Five Magic Observations: The “Plus-Minus” Edition

Welcome to Five Magic Observations. Last week I said the Magic would be in a decent position if they could just go 2-1 last week, and they managed to do just that...still, when you have to gut out wins against the Brow-less Pelicans and the D-League version of the Mavericks, it’s hard to get too excited. Just keeping telling yourself that wins are wins, and maybe you can ignore the reckoning that might be coming for this team when they start facing real opponents. That’s my plan, at least.

Let’s get started with some advanced stats, shall we?

Where do the Magic’s players stand by plus-minus metrics?

This week ESPN released their Real Plus-Minus data now that everyone’s played at least 10 games, so this is probably a good time to take a broad measure of how Orlando’s players are faring by a variety of plus-minus statistics.

For the uninitiated, advanced plus-minus (APM) stats take the standard, vanilla plus-minus you can find in any box score and try to improve the metric with a variety of statistical techniques, the most important consideration being an adjustment for who each player is playing with and against.

For example, if you threw me on a team with LeBron James, we’re probably going to affect each other’s regular plus-minus dramatically: with my devastating triple-threat combination of “short, slow, and unskilled,” he’s going to look a lot worse playing with me, and I’ll probably look a lot better playing with him.

Advanced plus-minus, in theory, accounts for that by more or less saying, “Look, every lineup with Cory Hutson is truly awful, while LeBron actually does just fine when he’s not playing with a 5’10” blogger. LeBron’s “true” plus-minus is higher than what the box score tells you.”

Each APM stat does this in their own sort of way and mixes in their own special blend of other factors, like box-score stats, player height, age, etc. I’m going to review three plus-minus stats today: regular plus-minus, or “net rating;” ESPN’s Real Plus-Minus (RPM); and Basketball Reference’s Box Plus-Minus (BPM). Each of these can be broken down into offensive and defensive components, and in the interest of brevity I won’t list all of those in the table below, but I’ll note any extreme cases as we break down the data.

Don’t take any of these numbers as gospel on their own. These aren’t meant as “rankings” of the players on the team, and after just 13 games, they’re far from perfect. It’s probably fair to say that some of these numbers are outright wrong, even, in terms of measuring the “true” impact of each player. After all, the Magic look like a much different team even if you only compare their first six games and their last seven. Plus-minus data is just one tool of many we should use to understand how the NBA works.

The players are listed in descending order of minutes per game and include anyone that’s played at least 10 MPG. All the data is listed in terms of points per 100 possessions. Net rating describes the actual change in score while that player is on the court, while RPM and BPM describe the player’s estimated contribution above a replacement level player’s production. That’s why those numbers are generally smaller than the net ratings: if five players together are a +10 on the court, RPM and BPM will try to divvy up those 10 points among the whole group.

Across the board, you’ll notice a lot of large negative numbers. That’s because the Magic, while close to a .500 team by record, are deep in the negative by point differential. Those blowouts drag everyone’s numbers down, and for good reason: statistically, blowouts say more about a team than close wins and losses do in the long run. Of course, the Magic have plenty of time to show that the blowouts were temporary setbacks, but for now that’s why nearly everyone looks like a minus on the court.

By the advanced stats, Aaron Gordon looks like the Magic’s best-performing player, a trend that continues from last year when his plus-minus ratings were similarly favorable. It’s encouraging, in some ways, given that’s he’s still just 21-years-old with so much more room to grow into his prime. While he’s looked uncomfortable at times in his role as a small forward, and while his shooting and ball-handling skills leave much to be desired, there’s simply things he can do that nobody else on the team is capable of. His athleticism is the reason he was drafted so high to begin with, and he’s proven he can be a capable on-ball defender.

Hezonja is on the other end of the spectrum, not just the worst plus-minus on the team, but among the worst in the entire league (e.g. 399 out of 407 by RPM). Again, these aren’t “ranking stats,” but you certainly don’t want to be at the bottom of the list. Most of his negative contributions come offensively: if he can’t shoot and he makes poor decisions, what can he do? He, too, has time to grow, but I won’t lie and say I’m unconcerned about these numbers.

The biggest surprise hidden in this data is Vucevic, specifically on defense. He’s a big negative offensively, but he’s got the highest DRPM on the team. He gets a few built-in advantages in RPM for things like “being tall,” but RPM’s not been kind to him in the past on the defensive half of the stat. By on-off metrics, he looks like the better defender than Biyombo, whose DRPM is barely positive. On the other hand, BPM likes Biyombo a lot more than RPM defensively, likely due to his high block count.

I don’t think I really believe that Vucevic is a better defender than Biyombo, but Vucevic’s defensive improvement is definitely real. On the court, he gets to where he needs to be more often than in years past and does a better job of contesting with his limited vertical ability. He’s never going to be an average defensive center, but an upgrade from “terrible” to “subpar” goes a long way. If he could just fix his shooting problems, Vucevic could be the best player on the team again.

Payton’s defensive numbers are way up, too, compared to last season, when he looked like one of the worst defensive point guards in the league. Fournier, on the other hand, continues to rate very poorly defensively. I’m not totally sure where he got the reputation for being a solid two-way player, but I’m fairly convinced, based on both this data, last year’s stats, and the eye-test that he’s pretty bad on that end. That’s not the end of the world, and the Magic certainly didn’t re-sign him for his defense, but it’s hard to avoid thinking about in the context of having traded away Victor Oladipo, who’s almost certainly the superior defender.

There’s plenty more to unpack here, but there’s also plenty that will change. Overall, these numbers tell a story we’re pretty familiar with at this point: the Orlando Magic are bad on offense, better on defense, and have a long way to go.

Checking in on rim protection and shooting

The first two weeks of this column focused on the Magic’s problems protecting the paint and getting their own shots to fall, so I thought it’d be good to check in on how they’re doing with a few more games of data to work with.

On the rim protection side, things are looking up. The Magic have improved significantly on the defensive end of the court, posting the third-best defensive efficiency in the league since November 11th, when they played the Utah Jazz. That 94.6 rating compares against the 109.0 rating from the first 8 games of the season, and overall the Magic are 14th in defense on the season.

Rim protection has been a big part of that improvement. After starting off with a collective 61.3% shooting around the rim on defended shots when we looked at the stat after six games, the Magic are down to 55.8 DFG%. That’s still near the bottom of the league, but it’s trending positively thanks to an above-average 50.8% in the 7 games since. Vucevic, Ibaka, Biyombo, and Green have all improved individually, and Gordon’s jumped into 4th on the team in defended field goal attempts at the rim, allowing just 29.4% shooting on those shots.

Offensively, things are not nearly so rosy. The team was shooting a collective 41.9% from the field when we last checked in, and that number’s dipped down to 41.0%. That pulls them down to 8th on the leaderboard of offensive ineptitude in the 3-point era, per Basketball Reference. They’re managing to get by with riding the hot hand of whichever guys happens to get going, like with D.J. Augustin on Saturday night, but there will be a reckoning against the good teams of the league if they can’t figure out something soon. More on the offense later this week.

Some odd pass-percentage stats

Among the myriad stats tracked by the NBA is the ability to see how often players pass to or from each other on a team. For example, most of Biyombo’s passes both come from and go to D.J. Augustin, which sounds about right.

I’ve been trying to understand Vucevic’s murky spot in the offense this season, especially his precipitous drop in post-up attempts compared to last year (6.9 to 4.5 post touches per game, per NBA.com), and I wondered if guys were just passing it to him less, but that doesn’t appear to be the case; on a per-minute basis, it’s about the same. He’s still passing a lot to and from Payton (like many of the players, unsurprisingly). It’s just that he’s getting fewer passes in post-up situations; his elbow touches have gone up slightly on a per-minute basis.

Over half of the passes Ibaka receives come from Payton, on average about 17.3 passes per game. He’s definitely increased the number of post-up touches he gets compared to his OKC days. While Vucevic is still getting slightly more post-up touches, it often seems like the Magic go out of their way not to throw him post-ups, and that the opposite is true for Ibaka. I’m not sure if that’s a sound strategy, and it’s worth monitoring going forward.

That one time Jeff Green dunked all over Andrew Bogut, sort of

Look, it was a nice dunk, I’m not gonna argue against that. Still, does this really count as a “poster” dunk? Maybe my standards are too high, but it seemed to me that Bogut got far enough out of the way for it not to count as such. I feel like you either have to dunk past a block attempt or be right above the guy to earn “posterization” status.

Maybe I’m just being grouchy because of Green’s low productivity. He’s been almost invisible since joining the starting lineup, and while the Magic’s defense has been better since the lineup change, I’m not sure how much the Gordon-Green swap had to do with it. Then again, Vogel said pregame on Saturday that Green looked really good on defense during the team’s film study, and you’re probably better off taking his word over mine.

What’s coming up next?

11/21 - Magic @ Bucks - While these teams have similar records, the Bucks have certainly been the more exciting of the two, even in losses, like when they went toe-to-toe with the Warriors on Saturday. Without the potential of an annual Tobias Harris revenge game, the Magic will have to find production elsewhere.

11/23 - Magic vs. Suns - This week is really a continuation of last week’s trend of “totally beatable teams that the Magic could blow out on a good night hahahaha just kidding the Magic never get blowout wins.” Devin Booker might roast Evan Fournier on the perimeter, but the Magic really have shaped up defensively, so it might be enough to beat another bad team.

11/25 - Magic vs. Wizards - Orlando’s first rematch of the season, and this time we should expect John Wall to participate. As dysfunctional as the Wizards have been lately that might not matter, though last game was certainly too close for comfort.

11/ 27 - Magic @ Milwaukee - Orlando’s second rematch of the season! A 1-1 split is sort of what I’m expecting by the week’s end, though it might be the case that Milwaukee’s win comes in a blowout and Orlando’s comes in another unnecessarily close contest. Seriously, a double-digit win would be nice, if only to prove that it’s even possible.