseason, this figure jumped to 81.4%, good for the highest number in the preseason. The difference represented a drastic change of 17.1% more smart shots coming into the season, which happened to be the greatest improvement in the league. In the regular season that number dropped down to 79.3%, which was good for 2nd-best behind the Houston Rockets.

What Effect Did the Bucks’ Shot Selection Have?

Looking at the calculations below reveals how we can actually quantify these changes. Below is the chart for the 18–19 Bucks.

The offensive rating for the Bucks based on this shot selection while shooting at league average over 100 possessions would have been 108.47.

Looking below at the 17–18 Bucks but adjusting for the number of shots to be equal to that of the 18–19 season while assuming the league average shooting of the 18–19 season we get:

The offensive rating for the Bucks based on this shot selection while shooting league average over 100 possessions would have been 104.23.

A difference in 1 point differential over the course of a season is roughly equal to 2.5–2.7 wins. Assuming all else equal, the change of 4.23 in offensive rating based on shot selection would result in 10.58–11.42 wins.

The Milwaukee Bucks jumped up from 44 to 60 wins over the course of one year due in part to the massive system change in terms of shot selection (I will cover the changes they made to their defense in another article). A huge portion of the improvement needs to be credited to Giannis Antetokounmpo for being able to generate looks at the rim at a high rate while also collapsing the defense to generate three-point opportunities. The shots Antetokounmpo got at the rim increased from .454 to .573 while also improving his efficiency from that range year-over-year. The Bucks did a great job to change up their personnel to match their new shot selection. This included bringing in big men who could shoot, such as Brook Lopez and Ersan Ilyasova. Milwaukee was also able to change the habits of mid-range sniper Khris Middleton to reduce his volume of contested mid-range jumpers and start taking more threes.

Switching Shot Selection Profiles for Two Teams

Let’s also take a look at what would happen if we switched the shot distribution of two teams. The Utah Jazz ranked 4th in “smart shots” during the 18–19 season, taking 75.3% of their shots from those area. The Pacers ranked 28th in shot selection, taking 64.0% of their shots from the preferred areas.

Below I have outlined the two teams shooting numbers compared to league average:

As one can see from the above table, the Jazz had below-average shot makers in three of the four zones and conversely, the Pacers had above-average shot makers in three of the four zones. Surprisingly, however, the Jazz had the 15th-best offense in the league while the Pacers had the 18th-best offense (Note: the Jazz also had a higher Free Throw Rate). However, we can see that the better shot distribution of the Jazz is pulling their offensive rating up while the Pacers’ poor shot selection is holding them back.

Below are the shot tables for both the Jazz and the Pacers:

Jazz 18–19 Season:

The offensive rating for the Jazz based on this shot selection while shooting league average over 100 possessions would be 107.53.

Pacers 18–19 Season:

The offensive rating for the Pacers based on this shot selection while shooting league average over 100 possessions would be 104.12.

There is a difference of 3.41 in terms of offensive rating based on the shot selection between the teams. If the Jazz changed their shot selection to match the Pacers’, it would result in a drop approximately 8.53–9.2 wins, assuming all else equal. Also, considering the Jazz have below average shot makers, the drop would be even more than that. This change can be applied to the Pacers as well (moving in the opposite direction).

Clearly, shot selection can have a drastic impact on a team’s performance. It can take a mediocre team and turn them into a great one, but can also take a talented team and turn them into a disappointment. It’s only a matter of time until all teams try to implement it and reap its benefits. I will be writing about how the Jazz seem poised to take a major jump for the 2019–20 season now that they have added above-average shot-makers based on Utah’s shot selection, so stay tuned for that analysis.


2 thoughts on “The Importance of Shot Selection

  1. It would be incredible if there were an automated way to do this for all the teams. It would also be interesting to run the analysis for individual players to show the breakdown between shot selection and shot making.

    1. I basically have the offensive rating part automated, I could post something that shows all 30 teams offensive rating if they were to shoot at league average efficiency on here or on my twitter if you’d like


Leave a Reply