Ever since the acquisition of Rudy Gay, people have wondered if and how Gay and DeMar DeRozan could play together.

Floor spacing concerns and an offense built around two low-efficiency, high-volume players running isolations led many to believe the duo wouldn’t co-exist capably.

Of course, the Raptors’ new starting unit surprised many by being extremely effective last season, outscoring opponents by 12.9 points per 100 possessions over a 343-minute sample. The team’s offense improved slightly with this five-some, and the defense was dramatically better.

No worries, then, right?

Well, that would make for too short an article. And 343 minutes is hardly enough of a sample. There were other concerns as well, namely that the success was coming primarily on defense while the offense wasn’t necessarily going gangbusters.

In 173 minutes to start the year, the starting five is actually being outscored by 5.3 points per 100 possessions, with the team taking a step back on both ends of the floor. The starting unit is actually performing worse than the team in general, and with a sample of 516 minutes now, we’re starting to see that the starting five may be good-not-great.

The core of the issue, offensively, is the DeRozan-Gay combination.

Consider the following table, which shows the team’s O-Rating and D-Rating with certain lineup iterations (don’t cringe, anti-stats people…this is just points adjusted for pace of play, or “points per 100 possessions”):

Unit Time Min Ortg DRtg Net
Starting Five 2012-13 343 105.4 92.5 12.9
Starting Five 2013-14 173 101.1 106.4 -5.3
Starting Five TOTAL 516 104 97.2 6.8
Gay-DD 2012-13 923 103.4 101.3 2.1
Gay-DD 2013-14 316 98.6 101.9 -3.3
Gay-DD TOTAL 1239 102.2 101.5 0.7
RAPTORS Since Trade 2286 101.6 102.9 -1.3
LEAGUE 13-14 Average x 101.4 101.4 0

(Stats via NBA.com/Stats)

What that’s showing us is that the Gay-DeRozan combination is roughly average offensively and defensively, which isn’t really good enough when these are supposed to be your two best offensive players. This year, in particular, the duo has struggled mightily at the offensive end.

Opening this up further, we see what happens to the team and individual when both are on the floor together, compared to just one or neither (for this season):

Unit TS% PPP 3PA/FGA FTA/FGA % FG Ast Midrange% Ortg DRtg MIN
Gay+DD 49.1 0.981 22 27.8 45.6 30.7 98.6 101.9 316
Just DD 56.7 1.17 24.9 38.5 50 30.1 111.9 94.1 103
Just Gay 49.7 0.993 25.4 33.8 38.9 26.9 98.8 99.4 79
Gay OR DD 53.7 1.09 25.1 36.5 45.2 28.7 106.2 96.4 182
Neither 47.3 0.952 33.3 33.3 63 17.4 93.1 102.4 45

(Stats via NBAWowy.com)

What we see here is that the team’s offense has been better with just one of the two players on the floor, specifically with just DeRozan out there. With neither, it obviously suffers.

It’s worth noting now that these are small samples, though I think it backs up what we all a) expected and b) have witnessed.

The floor spacing improves, the team gets to the line more and cuts down on inefficient mid-range jumpers.

At the individual level, the results are a bit more surprising:

DeMar TS% USG PPP FGA/36 % FG Ast Midrange%
w Gay 46 25 0.96 15.7 49.1 50.7
w/o Gay 55.2 36.2 1.15 22.7 41.4 42.9

(Stats via NBAWowy.com)

DeRozan has killed it without Gay and has shown a knack for being the primary scorer on the floor. It does appear he goes iso a little more (lower rate of assisted field goals), but we saw above that team as a whole sees better ball movement with this look.

Gay TS% USG PPP FGA/36 % FG Ast Midrange%
w DeMar 47 30.9 0.97 19.1 39.7 36.9
w/o DeMar 43.9 37.9 0.87 23.7 23.8 42.3

(Stats via NBAWowy.com)

Gay, however, looks fairly lost without DeRozan. Again, it’s a small sample, and you can perhaps chalk it up to Gay’s poor start rather than some larger trend (and, in fact, the roles were reversed last season), or even the competition each may face when rolling solo. The data isn’t rich enough – and/or my management of the data capable enough – to dive in.

(On that note, it’d be very cool to see the SportVu data teams work with, to see what kind of movement, spacing and opponent defense the team sees in these different instances.)

But all of this leads us to a big ‘so what?’

Sure, the team looks better with one or the other, but not both, on the floor. We’re still dealing with a small sample, Gay has been uncharacteristically bad and, to his credit, Dwane Casey has actually staggered the playing time fairly well, such that there have been only four minutes a game, on average, where neither player was on the floor.

Step one would be to get that number closer to zero, minus garbage time. (There are still issues with his rotations, to be sure, and I’d get even more aggressive in leveraging sub patterns to make one of these guys the de facto scorer on the second unit.)

Step two would be to do what you can to get Gay going (perhaps by having him not dribble in isolation into two defenders and hoist up a jumper with a safety valve available, and then 3-point shooters a quick pass away).
gay1
It’s not like the team can make a change to the starting line-up to better stagger the presence of these two wings. You can’t very well tell DeRozan, who has been working his ass off, that he’s going to the bench, and you certainly can’t do it to Gay if you wan’t to maintain any semblance of trade value while he works things out.

I think the point here, then, is that the experiment that most thought made no sense from an Xs-and-Os standpoint, isn’t working at the offensive end.

Whether that can be remedied by the coaching staff, some creativity on offense or changes to the rotations is unclear. After all, it worked to some degree for a 30-game stretch last year.

In all likelihood, though, it’s just further proof that the experiment wasn’t a smart one to undertake (I seem to remember someone trying to save his job at the time) and yet another piece of ammunition for the “blow it all up” crowd.