This is the 2nd of 4 articles about the 2014-15 Kentucky Wildcats. Part 3 will be published on Mon, Oct 26, and Part 4 on Thurs, Oct 29. Check my archive for previous articles.
Before the 2014-15 season, fans were eager to see how John Calipari would manage the sheer amount of talent on the Kentucky Wildcats. 9 McDonald’s All Americans! 4 highly touted freshmen joining 3 returning starters and 2 accomplished bench contributors! 7 players 6’8” or taller, and only 2 shorter than 6’6”!
During a preseason exhibition schedule in the Bahamas, the platoon system was unveiled.
2 groups of 5 swapped in and out in hockey-style line changes. The Blue platoon started games and consisted of Andrew Harrison, Aaron Harrison, Alex Poythress, Willie Cauley-Stein, and Karl-Anthony Towns. The White platoon would relieve them, and consisted of Tyler Ulis, Devin Booker, Trey Lyles, Marcus Lee, and Dakari Johnson. Due to the confluence of a preseason #1 team and such an uncommon substitution pattern, the Wildcats and their platooning were the talk of the preseason. Once the season got rolling, 2 major factors caused the platoon system to fizzle out: foul trouble and injuries.
The best laid platoons often go awry….
During the first 9 games of the season, the two platoons were far and away the most common lineups for Kentucky:
Even the third most played lineup was effectively an alternate Blue platoon, as Dominique Hawkins filled in for Alex Poythress due to injury in 2 games. However, this left 236 total (non garbage time) possessions that were being filled by non-platoon lineups.
The primary reason was foul trouble. In 5 of the first 9 games for UK, at least 1 player ended up with 4 or more fouls. This is in spite of the fact that no Wildcat played more than 26 minutes in any game until the 8th game. In each of these 5 games, non-platoon lineups played more possessions than any individual platoon lineup. In the other 4 games, platoon lineups each played more than all combined non-platoon lineups. Interestingly enough, these makeshift non-platoon lineups actually performed better than either platoon:
None of these lineups performed poorly; even the Alt Blue platoon’s +31.7 adjusted margin (click here for an explainer of adjusted margin and other stats I reference) would have ranked in the top 5 KenPom ratings for any season since 2002.
The platoon system was pretty clearly a clever way to manage playing time. Everyone was designed to get about 20 minutes of playing time, and when given a chance it mostly worked. It removed the need for the coaching staff to track minutes and adjust lineups on the fly. But it didn’t really do much to maximize Kentucky’ chances of winning. A platoon system is most useful when a team has a single lineup or two that vastly outperforms the other options. For example, let’s say Fake State has a starting lineup that dominates together, but when even 1 of them leaves the game, they struggle mightily. That team would be best served by substituting their starters in and out all together, so that the starters only play together. They may take their lumps when the starters sit, but if they can have their best lineup together for 30+ minutes they will maximize their chances of winning.
Kentucky went into each early season game planning to lean on the platoon system, but they had the ability to still field great lineups even in the event of foul trouble or injuries. That’s the advantage of having a roster full of McDonald’s All Americans!
The tallest team gets a little smaller…
In the 10th game of the season, against Columbia, Alex Poythress suffered a season-ending knee injury. Rather than continuing to platoon with another player taking Poythress’ spot, Kentucky largely abandoned the platoon system. One of the oddities of UK’s roster composition that season was that they had several guards, power forwards, and centers, but almost no wings. Alex Poythress (6’8”, 238lbs) and Trey Lyles (6’10”, 235 lbs) masqueraded as small forwards; both later would play power forward in the NBA. With Poythress’ injury, Kentucky was down to 1 quasi-small forward. The response was to play 3 guard lineups much more often:
Prior to Poythress’ injury, Kentucky had played 3 guard lineups for just 100 of their 600 non-garbage time possessions. That is even inflated a bit because 36 of those possessions saw Dominique Hawkins filling in as a third guard and nominal starter for Poythress; Hawkins was essentially out of the rotation other than these initial appearances. For the season, over 75% of Hawkins non-garbage time minutes came as a part of the starting lineup. Kentucky would start with 3 guard lineups but when Hawkins was first substituted out of the game he seldom returned.
…and gives up some of their advantage
During the final 30 games of the season, Trey Lyles was elevated into the starting lineup for 21 games. Whenever Lyles was out of the game, Kentucky would feature some combination of 3 of the Harrison twins, Booker, and Ulis. These 3 guard lineups were largely just as effective as 2 guard lineups, albeit in different ways:
2 guard lineups clearly benefitted from the extra length in the frontcourt. They were more accurate on 2 pointers, defended opposing 2 point shots better, fouled less often, and rebounded better. 3 guard lineups made up the difference by winning the turnover battle and drawing more fouls. Both groups played at a very high level, as measured by adjusted margin.
In my introductory article about the 2014-15 Wildcats, I laid out how their extraordinary length contributed to their historic greatness (especially on defense). When we look at their halfcourt shooting with 2 guards vs 3, we can see some of the impact from losing that length advantage.
With 3 guard lineups in, Kentucky swapped some midrange shots for three pointers. That’s normally a good tradeoff, but a drop in accuracy at the rim and in the midrange gave back the benefits. On defense, Kentucky clearly lost some rim protection despite continuing to feature some monster shot blockers even with 3 guard lineups. 2 guard lineups didn’t give up any perimeter defense with their additional size, either. Any benefit from playing 3 guard lineups was almost entirely due to committing fewer turnovers and forcing more from opponents.
More pressure on the coaching staff to juggle lineups
Without the platoon system, Kentucky’s coaching staff was much more flexible in which lineups to play. If you take a wide view of the season, there weren’t many bad options. Kentucky had 15 lineups who played 40 or more non-garbage time possessions after Poythress’ injury, and 14 of them outscored their opponents; the 15th was outscored by 1 point. From that perspective, Kentucky almost couldn’t go wrong no matter which lineup they leaned on in any particular game. But taking a closer look reveals some challenges in how to allocate playing time.
During the first 9 games, Kentucky had 3 different lineups that were the most played in any individual game:
The Blue platoon (5 times)
The White platoon (3 times)
The alternate Blue platoon with Hawkins in for Poythress (1 time)
The latter only saw so much time due to Poythress’ injury; otherwise, the Blue and White platoons were the most used lineups.
In the last 30 games, Kentucky never really seemed to get the same handle on the lineups they wanted to lean on. 16 different lineups finished games with the most possessions played (or tied for the most), excluding garbage time. The only lineup to do so more than twice was the most common starting lineup over this stretch, the Harrisons, Lyles, Cauley-Stein, and Towns; it was the most played lineup in 14 games. Lyles missed three games due to injury, but in the 27 other games it was 50/50 whether this lineup would be the most played. There were another 6 games where it finished as the 2nd most played, so there were 7 times when this lineup did not finish as one of the 2 most played in a game. While this was clearly the most preferred lineup during this stretch, it was far from a lock for Kentucky to lean on them. So who did Kentucky turn to when they needed a boost?
In order to determine which lineups the coaches trusted the most, we can look at which lineups were played when 1) the score was within 10 points either way, and 2) the game was in the final 10 minutes or overtime. Kentucky didn’t have a ton of close games, but the most played lineups in these situations were:
In these situations, Kentucky leaned even more heavily on 3 guard lineups with the Harrisons and either Ulis or Booker alongside Cauley-Stein and Towns. The lineup with Lyles at SF was the 3rd most used here.
We’ll never know how Kentucky’s season would have gone had Poythress not gotten hurt. Maybe Kentucky would have tried to continue the platoon system and benefitted from continuity. Maybe they would have gotten away from it anyway. It does seem that Kentucky never really settled into who their best lineup was, and relied more on their overall team brilliance than any singular group of players.
In my next article I’ll look at Kentucky’s guard rotation, how the rotations changed through the year, and how the pieces seemed to fit together in the backcourt.
ncG1vNJzZmign6S9tLXNrKCgoKRjwLau0q2YnKNemLyue89oaWlpZGJ%2BdnnKnqWtrZOgxm7DyKWbnJmkqHqttc2erKmr