When Orlando Magic coach Stan Van Gundy moved Mickael Pietrus out of his starting lineup in favor of Matt Barnes last week, he did it with the best of intentions. His team had come out flat of late, trailing at halftime in 7 of its last 9 games. Van Gundy thought Vince Carter, Dwight Howard, Rashard Lewis, and Jameer Nelson--the other four starters--would feed off Barnes' energy. Intuitively, it makes sense: Barnes makes more hustle plays than Pietrus does, and although he isn't as skilled a perimeter defender, he's a far superior rebounder and moves much better without the ball. Meanwhile, Pietrus could assume a bigger offensive role coming off the bench, reprising the role he held in the Magic's deep playoff run last year: in 24 games, all off the bench, Pietrus averaged 10.5 points on 62.2% True Shooting, while playing All-World defense on the likes of Paul Pierce, LeBron James, and Kobe Bryant.
What I'm getting at is that Van Gundy had the right idea when he made the move. The problem is that it hasn't worked. In the six games since Barnes re-assumed his starting role--he filled in as an injury replacement early in the year--the Magic are 2-4 and on a 4-game losing streak. There are myriad reasons and no one person is responsible, but this much is clear: the Barnes/Carter/Howard/Lewis/Nelson starting unit isn't cohesive or effective, and Van Gundy should table it. He'll be forced to tonight against Atlanta, with Carter likely out due to a sprained left shoulder, but even when Carter returns, it shouldn't be to the same starting lineup that Van Gundy has implemented in each of the last 6 games.
In those 6 games, Orlando has held a halftime lead twice and trailed 4 times, so the switch hasn't changed the Magic's course in that regard: they're still having to play from behind. And since the switch, that unit has played 84:07, or roughly 7 quarters' worth of basketball. Its raw +/- over that time is -15, meaning opponents have outscored it by 0.179 points per minute. Further, Orlando has come out ahead just 5 times in its 13 stints on the floor.
As for why that unit has struggled, well, I'm not quite sure. My theory is that putting the newcomers Carter and Barnes together with the mainstays Howard, Lewis, and Nelson disrupted the on-court chemistry. I stress "on-court" because nothing I've heard suggests that Carter or Barnes is a negative influence in the locker room. In any case, the mainstays have a long history of success together. Per NBA.com's +/- Stat ("Driven by AutoTrader.com"!), they've tallied a +527 rating in 2493:58 over their first two years. Adding Pietrus to that trio yielded a +89 rating in 320:34 of playing time spread over last year and this one, so the Howard/Lewis/Nelson triad is amenable to playing alongside him. But losing Pietrus to the second unit and integrating two new players seems to have had an adverse affect on Orlando's cornerstones.
The variable here is Carter. Can he co-exist in a five-man unit with Howard, Lewis, Nelson, and Pietrus, who all know each other's games? The signs point to "yes," as that lineup is a solid +24 in 65:08 of playing time this year, outscoring its opponents by 0.369 points per minute. Orlando didn't hold many halftime leads with this group, but it's proven to be 0.548 points per minute more effective than the alternative.
Again, Orlando has several problems to deal with as it contends with its longest losing streak in nearly 3 years, and reverting to the previous starting lineup may not solve anything. However, there's substantial evidence to suggest that this original lineup has a greater chance of succeeding than the one with Barnes in it. That's not meant to be an indictment of Barnes, by the way, because he's played well individually this year. And in his last 6 starts, he's averaged 12.7 points on 67.3% True Shooting to go with 6.3 rebounds. He's just not a great fit with Orlando's "Big Four" of Carter, Howard, Lewis, and Nelson. Pietrus is, which is why Barnes should re-join the second unit once Carter returns to health.