A Look Back at the NBA Finals and LeBron James’ Supporting Cast
Before moving on to the draft, I thought I’d take one more quick look back at the Finals. Many articles are focusing on how great LeBron was and there has been plenty of debate on whether he should’ve gotten the Finals MVP. But what makes LeBron’s performance even more impressive is the relative lack of performance from his teammates. This isn’t a knock on them, as none of them were expecting to play second fiddle to the greatest player in the world. Injuries may have robbed the Cavs of a title. But the fact is the performance of the Cavs role players makes LeBron’s production even more impressive.
Let’s look at their performance in a few different ways. First, we’ll look at the non-LeBron Cavs shot selection, what they would be expected to score based on their shot selection and if they have underachieved on their expected point total.
We can do this by using advanced XPPS+, which I introduced in an earlier post. [1. Again, aXPPS+ adjusts for the shot distance, defender distance, number of dribbles, the time in the game, shot clock, home/road and the height difference of the shooter and defender. For shots near the basket, touch time is also a part of the model. For more details, see this post. For the gory math, the model.]
Yikes, that’s a lot of red. In total, the Cavs have lost about 7 points a game due to underachieving shot-making, given their shot selection. Of course, the word underachieving implies that the Cavs have a league average roster around LeBron[2. Minus Kyrie Irving and Kevin Love, who of course, didn’t play/barely played in the Finals.], which is highly unlikely. So a more likely explanation is that 7 points a game is the result of a lack of talent and players who are likely to underperform their shot selection. How about the King himself?
LeBron has underperformed his shot selection?!?! Well, that’s certainly different then your typical narrative. But there’s a reason for this. Because aXPPS+ factors in the defender distance on each shot, LeBron is going to naturally be taking tougher shots given the amount of attention he’s drawn. However, the main reason for his numbers here is the sheer volume of shots he’s taken. We can look at the overall relationship between aXPPS+ and Shot Number within the game[3. Found in the SportVU shot logs.]
As we see here, there’s a negative correlation (-0.93) between shot number and aXPPS+. As a player takes more shots, his shot selection goes down. Part of this is that “stars” are the ones typically taking a high volume of shots and they’re going to typically draw more defensive attention but even so, there’s a clear relationship between Usage and Shot Selection. Given LeBron’s 32.7 FGAs per game, we can run a regression and predict his aXPPS+ given only the number of shots he has taken per game. His 32.7 FGAs per game would be expected to yield a 0.837 aXPPS+, which would put him in the green in terms of overachieving his shot selection. So this says a lot about LeBron’s abilities and basketball IQ to be able to get better shots given those high number of FGAs. Many players/stars can’t do this.
What if we look at a few other measures? Like
(DRE),
(AWS) and
? How have the rest of the Cavs performed?
We can see that Kyrie was obviously a big loss given his performance in game one. We can also see that Tristan Thompson and Timofey Mozgov were decent players in the Finals, mainly due to their offensive rebounding[4. From what we saw above, they weren’t exactly decent in regards to performing to the levels of their shot selection.].
According to Game Score, both Mozgov and Thompson rate as about average. The same is generally true for both DRE and AWS. But how about the wings and guards minus LeBron? Since DRE approximates RAPM and a RAPM of -2 is considered to be replacement level, we can see that J.R. Smith, Shumpert and Delly all played below replacement level during the course of the series. What exactly does this mean? Well theoretically, you would be able to replace these three players with Joe Schmo from the D-League and receive similar production[5. But to be honest, I’m not entirely sure what it means and why the term is necessary. Why not compare to the average player?]. But I remain skeptical of this proposition as these three guys all bring certain skills that might not be easily replicated by a replacement level player[6. And stay tuned for a larger discussion on replacement level from the group. It’s been quite the excellent discussion.]. However, whether you want to get technical with a term like “replacement level” or not, there’s a pretty clear point that these three players did not play very well and were a big part of the problem in the series. Delly had a few good games at the beginning of the series but after Game 3, did anyone even notice he was on the court?
All in all, LeBron was mostly a one man team with Thompson and Mozgov occasionally providing help to clean up the boards.