Will Dwight Howard be better now that he's playing with the Houston Rockets instead of the Los Angeles Lakers? How about Andre Iguodala with the Golden State Warriors?
Those questions are on everyone's mind for a reason. NBA superstars don't change teams too often, and as a result, we wonder how they'll perform during their first year with a new team. Will the changing surroundings have a positive impact on them? Will the uncertainty of a new situation pose problems?
We just don't know yet, and it's not like us human beings to accept waiting for a whole year to answer our questions. That's why I'm turning to the numbers to figure out whether superstars are better or worse during that first season in a novel situation.
Over the last decade (2004-2013), the NBA has seen a number of superstars switch teams while still playing at a true All-Star level, but we're going to focus on seven of them here: Shaquille O'Neal (2004), Kevin Garnett (2007), LeBron James (2010), Chris Bosh (2010), Chris Paul (2011), Dwight Howard (2012) and James Harden (2012).
Let's move in chronological order. Keep in mind that any stats you see presented graphically will be per-36-minute numbers. That way, any subtle changes in playing time are negated.
Shaquille O'Neal: Los Angeles Lakers to Miami Heat in 2004
Driven (to some extent at least) by the feud with Kobe Bryant, Shaquille O'Neal was the first superstar in the last decade to take his talents to South Beach. He was traded from the Los Angeles Lakers to the Miami Heat for Caron Butler, Lamar Odom, Brian Grant and a few draft picks.
As you can see below, The Big Aristotle maintained his dominance despite the transition:
While leading the league in field-goal percentage for the second season in a row, Shaq upped his scoring rate by a little more than three points per 36 minutes and kept everything else in line with his final Lakers season.
He was transitioning from playing alongside Kobe Bryant to joining forces with Dwyane Wade, and there wasn't much of an adjustment necessary. This was also aided by the fact that O'Neal was still a dominant physical presence, one that offenses and defenses were centered around.
Instead of adjusting his game, O'Neal was adjusted to. That change from active to passive voice presents a crucial difference, and it's a major part of the reason that The Big Diesel was able to remain such a consistently dominant player.
Kevin Garnett: Minnesota Timberwolves to Boston Celtics in 2007
Kevin Garnett was an absolutely unstoppable force while he played for the Minnesota Timberwolves, even if it never resulted in a championship. There was no doubt that he was "the man." That changed once he joined the Boston Celtics.
The above statistical differences are explainable once more.
Garnett was suddenly thrust into a completely different situation, one in which he was just a member of the Big Three. He had to share the rock with Paul Pierce and Ray Allen, and he did so while moving further away from the basket in order to space the court more.
There are a lot of numbers and images in this article, but the single most impressive one comes in the graph up above. Despite joining forces with a much stronger team and accepting a reduced role—KG's usage rate declined from 27.4 to 25.5—he experienced an uptick in scoring rate.
That simply isn't supposed to happen, and it's a testament to the greatness of Garnett when he was in his prime. The willingness to sacrifice right away in his new role played a large part in both KG's ability to maintain his scoring prowess and the Celtics' eventual title.
LeBron James: Cleveland Cavaliers to Miami Heat in 2010
Now we come to the first player for whom facilitating is a crucial part of the game. Kevin Garnett is a great passer from the low post, but he's still not a primary distributor.
We'll find out later whether or not this is the beginning of a trend, but LeBron's assists per 36 minutes did go down as he transitioned from the Cleveland Cavaliers to the Miami Heat.
His scoring rate went down as well, and both drops were the result of a new system. Whereas LeBron controlled everything and had constant command over the rock with the Cavs, he was one of three stars in Miami.
During the 2010-11 campaign, there was a constant tug-of-war between LeBron and Dwyane Wade for ultimate control over the team. The Heat weren't able to win a championship until Wade stepped down and acknowledged that LeBron was the No. 1 player in Miami.
But that initial struggle resulted in LeBron handling the ball less and a decline in both points and assists recorded per 36 minutes. Even the best player in the world isn't immune to the phenomenon that occurs whenever one basketball needs to be divvied up into more slices.
Chris Bosh: Toronto Raptors to Miami Heat in 2010
LeBron wasn't the only one who experienced a decline in individual numbers in Miami after that fateful 2010 summer. Chris Bosh falls into the category as well after he joined forces with LeBron and Wade to form the NBA's newest Big Three.
While many of the other stars featured in this article experienced declines in their new situations, none were as precipitous as Bosh's.
The 5.4 points that he lost per 36 minutes were more than double the difference suffered by anyone else, and he also managed to drop 2.5 rebounds and 0.4 blocks—lowlights among the seven players. Bosh's situation was unique, though.
Instead of going from Batman to Robin when he moved from the Toronto Raptors to the Heat, he changed from Batman to Alfred Pennyworth. Alfred is still a crucial member of Batman's crime-fighting team, but it would be tough to justify calling him the true second fiddle.
The only other star in the last decade who has experienced a similar role change has been Ray Allen, who is not featured in the article. But interestingly enough, Allen's points per 36 minutes declined by 6.1 when he put on a C's uniform for the first time.
As has been the case throughout, points still seem to be a function of role.
Chris Paul: New Orleans Hornets to Los Angeles Lakers Clippers in 2011
Unlike many of the other players featured above Chris Paul, this point guard didn't experience much of a role change when he left the bayou for the allures of Los Angeles.
While he was the unquestioned leader of the New Orleans Hornets, that remained true with the Los Angeles Clippers. He just traded playing alongside David West for throwing lobs to Blake Griffin. Let's take a look at what happens when teammates and jerseys change but roles don't.
CP3 actually experienced a rather significant increase in points per 36 minutes, but it was accompanied by a simultaneous decrease in both rebounds and assists. The former isn't particularly concerning, seeing as Paul is a pure point guard without much size, but the latter most certainly gives reason to pause.
Although Rajon Rondo has consistently beaten him out for the assists crown over the past few seasons, CP3 is still the league's best point guard, and it's a title that stems from the many parts of his game. His incredible distributing skills are most certainly a crucial aspect of his greatness, and the loss of nearly an assist over the span of 36 minutes is worth noting.
Dwight Howard: Orlando Magic to Los Angeles Lakers in 2012
Ah, the curious case of Dwight Howard.
Many caveats apply to this move from the Orlando Magic to the Lakers, most notably Howard's health throughout his one season wearing the Purple and Gold. However, Howard still played, and he still performed at an extraordinarily high level.
As you can tell from the graph above, Howard declined across the board, with the exception of blocks per 36 minutes. Considering his overall defensive impact was lessened because he was playing with his hands rather than his feet, that doesn't speak particularly well about the validity of blocks when analyzing defensive impact.
While some of the decline can be attributed to his back issues, it can also be credited to yet another change in role.
The Magic built their offensive and defensive systems around Howard, but he was the No. 2 option in Los Angeles. It's the exact opposite situation to the one experienced by Shaq when he was leaving Kobe Bryant's side, so it shouldn't be too surprising that their numbers trended in different directions.
James Harden: Oklahoma City Thunder to Houston Rockets in 2012
The bearded one may not have been a true superstar while he was serving as the Oklahoma City Thunder's sixth man, but he was most assuredly one of the NBA's biggest names as soon as he transitioned into the featured role for the Houston Rockets.
It's for that reason that Harden's included as one of the seven featured players. We have plenty of examples of what happens when the role is lessened, but not vice versa.
You can see above what happened to Harden's stats. It was undoubtedly impressive. With the exception of his rebounding figures, everything went up. His points in particular skyrocketed, and the young shooting guard was the only player who experienced an uptick in assists per 36 minutes.
So, what does all of this mean? Let's take a look at the bigger picture now that the minutiae are all laid out.
Above you can see all the differences ([stats on new team]-[stats on old team]) per 36 minutes in each category. For example, the bars for Shaq and KG's assist differentials don't appear because they experienced no change in the category.
There are a few key takeaways from this:
- I'm going to continue ignoring blocks and steals because they really don't matter too much. We'll be looking at overall defensive impact in a bit.
- Rebounds and assists both experienced overall declines.
- Harden was the exception to the rule in assists, and that stems from a complete change in role. He was the scoring spark plug off the bench in OKC, but he was the man in Houston.
- Somewhat surprisingly, rebounds seem dependent on team play as well as individual prowess on the glass.
- Scoring is strange.
Fortunately, some of the weirdness in No. 5 can be explained when you look at the correlation between the increase in usage rate and the change in scoring.
As you can see, there's a pretty strong correlation between usage rate and scoring rate. As the former increases, so too does the latter.
In fact, the R^2 value for a trendline fit to the data is 0.907, which indicates an extremely strong fit. Anything greater than 0.9 qualifies as quite strong, so it's rather telling that that correlation is already up there despite only having seven data points to work with, one of which (KG) is an outlier.
I want to be careful not to infer causation from mere correlation, but it seems as though usage rate is what's affecting scoring rate. Elite NBA players are going to score around a point per possession, which means that you shouldn't be surprised that they score more when they're given more possessions to work with.
Scoring is a surprisingly individual box-score metric. While teammates certainly have an influence, good scorers are going to remain just that: good scorers. So far, we've learned that rebounds and assists will likely decline (unless he pulls a Harden and has an insanely high increase in ball-handling responsibility) and that points seem simply to be a function of role.
There's no way to flat-out predict scoring changes without first analyzing how a player's new role compares to his old one. But now it's on to the two aspects of basketball that I've neglected thus far: offensive efficiency and defensive play.
Overall Offensive Efficiency
There's no one perfect way to quantify a player's overall offensive efficiency, but the closest metric is offensive rating from Basketball-Reference. It represents the number of points a player produces per 100 possessions.
You can find the full complicated breakdown of the calculations here, but if you're not interested in the nitty-gritty of it all, just know that it factors in everything that a box score can capture about offense. Below are the offensive ratings for each of the seven players before and after their moves.
There isn't much of a consistent trend here, as players get increasingly or decreasingly efficient regardless of changes in roles. This is supported by the next graph, which shows the correlation between changes in usage rate and differences in offensive rating.
Good luck finding a trend there. If you do attempt to fit a best-fit line to the data, you'll get one that has a correlation coefficient of just 0.113. Essentially, the data is almost completely patternless. In a lot of ways, this makes sense. NBA players typically experience a tradeoff between efficiency and volume, but the system they're playing in has a large impact on their success.
Take Harden, for example. His numbers take the pattern I would have expected; the offensive rating severely declined as his importance to the offense went up. However, that's not the sole reason for the change. The Rockets ran a system that was better tailored to his talents—a fast-paced offense that relied on drives to the basket off screens and plenty of three-point shooting.
That has to be factored in somehow. Just as was the case with scoring in general, offensive efficiency depends on the situation.
In order to quantify defensive performance in just one number, I'm going to turn to the impact a player has on his team's overall defense—the best metric for our purposes here. Individual PER against, defensive rating, Synergy numbers and more all have their own merits, but none come as close to the entire picture.
The metric we'll be using is calculated by subtracting the number of points per 100 possessions allowed by the team when a player is on the court from the number allowed when he sits. A positive score is a good thing, as it means that fewer were allowed when the player in question was active.
Let's just call this "defensive impact," and you can see it along the Y-axis in the graph below.
Of the seven players, O'Neal, Garnett, Howard and CP3 all got markedly worse defensively. Their teams were effective to varying levels when they played, but their impacts seemed minimal compared to the years prior.
There were three exceptions to this rule of thumb: LeBron, Bosh and Harden.
Harden is pretty easy to explain. He simply couldn't have been much worse than he was for the Thunder, so a regression to the mean was inevitable. The bearded shooting guard was still a horrific defender, but he was slightly better than his final season in Oklahoma City.
As for the other two, there's an obvious explanation as well. They went from systems in which they were the primary pieces to a team on which they could buy into the group concept. Miami's defense was terrifying given the versatility and effectiveness from top to bottom, and that made Bosh and LeBron have an easier go of it.
In general, there will be exceptions. That's an inevitability of life, and basketball is no, well, exception.
But for the most part, team defense is something that comes with time. Unless the burden is lifted by exigent circumstances (regression to the mean, major upgrade in surrounding pieces, etc), an individual will suffer defensively when he changes teams.
Seven players obviously makes for a ridiculously small sample size, but it's not like superstars change teams every day. We have to work with a limited amount of data, regardless of how much I've trimmed it down.
Looking too far into the past for examples doesn't help out either, as basketball is a rather amorphous entity; the game changes by the year, and so too would the conclusions drawn here. Given the inherent limitations of this brief overview, we can glean a couple main pieces of information.
With a few logical exceptions, the defensive play, rebounding and distributing of a player will inevitably decline once he puts on a new jersey. Those aspects might eventually jump back up to their pre-move levels, but they're all dependent on teammates. This is especially true for distributing and defense.
Scoring and offensive efficiency, though, are more tied to the change in roles.
The former is almost directly correlated with usage rate—the change comes at a nearly one-to-one rate—and efficiency is a byproduct of the new system, something that can't be quantified but rather must be interpreted subjectively.
Now, the original question posed in this article was: Are NBA superstars better or worse during their first year with a new team? Unfortunately, there is no simple answer. It depends on what part of the game you're talking about, as well as the factors surrounding the move.
Remember that patience is a virtue. You'll have to if you want to keep your sanity while trying to figure out what exactly will happen to D12, Iggy and the other big-name players who changed teams. We can make educated guesses, but as this article shows, there are no hard-and-fast rules.
Note: All statistics, unless otherwise indicated, come from Basketball-Reference.