Kentucky is wasting a perfectly good defense
Kentucky is defending at their normal strong level, but the offense is a complete disaster. How bad is it, and what could help fix it?
I’m going to have to respectfully disagree with John Calipari when he cites “toughness” as the problem holding back Kentucky basketball this season. Sure, the Wildcats have lost 3 1-point games where just a little better focus would have put them over the top. The defensive breakdown on the last play against Georgia was a complete failure to understand and process a late game situation. But to focus on those individual opportunities is to miss the overarching theme to this season for Kentucky, which is that this team is good enough on defense to be at the level Big Blue Nation is accustomed to, but is a complete trainwreck on offense. Today I’ll explain how bad it is, and offer a suggestion of how to improve.
The defense is fine! Really!
As of this writing, Kentucky ranks 20th in the nation in defensive efficiency per KenPom.com. That is plenty good enough, even by the high standards to which Big Blue Nation is accustomed. That would be the 7th best in Calipari’s tenure at UK.
Kentucky has some areas on defense where they are legitimately good! They generally avoid giving out free throws (especially by Kentucky standards), they defend the three fairly well, and are elite at blocking shots. They also force turnovers at a better clip than many recent UK teams. It’s not all great, as they are the worst defensive rebounding team under Calipari at UK, but overall they are able to make life difficult for opponents. They had the best defensive effort of the season by an opponent against Morehead State or Florida, and the 2nd best by an opponent against Richmond, Kansas, and Louisville. In short, UK has been able to defend at an elite level, even against its toughest opponents.
The offense is not fine, however, it’s awful
The offense is another story. Kentucky ranks 98th in offensive efficiency, easily the worst figure of any Kentucky team since the coach prior to Calipari (who I will not name out of decency!). Kentucky is performing worse than they ever have under Calipari in several advanced stat categories:
Effective FG% (45.9%, 2nd worst is 49.9% in 2014)
Turnover rate (22.0%, 2nd worst is 20.3% in 2010)
Free throw attempts per FGA (.324, 2nd worst is .361 in 2011)
All this adds up to disaster. Per KenPom this season, Kentucky has had the worst offensive game by an opponent against Richmond and Notre Dame, the 2nd worst by a D-1 team against Kansas, and the 3rd worst against Auburn. Among the teams who have performed better on offense than UK against these opponents are Texas Southern, North Dakota State, Morehead State, Detroit, and Bellarmine.
I could go on, but the horse has been long dead and I’m tired of beating it. I’ve seen people attribute the problems to roster talent, coaching, focus, heart, leadership, COVID-19, and everything else you could imagine, but I am here to propose a somewhat radical but effective solution that could help the Wildcats improve their offense without needing them to vastly improve.
Offensive structure can paper over the issues
Nobody can wave a magic wand and make this team stop turning it over, or start hitting threes, or draw free throws. What could be done, if one was so inclined however, would be to change the offensive structure and gameplan to become much more efficient
The inspiration for this idea comes from what Nate Oats is doing with Alabama basketball. This season, Alabama has decided to take field goal attempts at the rim or three pointers, and almost nothing else. Here is their shot distribution:
43% of their shots are at the rim, where they hit 57%
47% of their shots are from three, where they hit 36%
10% of their shots are in between, where they hit 25%
Alabama is not especially elite at the rim or from three. Per hoop-math.com, They are 262nd in FG% at the rim and 87th from three. They are not good in transition, with an eFG% of 50% (276th). But in halfcourt offense they are 62nd in eFg%, and they rank 13th in offensive efficiency. It’s because they take shots are likely to create points, plain and simple. Why take shots that generate 0.50 points (like midrange shots where you hit 25%) when you can take shots worth 1.14 (at the rim) or 1.08 (from three)?
Kentucky takes a differnt approach. Here is their shot distribution:
32% of their shots are at the rim, where they hit 61%
28% of their shots are from three, where they hit 29%
40% of their shots are from midrange, where they hit 36%
Kentucky is not good at shooting from three, but they generate more points per shot from there (0.87) than from midrange (0.72). They are even better at the rim (1.22). Every time Kentucky takes a midrange shot, they are missing a chance to to take a higher value shot. The result is that Kentucky is 302nd in eFG% in halfcourt offense, and 98th in offensive efficiency. But what if Kentucky improved their shot distribution?
Kentucky could structure their offense like Alabama’s and try to match their shot distribution. It clearly is possible, because Alabama does it. How would Kentucky’s offense improve if they took more shots at the rim and from three, and much less from midrange?
Even assuming Kentucky does not improve their shooting from any one area, their overall shooting would drastically improve. Kentucky’s effective FG% would improve from 45.9% (313th in the country) to 51.2% (about 150th), even with one of the worst 3 point percentages in the country! Kentucky’s offense would improve by about 8.5 points per 100 possessions, moving them up from 98th to about 22nd. That would make Kentucky very similar to the 2018 Wildcats, who ranked 24th in offense and 22nd in defense.
I even went a step further and looked at how each game would have changed if Kentucky had redistributed their shots in the manner laid out above. Some games would still be disasters due to horrendous three point shooting (Georgia, Richmond). But when I ran the numbers, it had Kentucky picking up wins against Kansas, Notre Dame, and Louisville in games they actually lost. That would put UK at 7-6 (3-3 in SEC), on a 3 game losing streak still but with 2 quality non-conference wins and a realistic chance of an NCAA Tournament bid. All from making better choices in how to structure offense…now who wouldn’t take that right about now?
Good stuff.
To play devil's advocate, perhaps this just kicks the can down the road as an explanation. The question becomes: how? They're not really passing up open 3s, and it's hard to generate looks at the rim when your only dynamic penetrator is wearing a walking boot. I take your point that they don't really have to be good at any of those, just a better distribution by itself would be a plus.
But I think it's high time to talk about tempo. I've seen it commented upon to a degree over the years, but it has still flown under the radar somewhat the degree with which Cal has turned his offense into legit "slow". He's no Tony Bennett, but with exception of the 2010 and 2018, years when he had two of the fastest PGs in CBB history, he's played at a below average pace, getting slower over time, and is the slowest ever this year.
This has always been a curious choice to me: when you have superior talent and athleticism, you want to maximize your advantage by increasing the number of possessions, and take luck out of it. But you can't really argue with the previous results.
However I think it's possibly a disastrous choice for this group. Slow offense requires you take care of the ball well (more dribbling and passing gives more time to turnover) and execute to get open shots later in the clock (again, penetrator with a boot). They do neither. My hypothesis: a lot of those midrange attempts are choices forced upon them by draining the shot clock with false action and taking forever to get into the offensive movement. Instead, even when not on break, they should be forcing the action in the first few seconds, when defenses are back but before they are truly set and ready to help on drives/posts.
It's really hard to test, because when shots happen with regard to the shot clock is not a random variable, as you of course know. I'm not really sure how to test. Maybe look at possessions after the other team has scored? Or at least, defensive possessions where no turnover occurred?
Do you think there is anything there? How would you approach looking at this?