There has been plenty written on the Kawhi Leonard trade. Most of the ink has gone towards the reaction DeMar DeRozan had to the trade, and the mixed emotions of many Raptors fans in losing a loyal All-Star for a player the team may have for only one year (if that, to hear some tell it).
One aspect of the trade that seems to be glossed over, accepted as fact and yet underplayed for its value, is that the Raptors got a lot better. Not a little better, not moderately better.
The Raptors got way, way better.
Superstar for an All-Star
Some may question how Toronto got so much better. Sure, if the team had simply added Kawhi Leonard they’d be much better, but they lost one of the top scorers in the NBA and a perennial All-Star and all-NBA selection in the process.
This makes sense on the surface, but to take this stance is to not understand the difference between a Kawhi Leonard-level superstar and a DeMar DeRozan-level All-Star. There is no comparing the two.
Let’s compare them anyway. We’ll use DeMar’s latest season, and Kawhi’s latest healthy season. All indications (Kawhi’s reported intention to showcase his health at the USA Basketball camp were he not traded before then, him passing his physical) are that Kawhi is back in good health, so we’ll assume that until we have a real reason not to.
There are several things one can look at when evaluating a star player — but some pretty logical starting points are volume production, efficiency, and defense.
DeRozan was a solid producer for the Raptors these past five years, and last year was no exception. Some common box-score aggregation statistics that attempt to capture the value of a player’s production are Player Efficiency Rating (PER) and Win Shares (WS, often expressed in WS/48, a per-48-minute version).
DeMar has been between first and third on the team in PER the past three seasons, peaking at a PER of 24, and posting a PER of 21 this past season. A PER of 15 is considered average, while 30 is an almost unreachable value, the very best players in the league usually topping out just below that. The lowest PER of any rotation player with at least 1000 minutes played last year was about 5.0. A PER above 20, like DeMar’s, would indicate roughly a top 30 player in terms of production.
DeMar has produced very well by WS as well, placing second on the team each of the past three years in total WS, and posted a WS/48 of .170 last season (a typical value for all three of the most recent seasons). The league average WS/48 is .100, while the very best players in the league tend to approach .250. DeMar’s 9.6 total win shares last season ranked 16th in the league.
There is also the oldest of production measures: points per game. DeMar scored 23 PPG last season, ranking 14th in the league. He has led the Raptors in raw scoring every year since 2013-14, and peaked at 27 PPG in 2016-17.
So what about Kawhi? This guy is known as more efficient and a better defender, but can he hope to replicate the raw production that was DeMar’s calling card?
Let’s start with points per game. His 2016-17 season, he scored 25.5 points per game, just behind DeMar’s peak in that same season. Already, that’s a solid sign he can produce at a similar rate. But consider that Kawhi played two fewer minutes per game than DeMar. Adjust both players to points scored per 36 minutes of playing time, and DeMar scored 27.8 PP36 and Kawhi scored 27.5 PP36. Practically identical scoring rates.
But scoring isn’t everything. We’ve seen DeMar expand his game and pass more, and he’s a solid rebounder for his position, so we should look at those catch-all production numbers as well.
In 2016-17, Kawhi put up production numbers to the tune of a 27.6 PER, 13.6 total WS, and .264 WS/48. That would tie him for second highest PER in the league, give him the fourth most WS produced (while playing fewer minutes than any other player in the top 10), and he’d place second in WS/48. That was no outlier season either; his previous season he posted values of 26.0 PER, 13.7 WS and .277 WS/48.
So, Kawhi Leonard seems to be easily capable of reproducing the raw production that DeMar DeRozan provided for the Raptors.
One of the big differences in the two players is how they go about posting their 25+ PPG. That shows up a bit in their catch all box-score stats above, but let’s look more closely at it.
First off, the mark of a good first option in the NBA is to be able to finish a lot of possessions at an above average efficiency. DeRozan meets those criteria — he’s carried varying high usage rates (how many of his team’s possessions he finishes) over his seasons here, peaking at an incredible 34% in 2016-17, but for the most part floating just a hair below 30%. This is a very high usage rate. His 34% ranked 3rd in the league that season, and his typical 30% usually sits around 10th. Even this past season in Nick Nurse’s offense, DeRozan carried a near-30% usage rate, so we should expect Kawhi will need to replicate that high usage when he plays here.
As for efficiency, DeRozan has done a good job over time developing from a far-below-average efficiency scorer to a league average efficiency scorer, posting a true shooting percentage (a version of FG% adjusted for the value of threes and free throws) of almost exactly 55% the last three seasons (league average last season was 55%). That’s fairly impressive at such a high usage. Russell Westbrook posted a similar 55% TS% with his (insanely) high usage. But DeMar is not what you’d call efficient, even qualifying for his usage.
So, how does Kawhi stack up here?
Well, in 2016-17, he posted a TS% of 61% on a usage rate of 31%. The year before, a TS% of 62% on a usage rate of 26%. On the face of it, Leonard is incredibly efficient.
For reference, those 60+ TS%’s compare very similarly to two very efficient players already in Toronto, who carry a lower usage rate to generate those efficiencies: Kyle Lowry and Jonas Valanciunas. Lowry posted a 60% TS% on 19% usage last season, and Valanciunas posted a 63% TS% on 23% usage.
Valanciunas carrying a TS% that high on a pretty high usage rate is already very impressive — only four other players with 1500 minutes played managed a TS% and usage rate as high as Valanciunas’: LeBron James, Kevin Durant, Steph Curry, and Karl-Anthony Towns. And Kawhi Leonard is able to post similar efficiency on a usage rate nearly 10 points higher, in line with the first three names on that list.
Now, being efficient on high usage is all well and good, but how was Leonard scoring his points? If he’s scoring mainly off cuts, post ups, rolls and catch-and-shoot opportunities (similar to how Valanciunas scores his points), Kawhi would struggle to repeat that efficiency if forced into DeRozan’s role in the offense.
So, let’s use play type data to examine what plays each player scores most on, and where on the floor they score from (and on what efficiency, shown in points per possession). I’ve only looked at the play types where DeRozan was asked for at least one possession per game.
Pick and Roll Ball Handler: 9.3 possessions per game, 0.92 PPP
Isolation: 3.0 possessions, 1.03 PPP
Transition: 2.2 possessions, 1.26 PPP
Spot-up: 2.2 possessions, 1.00 PPP
Off-screen: 1.8 possessions, 1.06 PPP
Post-up: 1.6 possessions, 0.83 PPP
Pick and Roll Ball Handler: 5.6 possessions per game, 1.01 PPP
Isolation: 2.9 possessions, 0.94 PPP
Transition: 2.9 possessions, 1.33 PPP
Spot-up: 4.3 possessions, 1.24 PPP
Off-screen: 1.7 possessions, 0.91 PPP
Post-up: 1.9 possessions, 0.99 PPP
Very similar numbers across the board, with slightly better scoring in the pick and roll and post, and slightly worse scoring in isolation and off screens. Similarly dominating transition game. The one area that has driven his efficiency way up is his increased attempts and far greater efficiency on spot up shots. Leonard is far more of an off-ball threat than DeMar is, which works out nicely with Nurse’s stated intention to run more offense through Kyle Lowry this season.
In any case, Leonard has demonstrated competency in all the attacking plays DeMar tends to be used most in, so should have no issue filling into that role on offense.
And at this point, we’ve spent 1,500 words on offense. I guess we might want to get to the other half of the game at some point.
This end of the court is far harder to measure. We could go over raw steals and blocks (both of which favour Leonard by about a factor of 2), but really the best measures of defense right now in advanced stats are impact stats. They are still flawed (any defensive number is) but in extreme result cases, can usually be believed, and (spoiler alert) we’re going to be looking at two extreme cases here.
First, DeRozan. Much has been made of his defense in these latest playoffs, and it was indeed disastrous. However, it was a continuation of a trend, one that extends into the regular season.
One quick way to look at defense is simply how efficiently the opposing team scores when the player is on the floor. This is full of context and noise, which we’ll address in a second, but long term trends can still point at something. Here are DeRozan’s on-court and off-court DRTGs (defensive rating, points given up per 100 possessions) for the past few regular seasons and playoffs. A lower number is better.
2017-18: 105.4 On, 98.9 Off
2017-18 playoffs: 113.5 On, 109.6 Off
2016-17: 107.0 On, 100.7 Off
2016-17 playoffs: 109.2 On, 103.4 Off
2015-16: 104.1 On, 99.2 Off
2015-16 playoffs: 107.7 On, 98.4 Off
On average, over that sample, the Raptors’ defense was 6.1 points per 100 possessions better without DeRozan on the floor.
Now, there is all sorts of context to that — DeRozan played a lot of minutes against starters (though he also got a lot of bench minutes), the Raptors’ bench has been very strong defensively, etc.
So to try to remove that context, there are impact stats — essentially modifying simply plus-minus stats to account for what teammates a player plays with and what opposition they play against. These stats have gone through several iterations. The most easily available ones are currently basketball-reference’s Box Plus Minus (BPM, from box score data correlated to the plus-minus data) and ESPN’s Real Plus Minus (RPM, a combination of the adjusted plus-minus data and some box score data). They are hardly perfect, but they don’t need to be in our case.
Here are DeMar DeRozan’s three most recent seasons by defensive BPM and defensive RPM.
DeMar’s DRPM the past three seasons: -1.8, -2.0, -2.5
DeMar’s DBPM the past three seasons: -0.9, -1.5, -1.2
That comes to an average estimated defensive impact of -1.7 points per 100 possessions. That value would rank 464th of 521 players in defensive RPM last season.
Here are Kawhi Leonard’s three most recent healthy seasons.
Kawhi’s DRPM his past three healthy seasons: +1.3, +3.9, +4.6
Kawhi’s DBPM his past three healthy seasons: +1.5, +2.8, +3.5
That comes to an average estimated defensive impact of +2.9 points per 100 possessions. That would rank 20th of 521 players measured last season.
464th. To 20th.
All that time discussing the offense really is just to get a feel for if Leonard can even replicate what DeRozan did on that end with the Raptors, because his defensive impact seems likely to be dramatic, to say the least. Leonard seems likely to be more efficient and more able to play off the ball, which is great news. But this fact alone can be completely overshadowed by the potential defensive upgrade his presence brings to the team.
Which was the point of this article. None of this stuff should be getting overlooked. The Raptors should be significantly better on offense with Kawhi in DeRozan’s place. The Raptors should be dramatically better on defense with Kawhi in DeRozan’s place. And considering they had the third best offense and fifth best defense in the 2017-18 regular season, and that offense finally held up in the playoffs (while the defense did not), that should be a scary thought for other NBA teams.