How Far Back Would We Have To Move The Three Point Line To Bring Back The Midrange Shot?
Lots of “old heads” as they’re called — those members of the basketball community who complain that today’s game is soft, that the players are worse, that defense doesn’t exist and that the game was significantly better back in their day—these people are upset that the midrange shot is being phased out of the game.
And maybe they have a point. It can be really interesting to see players take shots from everywhere on the court, and watching a midrange specialist go to work can be an extremely exciting experience.
The problem is, teams (like the Houston Rockets) have been paying attention to the numbers, and those numbers are telling them how extremely inefficient the midrange shot is. Shots from midrange are less likely to pass through the net than than shots taken in the paint, and are worth less than shots taken from three. Someone shooting 40% from three would have to shoot 60% from midrange make the two shots have equal value (40/100 made 3's=120 points, 60/100 made 2-pointers=120 points).
In the past decade, exactly zero players have shot 60% from midrange over the course of a season (of players that shoot at least three midrange shots per game and to have played at least half the season). By the same criteria (>3 3fga, at least 41 games played), 26 players reached the 40% mark from three. And that was only last season.
The average team makes about 40% of their shots from that zone between 8 feet from the basket and the three point line.
Midrange shots are great not only because of players like Michael Jordan and Kobe Bryant, but also because they open up tons of new schemes in playbooks and utilize every inch of the floor.
All of which begs the question: How do we bring this shot back? In the modern NBA, we’d have to make the midrange jumper a statistically viable shot.
To get a shot’s expected value, we take the number of points received for a made bucket (Free throw=1, Shot within the arc=2, Shot outside the arc=3) and multiply it by how frequently it goes in (if a 2-point shot goes in 50% of the time, the expected value is 1. If a 3-point shot goes in 50% of the time, the expected value is 1.5).
A mid-range shot right now is considered a shot from 8 feet all the way out to the three point line (23.75 feet out, 22 feet in corner). In this new scenario, I test the expected value of shots from 8-x number of feet, where x is the distance of the three point line (and a three point shot would be from roughly one foot behind it).
My scenario eliminates the corner three, making that space now part of the midrange zone. The corner three is closer than other areas of the three point line, and if we didn’t move it in, we’d see a whole lot of shots coming from those corner spots.
All data collected from the 2016–2017 nba season.
I gathered data from 8 ft to x number of feet. I started with x at 23, so the three point line would be 23 feet out, a three pointer would be taken from about 24 feet out, and there would be no corner three. I selected the 300 players with the greatest number of field goal attempts from 8–23 feet. I took the total shots made by those 300 and divided them by the total shots attempted by players from that range. That gave me the average field goal percentage in the designated midrange zone. I multiplied this percentage by the 2 points a made shot would produce to get the points expected per shot.
Next, I took the 300 players with the most shot attempts from 24+ feet, the range which in this scenario is considered a 3 pointer. I again divided the total shots made in this range by total shots attempted in this range to get the average field goal percentage. Next, I multiplied this number by 3 to get the expected number of points shooting a three pointer would amount to.
Midrange (8–23 ft) fg%: 40.81%
2 points*0.4081=0.816 expected points
Three point (24+ ft) fg%: 35.33%
3 points*0.3533=1.0599 expected points
As you can see, the expected outcome of shooting a three pointer is much higher (0.24 points per shot) than shooting a shot from midrange.
So I changed the parameters, extending the three point line by 1 foot to make the midrange zone 8–24 feet and the three pointer 25+ feet. I again took the top 300 in fga from both areas, and found field goal percentages and expected values for each.
The midrange shot gained a bit of value relative to a three pointer when extending the line by a foot, but it was still not worth shooting — a 3 pointer was still likely to yield 0.218 more points per shot attempt than a shot from midrange. I did this a few more times, gathering data and tediously copying-and-pasting for each 1-foot increase to the distance of the three point line.
It was once we moved the three point line out to around 27 or 28 feet that we got roughly equal expected values for a midrange shot and a three point shot.
Let’s put the three point line at 27 feet. For most players, they’ll be taking threes from about 28 feet out (as they have to be behind the line).
Where would this be on a basketball court? More MS Paint please:
What would this look like in a game?
Here are some shots from 27 feet out:
And here are some shots from 28 feet out:
The caveat to this extended distance is that there isn’t a great way to project how teams will play under these new circumstances. They may start shooting midrange shots, but they also might force it in the paint, or maybe they’ll continue to pour it in from deep. We’d probably have a ton more shots in the paint, though teams might place more bigs down low, causing field goal percentages in the paint to decrease. For now, we don’t really know what would happen in this new NBA. A deeper dive into the effects of this new scenario could be a source of exploration for a future article.
So who are the best 3 point shooters (based on the new criteria) so far in the small-sample-size 2017–2018 season?
Bonus: top shooters from 27+ feet (there are a lot more shots taken)
I think I finally found a fit for Marcus Smart on offense.
Credits to basketball-reference.com for the data
