What Makes for a Bad Matchup for UofL?
A Rough Stretch for the Cards...
Losing 3 out of 4 (and nearly 4 out of 4) has dampened the enthusiasm of Cardinal fans who had been eagerly anticipating a top-4 seed in the NCAA Tournament.
UofL Stats Profile vs Clemson, Feb 16th
UofL Stats Profile vs Syracuse, Feb 20th
The Syracuse loss in particular was a bad one, and may have discouraged Cardinal fans. But according to my analysis, Syracuse is possibly the worst matchup of all UofL opponents this year. That may seem surprising, but let me walk you through my work to see what makes a bad matccup for UofL.
What Makes for a Bad Matchup for the Cards?
First, we can look at what UofL is really good at this season. The Cards are among the top 100 nationally (per KenPom.com) in effective Fg% defense, 2 point and 3 point shooting defense, offensive rebounding, free throw rate and free throw shooting. They are among the bottom 100 in having their shots blocked and generating steals; they are also poor in blocking opponent shots and forcing turnovers. That gives you an idea of how they've done in these stats, but it also leads to a philosophical question. Does UofL play better when their strengths match the opponents' weaknesses, or when their strengths counter the opponents' strengths? In order to answer that, we need to bust out some statistical analysis.
I like to use adjusted margin to evaluate performances because it adjusts for opponent strength and is a pretty good "common currency". We can use regression to determine how much opponents' ratings in key stats impact UofL's adjusted margin in a game, and determine what UofL should look for in a good or bad matchup.
***NOTE: Skip this next paragraph if you feel overwhelmed by stats talk***
To do this analysis, I used the KenPom rankings in the Four Factors categories for offense and defense. I converted the raw numbers into standardized z-scores to state everything in terms of the number of standard deviations away from the mean. Then, I ran a regression using each Four Factor stat on offense and defense as the independent variables, and UofL's adjusted margin per possession as the dependent variable.
***END of super-wonky talk***
Doing this yielded some interesting results. Per the regression, a good matchup for UofL is a team who is:
Above average in shooting
Below average in forcing turnovers
Above average in allowing free throws
Below average in turning the ball over
What this tells us is that UofL is at its best this year when it can match its strengths against opponent strengths. UofL's strengths seem to legitimately come from imposing its will, and the Cards are able to do this to take away what an opponent likes to do. The Cards also are at their best when an opponent isn't able to force a lot of turnovers, and when the opponent is more likely to turn the ball over themselves. UofL isn't going to be able to impose their will in these areas, so it's best for the Cards to benefit from the opponent weakness.
Looking at UofL's opponents so far this year, the opponent who most clearly met all 4 of these criteria was Michigan State. They rank 14th in shooting and 64th in allowing free throws, but only 200th in committing turnovers and 326th in forcing them. UofL held Michigan State below their season average in eFG%, drew 41 free throws, and won the turnover battle 10 to 17. While it seemed like a huge upset at the time, it doesn't seem so surprising now when looking at UofL's season and these regression results. Other opponents who stand out as good matchups are Marquette, UNC, and Virginia Tech; UofL has had strong performances against all of these teams this year.
On the flipside, one opponent stands out as meeting all 4 criteria for being a bad matchup for UofL: Syracuse. The Orange rank only 244th nationally in shooting and 195th in allowing free throws, but they rank 135th in committing turnovers and 10th in forcing them. Syracuse wins by drawing free throws, limiting opponent shooting, forcing turnovers, and blocking shots; this amplified UofL's weaknesses as UofL struggled to shoot, turned the ball over on 20% of their possessions, and had 5 of their 26 2 point shots blocked. Syracuse is used to winning despite shooting poorly and letting opponents get to the foul line, and that's what happened last game. Two surprising teams who profile as bad matchups for UofL are Nicholls State and Kent State; in both games UofL struggled more than expected against relatively soft opponents.
Looking forward to the rest of the regular season:
Virginia is a bad matchup due to their amazing ability to avoid turnovers
Boston College is a relatively good matchup due to not being excellent in any one area, and rather poor at forcing turnovers
Notre Dame is a bad matchup due to their ridiculous ability to avoid turnovers, and ability to play well when shooting poorly...in fact, my analysis has Notre Dame as an even worse matchup than Syracuse!
I wouldn't be surprised or too discouraged by a 1-3 finish to the regular season. Even with a poor run heading into the ACC Tournament, there are good matchups available. Cards fans should hope for matchups against teams like Virginia Tech, Georgia Tech, Clemson, and North Carolina, who all have profiles that UofL can match up well against.
Thanks for reading my newsletter. If you have any questions, want to argue a point, or have some feedback, feel free to reach out via email at sean@hoopsinsight.com, or on Twitter @hoopsinsights. If you liked this, let me know as well, and tell your friends to subscribe at www.hoopsinsight.com.
You can forward this to others, but please ask them to subscribe as well so I can keep track of who's enjoying my insight and analysis