Louisville is Elite, 3/4 of the Time
If UofL's regular season had ended on February 10th, they would have completed 3/4 of their season. At that point they were 17-7, and per BartTorvik.com, they rated as the 15th best team in the country. They would have likely been a top-4 seed in the NCAA Tournament.
If UofL could end all of their games after 30 minutes, they would have a record of 25-6, including wins over Virginia (twice!) and Duke. Their record in their last 8 games would be 6-2 instead of 2-6, with losses only to Clemson and Syracuse. Technically they were tied against Boston College with 10 min left, but UofL scored first after that and won, so I'm giving them the win in this made-up scenario. A 25-6 record with wins over Michigan State, UNC, Virginia, and Duke would probably get the Cards in contention for a top-2 seed!
The point is, UofL would be an elite team in a world where they either played only 3/4 of their schedule, or only played 3/4 of a game. Unfortunately, NCAA rules mandate you finish what you start, and that finishing stretch is what has doomed the Cards. Before I get into my analysis of why, here are UofL's stats profiles for the full season and at the 3/4 mark of the season for reference:
UofL Stats Profile through March 9
UofL Stats Profile through February 9
What You Should Know: UofL's performance this season has fallen off considerably in the final 10 minutes of games and overtime. If we separate games into the Opening (the first 25% of the game, or about 10 minutes); Middle (25%-75%, or about 20 minutes) and End (last 25% of games, or last 10 minutes plus overtime), we can see a stark dropoff at the end of games:
Looking at these numbers, we can see that UofL's defense falls off lat. Opponents shoot better than the Cards, and draw more free throws. But the story gets more interesting when we go deeper. Here are UofL's stats for the end of games, separated by the difference in score:
Looking at these numbers we see a bit of a different story. UofL's defense holds up at the end of games when the score isn't within 10 points (with UofL either up or down by that much). When the game is within 10 points, however, UofL's opponents shoot incredibly well. To make matters worse, when UofL is down by 1-10 points, it's as though the Cards forget how to shoot OR defend! During these stretches, UofL plays like the worst college team in the country, at 48 points per 100 possessions worse than average. Needless to say, this makes a comeback basically out of the question. UofL's adjusted margin per 100 possessions at the end of games when up by 10 or less (+17.4), or when the game is out of hand (+19.0) is not much worse than their margin in the rest of the game, but the Cards are falling apart at the end of games when they are down by 10 or less.
What Is Happening? The biggest difference in these situations is that UofL is playing very different lineups when down by 10 or less, compared to other situations. Over the full season, the lineup UofL has played the most is Cunningham/Fore/Sutton/Nwora/Williams. That lineup has also played the most during the opening and middle segments of games, and they've played very well. That group has an adjusted margin per 100 possessions of +45, and that's consistent across opening and middle segments. But, during the closing 25% of games, that group has played only the 5th most possessions. In the ending segment of games when UofL is down by 10 or less, that group has only played 4 possessions together all season. UofL is instead relying on 2 other lineups to play the most when the Cards are down late. These lineups are:
Cunningham/McMahon/Sutton/Nwora/Enoch, who have played 18 of these possessions and have been outscored by 9 points
Cunningham/McMahon/Sutton/Nwora/Williams, who have played 17 possessions and been outscored by 18 points
It's not entirely unreasonable to lean on these lineups to lead a comeback, since they've been quite good during the first 3/4 of the game. They have an adjusted margin of +31 points per 100 possessions, above UofL's average. This group hits threes at a 39% clip during the opening and middle segments of games. They're not very good at rebounding or forcing turnovers, but the three point shooting makes up for a lot.
The problem is, they've been dreadful during the end of games when uofL needs to make a comeback. As noted above, these lineups are -27 in 35 possessions when UofL is down by 10 or less in the last 10 minutes or overtime. While these lineups are likely being played in order to take advantage of their 3 point shooting, they just aren't getting that done at all. These 2 lineups, who UofL plays more than any others when trailing late, are 0-for-22 this season on three pointers in the last 10 minutes when trailing by 10 or less. This includes a combined 0-for-11 from 3 in the last 10 minutes in late losses to Virginia (twice) and Boston College in the last 2 weeks.
I went back and watched the Boston College game and the 2nd Virginia game to see how these lineups are playing, and if there are any clues as to the cause of their poor shooting. I noticed a couple of things. First, Christen Cunningham is not penetrating off of screens like he does earlier in the game. I'm not sure if this is a function of play design or if he is tired, but he isn't turning the corner to get into the lane when receiving a screen. This leads to him getting trapped or pressured outside the 3 point line, and allows the opponent to keep their defense extended instead of collapsing to stop his drives. This leads to the second thing I noticed, that the three point shots UofL is getting in these late game situations are not open looks off of a pass. The majority of these shots are off-the-dribble threes when guarded, or bizarre pick-and-pop threes from Enoch or Williams. These seem to be lower quality threes than you'd want, and they're just not going in. These are not rushed threes during the dying seconds of the game, either; only one of the attempts I saw came in the last 30 seconds, and that was with 27 seconds left. There should be time for UofL to run good sets and generate open looks, but it's not happening.
What Does This Mean? I can understand why UofL is playing these lineups late in the game when a comeback is needed. As I laid out above, these lineups are great at hitting threes during the rest of the game. However, something is not clicking late in games. TO me, it looks like Cunningham, Nwora, and Sutton are fatigued late in these games. Each of these players averages over 31 minutes per game for UofL, while nobody else averages even 20. It's possible that the demands on them to carry UofL's offense leaves them lacking during the end of games. It's also notable that Ryan McMahon is not much of a factor in these situations; only 1 of the attempts I saw came from him, and it was off the dribble...his catch-and-shoot opportunities vanish late when opponents don't help off of him.
I think UofL should consider keeping Khwan Fore in more in close games late. UofL’s pattern has been to fall behind and not be able to catch up, and the lineups they are using aren’t getting it done. It makes sense for UofL to put what I believe to be their best lineup in the game to close out tough games. While Fore doesn’t bring much shooting, he does bring offensive energy, and the team has been just fine with him in every situation where they’ve used him. His ability to drive might open up good looks for Nwora and Sutton, and he can take some of the ballhandling load from McMahon. This should help UofL hold onto leads and keep small deficits from turning into big ones.
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