2014-15 Kentucky Wildcats: The Backcourt Rotation
Kentucky had 4 guards who all made the NBA. How did they balance the backcourt combinations, and which ones worked best?
This is the 3rd of 4 articles about the 2014-15 Kentucky Wildcats. Part 4 will be published on Thurs, Oct 29. Check my archive for previous articles.
In the 2014-15 season, Kentucky had 4 guards who each averaged at least 20 minutes of playing time: Andrew Harrison, Aaron Harrison, Devin Booker, and Tyler Ulis. All 4 eventually made the NBA. Ulis was the only one not to turn pro after the season, and all he did in 2015-16 was win SEC player of the year, SEC defensive player of the year, and make 1st team All America. There weren’t many bad answers to the question of “which guards should we play?”, but that doesn’t mean there weren’t hurdles as the coaching staff managed the backcourt rotation.
2 guards or 3? And which ones?
Under normal circumstances, a team with 4 future NBA guards would try to get as many of them on the court as possible. But Kentucky’s 2014-15 roster also had 5 future NBA players in the frontcourt. No matter the lineup choice, several future pros were going to need to sit.
In my last article, I noted that Kentucky began the season with Alex Poythress and Trey Lyles alternating as super-sized small forwards in platoon lineups. This maximized Kentucky’s length, and paid dividends especially on the defensive end. Once Poythress suffered a season-ending knee injury, the plan shifted to incorporate more 3 guard lineups:
The question remained, however, about which combinations to play. The Harrison twins had been the starting backcourt a year prior and led Kentucky to a title game appearance, but Ulis and Booker brought their own accolades. The 4 guards could complement each other well, with Andrew and Ulis serving as point guards and Aaron and Booker bringing more of a classic shooting guard skill set.
Interestingly enough, Kentucky’s coaching staff didn’t really mix and match the 2 guard combinations very much:
When Kentucky played 2 guards, 86% of the time they went with the Harrisons or Ulis/Booker. Occasionally they would cross-match, but relatively rarely.
Both effective, but one combo was a better fit…
Both of the main combinations used were very effective, although the Ulis/Booker tandem contributed a bit more because of their offense:
However, these numbers don’t provide an apples-to-apples comparison because of how Kentucky used their platoon system early on. Ulis and Booker played more frequently against bench lineups. When looking just at how each tandem performed when the opponent had 4 or 5 starters in the game, we get a different picture:
When playing against opponents’ starting lineups, the Harrisons were still very effective but Ulis & Booker struggled more. While the Ulis/Booker duo still shot well from the outside, they took relatively few three pointers. They also took free throws less often than opponents, giving up a valuable source of offense.
A search for 3 complementary guards
When playing 2 guards, Kentucky had many options to match complementary skill sets. As the table above showed, Kentucky almost never played Ulis & Andrew, or Booker & Aaron, because their skill sets were too similar. But when playing 3 guards, Kentucky would have to have either 2 of their ballhandlers (Andrew/Ulis) or 2 of their shooting guards (Aaron/Booker) sharing the court. Which combos were the best? Let’s look at the team stats for each of the 4 main options:
There’s definitely more separation between good and bad than there was with the main 2 backcourt duos. Looking at adjusted +/-, we can see 2 clear patterns: 1) lineups with Ulis/Booker are better than the Harrisons, and having Booker & Aaron on the court together is not a great idea. Because most of Kentucky’s use of 3 guard lineups came after the platoon system was scrapped, there isn’t a significant bias in terms of playing against starters or bench lineups; each of these 3 man units played about 48-50% of their time against 4 or more opposing starters.
3 guard lineups which featured Ulis/Booker shot much better from three point range than those without that duo, and made use of that by taking threes more frequently than they did in 2 guard lineups. A major component of this success was Tyler Ulis’ versatility to be a distributor or a shooter. Ulis hit 43% of his threes on the season and was rated by Synergy in the 96th percentile nationally as a spot-up shooter, so he could play as a second spot-up shooter alongside Booker when they teamed with Andrew Harrison. Ulis ranked in the 99th percentile in halfcourt offense when factoring in his passing, so he could also serve as the facilitator with Booker and Aaron Harrison on the wings. While Arron Harrison shot only 32% from 3 point range for the season, that rose to 39% when he played alongside Ulis and Booker. So Ulis could either serve as a perimeter threat, or feed 2 other perimeter threats. Neither Andrew or Aaron Harrison brought this flexibility, as Andrew wasn’t the outside threat Ulis was and Aaron wasn’t anywhere near the distributor.
Three guard lineups with Aaron Harrison and Devin Booker suffered from a redundancy problem. The problem can be illustrated by looking at the most common offensive play types each ran for Kentucky, and their percentile rankings per Synergy:
Aaron Harrison: #1 Spot Up (56th), #2 Transition (59th), #3 Off Screen (51st)
Devin Booker: #1 Spot Up (89th), #2 Transition (85th), #3 Off Screen (56th)
They ran the same play types, but Booker was better at each and significantly so for the 2 most common. For contrast, here are the most common play types for Andrew Harrison and Tyler Ulis:
Andrew Harrison: #1 P&R Ball Handler (63rd), #2 Spot Up (69th), #3 Transition (56th)
Tyler Ulis: #1 P&R Ball Handler (45th), #2 Spot Up (96th), #3 Transition (49th)
While Andrew and Ulis overlapped, Ulis was such an elite spot up shooter that it could work. Additionally, Ulis was an excellent distributor, which Aaron Harrison was not. When Kentucky played Andrew, Booker, and Ulis, they could offer a flexible and complementary package of skills. When Aaron and Booker played, they duplicated and minimized each other.
Getting better as they go
Kentucky’s coaching staff seems to have recognized these patterns and adjusted as the season went along. The Andrew/Ulis/Booker trio improved as the season went along, likely due to more familiarity and the natural improvement of freshmen as they become more comfortable. Over the first 16 games of the season, they only had 1 game where they played at least 10 possessions together (vs UNC); over the final 23, they did so 14 times. Over these same 23 games, they put up an adjusted margin of +50 points per 100 possessions! The Harrisons + Ulis played at least 10 possessions together only once before February, but did so in 10 of the final 17 games; they posted an adjusted margin of +42 per 100 during this closing stretch!
On the flip side, the Aaron/Ulis/Booker trio played together for at least 16 possessions 6 times before January 24th, but only once the rest of the season. The Harrisons and Booker played more than 10 possessions together only 7 times all season, and did not do so even once during postseason play.
Earlier I noted that Kentucky played 3 guard lineups much more as the season went on, and they clearly adjusted to play better 3 guard lineups. They got better results from these lineups as the season progressed. 3 guard lineups were outscored by 6 points against UNC on Dec 13, by 16 points against Ole Miss on Jan 6, and by 6 points against Texas A&M on Jan 10. Over the next 23 games, UK’s 3 guard lineups outscored their opponent every single time. The 24th game, and the final one of the season, was the only other time UK’s 3 guard lineups were outscored. But we’ll dive more into that game in the next Hoops Hindsight article on Kentucky’s 2014-15 season…