DSC 106 · Final Project · UCSD

The Death of
the Midrange

The most efficient offense in NBA history killed the most beautiful shot in basketball.

scroll to explore

For most of NBA history, the midrange jumper was just part of the game. Guys like Kobe, Dirk, and Tim Duncan lived there. Nobody questioned it. Then front offices started running the numbers and realized those shots were basically the worst ones you could take.

Scroll through the data below and see how fast things changed once teams figured that out.

2003 – 2010

The Midrange Reigned

Teams ran their offense through the post or the elbow. Three-pointers existed but nobody was building a game plan around them.

14.9threes per game in 2003
2011 – 2014

The Numbers Don't Lie

Analytics started showing up in every front office. Teams realized the midrange was giving up value on every single shot. The math was pretty clear.

+43%rise in 3PA from 2003
2015 – 2017

Curry Changes Everything

The Warriors won a title shooting 31 threes a game and suddenly every team wanted to play that way. It went from a strategy to basically the only strategy.

31.3threes/game, 2015 champs
2018 – 2021

The New Normal

By 2021, four out of every ten shots in the NBA were threes. The midrange did not slowly fade out. It just kind of stopped existing.

40.1%of all shots were threes in 2021
2022 – 2030

Where Does It Stop?

If the trend keeps going, teams could be shooting close to 45 threes a game by 2030. The dashed line is the projection. The shaded area is the uncertainty around it.

~45projected threes/game by 2030

Then vs. Now

2003 to 2021. Same sport, completely different game.

Do more threes mean more wins?

Each dot is one team, one season. There is a relationship but it is not that clean. Plenty of teams shot a lot of threes and still lost a lot of games.

Champions led the shift

Every title team since 2012 ranked by threes per game. The jump after 2015 is hard to miss.

Where does your number land?

Drag the slider to any three-point attempts per game number and see which season that matches. A number that was elite in 2010 is below average today.

Predict your team's wins

Set a hypothetical 3-point shooting percentage and see what the historical data predicts for win rate — with a 95% confidence interval built from 569 team-seasons.

35.0%
predicted win rate
95% CI: — to —
Model: OLS regression on 569 NBA team-seasons (2003–2021). R² shown reflects 3PT% as sole predictor — real win% depends on many factors.