2015-16 Louisville Cardinals: More guards for the Cards
UofL fared much better when playing with 3 guards as opposed to just 2 during the season. What was the key that made 3 guards so good?
This is the 3rd in a series of 4 articles about the 2015-16 Louisville Cardinals. Part 4 is coming on Thurs Nov 12. Check out my archive for my previous articles and others in the Hoops Hindsight series.
During the 2015-16 season, nearly all of the positive scoring margin the Cardinals ran up against opponents happened when they were playing 3 guards. Louisville outscored their opponents by a combined +368 points for the year, and 362 of that came with 3 guards playing for the Cards. Just 6 came with only 2 guards playing. Louisville played 3 guards most of the time, but did give 17% of the minutes to 2 guard lineups. They didn’t get much return in those 17% of the minutes, but what exactly was the issue?
A leaky defense
We can use adjusted +/- per 100 possessions to see the magnitude of the dropoff. Louisville had an adjusted margin of +33 per 100 poss with 3 guards, and only +7 with 2 guards. That 26 point per 100 possession decline was driven by:
A drop of 5.5 points scored per 100 possessions
An increase of 16.1 points allowed per 100 possessions
3 guard lineups played opponents who were 4.1 points per 100 possessions better
On that last point, each possession’s scoring margin gets adjusted based on how good the opponent is. So, playing tougher opponents increases your adjusted margin, all else being equal. This is because it’s more impressive to post the same scoring margin against a better team.
By far the biggest factor above is the increase of 16.1 points allowed per 100 possessions. The drivers of this were better opponent shooting (mostly on 2 pointers) and better opponent offensive rebounding. 2 guard lineups actually allowed free throws less often, and forced turnovers just as often, as 3 guard lineups. Opponents shot 47% on twos and 34% on threes against 2 guard lineups, compared to 41% and 32% against 3 guards. They also snagged 35% of the available offensive rebounds vs 2 guards, compared to 29% against 3 guards.
Better at defending midrange jumpers
We’ll start with the improved opponent shooting first. The issue wasn’t giving up easy baskets in transition; 2 guard lineups allowed a 46% eFG% there compared to 49% for 3 guards. The issue wasn’t 3 pointers in the halfcourt; both 2 guard and 3 guard lineups allowed 34% from deep. The issue was 2 point shots in halfcourt situations, where 2 guard lineups allowed 46% shooting but 3 guard lineups allowed only 36%.
We can break 2 point shots into 2 categories: At the Rim, and 2 Point Jumpers. When we look at the frequency and accuracy of each, we see the problem with 2 guard lineups:
With 2 guards: 2 point jumpers were 36% of opponent shots, and they hit 45%; shots at the rim were 28% of opponent shots and they hit 49%
With 3 guards: 2 point jumpers were 36% of opponent shots, and they hit 25%; shots at the rim were 27% of opponent shots and they hit 50%
There was virtually no difference in the frequency of 2 point jumpers, or the frequency and accuracy at the rim. The difference is made up by the accuracy of opponent 2 point jumpers.
Two games in particular stand out in this area. When Louisville played Virginia on Jan 30th, the Cavaliers were 5-5 on 2 point jumpers against 2 guard lineups and 1-4 against 3 guards. When Louisville played Georgia Tech on March 1, the Yellow Jacket were 6-7 on 2 point jumpers against 2 guard lineups and 2-10 against 3 guard lineups.
I looked at who took the 2 point jumpers and found something interesting. For the season, the bigs on Georga Tech and Virginia took 51% of their team’s 2 point jumpers. Against Louisville, they took 65%. But bigs hit 8-8 in these 2 games against Louisville’s 2 guard lineups and just 3-9 against 3 guard lineups. I found video of these games and went back to watch, and it appears the issue was the drop coverage Louisville used for its bigs. The Cards had their big men hang back near the rim, and had their guards and forwards pressure midrange shooters. But there were big differences in how effective Louisville’s players were at guarding spot-up shooters.
Per Synergy, the 2 best Cards defenders on spot up shooters among the non-bigs were Quentin Snider (84th percentile) and Trey Lewis (62nd), and the 3 worst were Ray Spalding (16th), Jaylen Johnson (13th), and Deng Adel (5th). In 3 guard lineups, Louisville played Snider and Lewis together for 51% of the possessions, but in 2 guard lineups they only played together for 26%. Meanwhile, in 3 guard lineups Louisville had 2 of the Spalding/Johnson/Adel triumvirate in the game for just 1% of the possessions, but in 2 guard lineups Louisville had at least 2 of them in for every single possession. When Louisville played 3 guard lineups, they had more of their good defenders for mid-range spot ups and fewer of the bad defenders.
Bigger didn’t mean better when it came to rebounding
Teams which play a lot of 3 guard lineups usually give up something in terms of rebounding. I know that’s not exactly the sharp insights you expect when you read this, but it is just common sense. It’s not always true, however, and it certainly wasn’t with the 2015-16 Louisville Cardinals.
When Louisville went to 2 guard lineups, they inserted Deng Adel at small forward in place of one of the Lee/Mitchell/Snider/Lewis group of guards. At 6’7” 200lbs, Adel was a good bit bigger than Mitchell, Snider, or Lewis, and roughly the same size as Lee. Louisville kept their usual cadre of forwards and bigs at the 4 and 5 position. So why is it that their opponents were so much better at offensive rebounding against these “bigger” lineups?
Similar to the shooting defense analysis above, the answer seems to be related to Louisville’s defensive scheme. Louisville frequently dropped their big man to hang near the basket, and asked the other players to step up to pressure the ball. This type of scheme is more effective when you have quicker, more athletic guards doing the pressuring as opposed to forwards. When Louisville had three guards, the forward and big could hang closer to the basket, knowing that the 3 guards could apply defensive pressure. When Louisville swapped a guard out for Deng Adel, they lost some of that effectiveness, and the forwards were more frequently out of position for rebounds. The power forwards tended to see their rebounding drop off the most when Louisville played 2 guards:
Jaylen Johnson got 9% of available defensive rebounds as a PF in 2 guard lineups, but 13% in 3 guard lineups
Ray Spalding got 15% of available defensive rebounds as a PF in 2 guard lineups, but 18% in 3 guard lineups
You might expect that the PFs rebounded more in 3 guard lineups because they needed to pick up slack for the guards, but that’s not true. In 3 guard lineups, Louisville’s guards combined to collect 32% of defensive rebound opportunities; in 2 guard lineups, the guards + Adel combined to collect….32% of defensive rebound opportunities. Louisville’s big men also had almost no drop off in rebounding, collecting 24% of rebound opportunities with 3 guards and 23% with 2 guards. The decline in defensive rebounding can be traced directly to Louisville’s power forwards, who had to take on some additional defensive responsibilities when playing only 2 guards.
As good as the 2015-16 Louisville Cardinals were, they had another level to get to if they committed to playing 3 guards full-time. Any time they spent without 3 guards in the game was basically their version of treading water. In the next edition of Hoops Hindsight, I look at the big question that lingered in the heads of every Cardinals fan that season: what would Louisville have done with a postseason?