If you asked 100 UK fans who the best 5 players on the current roster are, the overwhelming majority would answer, in some order, “Julius Randle, Willie Cauley-Stein, Andrew Harrison, Aaron Harrison, James Young.” Those 5 Wildcats are all considered eventual NBA players, and have started nearly every game together this season. They are the top 5 in minutes per game and points per game, and the lineup of the 5 of them together has played more than 4 times as much as any other lineup. There’s a question nobody’s asking, however: is that truly the most effective lineup for UK? Keep reading…you may be surprised by the answer.
Let’s start by looking at some statistics for UK lineups with all 5 starters, along with statistics for lineups with 4, 3, and less than 3:
5 Starters: Played 155 minutes
Scored 315 pts/allowed 270 pts, for a +/- of +45 points (+71 adjusting for opponents)
Shooting 48.7% eFG%, allowing 45.0%
19.2% TO Rate
+0.26 points per possession better than opponent (adjusted)
4 Starters: Played 254 minutes
Scored 556 pts/allowed 411 pts, for a +/- of +145 points (+182 adjusted)
Shooting 57.2% eFG%, allowing 43.2%
19.5% TO Rate
+0.44 points per possession better than opponent (adjusted)
3 Starters: Played 233 minutes
Scored 426 pts/allowed 354 pts, for a +/- of +72 points (+91 adjusted)
Shooting 48.5% eFG%, allowing 40.5%
15.7% TO Rate
+0.25 points per possession better than opponent (adjusted)
Less than 3 Starters: Played 82 minutes
Scored 141 pts/allowed 154 pts, for a +/- of -13 points (-4 adjusted)
Shooting 51.6% eFG%, allowing 48.8%
24.9% TO Rate
-0.03 points per possession worse than opponent (adjusted)
If you had to rank these groups blindly, you’d say the 2nd group was easily the best. However, if you just asked a UK fan if they thought UK is better with a bench player in place of 1 starter, there’s almost no chance they’d agree. The perception is that playing the most talented players will get the best results, but that’s far from the truth. The bottom line is that Kentucky outscores opponents by almost twice as much when they mix 4 starters in with a bench player…regardless of which bench player is in or which starter is out! In fact, of UK’s 10 most played lineups, the 5 starters have the 7th best +/- per possession…the top 6 are all 4 starters + a bench player (Poythress, Hawkins, or Lee).
In SEC play, it’s been even worse. The starters have actually been outscored when they’re on the court, as you can see here:
5 starters: -0.03 pts per possession (adjusted)
4 starters: +0.40 pts per possession (adjusted)
3 starters: +0.22 pts per possession (adjusted)
Less than 3: +0.32 pts per possession (adjusted)
So, in SEC play, the 5 starters have been even worse than they have over the full season. Improved play from lineups with less than 3 starters (especially from Jarrod Polson, Alex Poythress, and Dakari Johnson) have helped pick up the slack. It’s mostly been a problem starting halves; the starters are -15 in 21 minutes when starting halves, but +8 in the other 16 minutes they’ve played.
My suggestion would be to rotate in Alex Poythress for one of the usual starters to start each half. I’d suggest keeping Andrew Harrison in as PG and rotating out the other 4. These 4 lineup combinations are the 4 best in +/- per possession for Kentucky. In conference play, the gap between winning and losing can be really small; witness the Arkansas game…UK was +14 pts in 30 possessions with 4 starters in that game, and -16 in 48 possession with any other lineup. Playing your most effective lineups can make a huge difference, but it takes a lot of guts to look deep enough to figure out which lineups those actually are.