Kentucky's love for midrange shots is costing them
The Wildcats make it a point to take midrange shots, but the data says it's not productive
The following exchange took place after UK’s 68-66 loss to Vanderbilt on March 1, 2023 (video here, exchange is at 6:40 mark):
Question: “Of the 59 shots you took tonight, 30 were from midrange. Is that what you guys were wanting to do?”
John Calipari: “We had 19 threes. We took 19 threes. And we wanted the midrange shots, yes. Because that’s how they play, and it’s one of the things we do so well. We missed a bunch of' ‘em today. Just missed a bunch. That’s a shot we make. Matter of fact, we play to get those, too.
John Calipari is a very good basketball coach, and I’m sure this didn’t come out exactly as he wanted it to given that his team just lost a tough game, but this answer had a defiance to it that struck me as very odd. The primary reason is that, while Calipari is telling the truth about UK playing to get midrange shots, they probably shouldn’t be. Simply put, UK takes more shots from midrange than just about any team…year in, year out. That’s despite the basic math showing that midrange shots are just about the most harmful thing you can do on offense besides turning the ball over.
I’m not sure Calipari is seeing the right data on midrange shots, quite frankly; he may not really be looking at any data. For example, against Vanderbilt UK was 6-18 on jumpers between 4 feet and the three point line. Calipari says “we missed a bunch”, but that’s 33%; for the season UK hits 38% on these. UK absolutely did not miss a bunch more midrange shots than usual against Vanderbilt. The Commodores went 2-10 themselves on these shots, so if either team had an unusually bad night from midrange it was Vanderbilt. The overall data on UK and midrange shots tells two things very clearly: first, UK takes a larger percentage of their shots from midrange than almost any team and second, the Wildcats score less effectively on these shots than on other types. I’ll lay out the data and explain why I’m so confused by UK’s strategy of living in the midrange, in this free Hoops Insight article.
A pattern of relying on midrange shots
I believe John Calipari when he says UK plays for midrange shots, because otherwise something odd must be going on. The Wildcats diet of halfcourt shots has consistently been skewed to midrange ever since Calipari took over. Hoop-math.com has data doing back to the 2012 season showing the percentage of a team’s shots coming at the rim, from midrange, and from three. Here are UK’s annual ranks from each location:
That’s 6 straight years UK has been in the top 20 in frequency of shots from midrange, and 10 times in the top 80 in 12 years. In the last 5 years and 7 of the last 9, Kentucky has ranked 220th or lower in frequency of shots at the rim. And of course, UK has never ranked higher than 268th in frequency of three pointers in the last 12 seasons, ranking 331th or lower in the last 6 seasons.
Most of D-1 has been exchanging midrange shots for threes over the last decade. In 2012, the % of shots from midrange for an average team was about 32%, while the average % of shots from three was about the same at 32%. Things felt balanced, people were watching “The Dark Knight Rises” in theaters and listening to “Call Me Maybe” by Carly Rae Jepsen….2012 was lovely. By 2017, things were shifting a bit. The average % of shots from midrange had dropped to 28%, and the % from three was up to 36%. Things were still lovely but starting to feel tilted towards the deeper shot. Opinions were divided on Star Wars Episode VIII, and there was a lot of Imagine Dragons on the radio. Now in 2023, the geometry of the court has clearly shifted to favor threes over midrange. Only about 25% of an average team’s shots come from there, and only 1 team (CSU Bakersfield) takes more than 39% there…in 2012 63 teams did! Now the average teams takes 38% of their shots from deep, a number only 49 teams reached in 2012. It’s a “Top Gun: Maverick” world, and I don’t even know what the kids listen to know (Harry Styles? Taylor Swift’s latest? TikTok?).
Over that same time, Kentucky has resisted the tides of change like a rock wall. The 2012 Wildcats took 38% of their shots from midrange and 27% from three. Now in 2023, UK’s split is 36% from midrange, 30% from three. Only 39 teams take more shots from midrange than three, and among Power 6 conference teams only Ohio State and Georgetown have a bigger difference between the two than Kentucky.
The simple thing to do would be to assume that midrange shots must be good for UK. After all, John Calipari is a very good and successful coach, and it’s pretty easy to assume he’s making the right decisions. However, I believe that asking questions is useful. It’s not terribly hard to figure out how much a midrange shots helps UK compared to other shots…so let’s do that.
The value of midrange shots
Let’s just start with basic field goal percentages and look at how often a midrange shot goes in, compared to a shot at the rim and a three pointer. I’ll use actual data for UK from this season:
At the Rim: 63.3% FG%
Midrange: 36.7% FG%
Three: 36.1% FG%
Midrange shots go in at a rate barely higher than three pointers! Extensive research in the NBA and college has found this pattern persists all over. While threes are deeper shots, midrange shots tend to be more closely guarded, and the mechanics of a jumper are pretty much the same from 12 feet as 22. Because threes are worth 50% more points than midrange shots, it’s pretty clear that threes are more valuable, strictly in terms of points scored per shot attempt. On average a midrange shot gives UK 36.7% x 2 points = .734 points per shot, while a three gives UK 36.1% x 3 point = 1.083 points per shot. Of course, shots at the rim trump them all at 1.266 points per shot.
Kentucky is an excellent offensive rebounding team, though. Maybe a lot of the value of midrange shots is that they are rebounded more often by the offense? I crunched the data on UK’s offensive rebounding by shot location this season:
This surprised me. It's logical that shots at the rim get rebounded at a high rate, but UK has actually rebounded threes at a higher rate than midrange shots this season! The reason has to do with the fact that Oscar Tshiebwe is a fairly frequent midrange shooter, and when he’s shooting those UK’s best rebounder is out of position. Here’s how UK does when Oscar takes shots vs everyone else:
UK rebounds a much lower percentage of Oscar’s shots, because he’s usually the one getting offensive rebounds. However, when Oscar’s teammates shoot, UK collects offensive rebounds on midrange shots and threes at very similar rates.
So midrange shots go in about as often as threes, are worth fewer points per make, and UK rebounds them at almost identical rates. UK scores about 1.01 points per possession after collecting an offensive rebound, so we can calculate the average points scored per each UK shot (including offensive rebounding value). It looks like this for each location:
It’s not even close. Against Vanderbilt, UK took 30 midrange shots, 10 at the rim, and 19 from three. Based on their average FG% and offensive rebounding rates, you’d expect the Wildcats to score about 69 points. If UK took just 20 midrange shots and 5 more from three and at the rim, you’d expect the Cats to score 73 points. That’s 4 points gained just from better shot distribution.
What’s really wild is that, because UK collects offensive rebounds at such low rates when Oscar Tshiebwe shoots, he’s not especially productive for UK as a shooter:
Oscar finishes at a high level at the rim, and that’s tremendously valuable for UK, but the fact that his teammates shoot a lower percentage if almost completely offset by UK’s offensive rebounding. On midrange shots, Tshiebwe shoots a little better but UK loses a ton of rebounding value, and a Tshiebwe midrange jumper is the least productive shot among the chart above.
So why does UK shoot so many midrange shots? I honestly don’t know, but I can’t imagine they’ve looked at the data this way and decided it’s a good idea. The only other factors that could come into play are if the same actions that lead to midrange shots draw a lot of free throws or lead to fewer turnovers. Midrange shots certainly don’t draw shooting fouls at a high enough rate to make them worth much more (UK has drawn 10 shooting fouls all season on midrange shots, by my count). In order for midrange shots (at 0.97 pts/shot) and threes (at 1.33 pts/shot) to be worth the same per possession, teams would need to commit over 3x as many turnovers when setting up threes as they do setting up midrange shots. It’s pretty much impossible to know this for sure since we can’t know what a team was trying to do when they turned the ball over, but SEC teams who shoot a lot of threes (Alabama, Missouri, Vanderbilt) haven’t been any more turnover prone than SEC teams who avoid threes (Mississippi, Arkansas). Also, despite UK ranking in the top 20 in midrange frequency the Cats haven’t been in the top 60 in turnover rate in the past 6 years, and 4 times they’ve been worse than 120th.
It’s baffling, but the data is pretty plain. It’s a pretty common conclusion in the basketball world over the past decade that midrange shots are largely unproductive, unless they are taken by extremely skilled players who hit them at 45-50% or more. For some reason Kentucky has decided to lean into taking midrange shots, and to be fair UK’s offense is largely in the top 25 nationally. However, that’s because of offensive rebounding, not because of shooting, and UK’s offense could routinely be one of the 2-3 very best in the nation if they had better shot selection. Maybe that’s the difference between a narrow NCAA Tournament elimination and survival.