FiveThirtyEight projections analysis: Orlando Magic

Mar 21, 2016; Boston, MA, USA; Orlando Magic point guard Elfrid Payton (4) reacts during the fourth quarter against the Boston Celtics at TD Garden. The Boston Celtics won 107-96. Mandatory Credit: Greg M. Cooper-USA TODAY Sports
Mar 21, 2016; Boston, MA, USA; Orlando Magic point guard Elfrid Payton (4) reacts during the fourth quarter against the Boston Celtics at TD Garden. The Boston Celtics won 107-96. Mandatory Credit: Greg M. Cooper-USA TODAY Sports /

Analyzing FiveThirtyEight’s CARMELO projections for clues about what we can expect from the 2017 Orlando Magic. Will the Magic exceed expectations?

Earlier this month, the folks at FiveThirtyEight released their CARMELO NBA player projections for the upcoming 2017 season. This projection system tries to use wins above replacement as a means to predict the career arcs for players in the NBA that FiveThirtyEight’s founder, Nate Silver, introduced yesterday. Yes, that is the guy who is good at predicting elections.

CARMELO (Career-Arc Regression Model Estimator with Local Optimization) is a system that uses a player’s fundamental skills and attributes and historically comparable players to project a player’s career arc in terms of season-by-season WAR (wins above replacement). Given how well this system performed in its first year (2016 NBA season), and the improvements being implemented in its second year, CARMELO might be as useful a tool as any to project how good the Orlando Magic will be next year.

Below is a table summarizing projected and actual WAR from last season and WAR projections for next season. After the table, there are observations and notes about what we can learn from these projections.

But before all of that, to help provide WAR benchmarks that can aid interpretation of the table, here are some WAR numbers of other teams’ players from last season as a reference:

Stephen Curry21.7
Russell Westbrook18.3
James Harden15.1
Paul Millsap10.7
Isaiah Thomas7.3
DeMar DeRozan5.3
Taj Gibson3.9
Kyle Korver2.4
Austin Rivers-0.2

And here is the table (players in alphabetical order by last name):

2015-16 Regular Season

 D.J. Augustin-0.5Offensive Specialist0.2+0.7
Bismack Biyombo1.3Rotation Player1.9+0.6
Evan Fournier1.7Rotation Player3.8+2.1
Aaron Gordon0.5Up-And-Comer3.9+3.4
Jeff Green-0.4Scrub0.4+0.8
Mario HezonjaN/AN/A-0.6N/A
Serge Ibaka5.7Borderline All Star2.6-3.1
Jodie Meeks0.7Offensive Specialist-0.1-0.8
Elfrid Payton5.7Future All Star0.8-4.9
Nikola Vucevic2.7Average Starter4.3+1.6
C.J. Watson1.3Rotation Player-0.5-1.8

2016-17 Projection

Player2016 SeasonProjectedDifferentialDescription
D.J. Augustin0.20.1-0.1Offensive Specialist
Bismack Biyombo1.91.8-0.1Rotation Player
Evan Fournier3.82.5-1.3Average Starter
Aaron Gordon3.94.2+0.3Future All Star
Jeff Green0.40.6+0.2Scrub
Mario Hezonja-0.60.5+0.9Project
Serge Ibaka2.62.3-0.3Average Starter
Jodie Meeks-0.10.1+0.2Offensive Specialist
Elfrid Payton0.83.4+2.6Future All Star
Nikola Vucevic4.33.7-0.6Average Starter
C.J. Watson-0.50.1+0.6Scrub

Here are some observations about the above projections:

  • Aaron Gordon: Orlando’s best player next year could be Aaron Gordon, who seems to be a bit ahead of schedule in his development. A healthy offseason (finally!) will also do wonders for him. However, given the roster makeover this offseason, Gordon is projected to start at small forward, which could require some adjustments from him. Playing on the wing could afford the Magic a size advantage there and enable Gordon to use some of his ball-handling ability (which is decent for a player his size), but Gordon’s in-progress jump shot will be more exposed on the wing. A borderline all-star season is not out of the question, especially if the Magic overachieve in the first half of the season, but there are also reasons to believe that there might be an adjustment period for Gordon as well.
  • Elfrid Payton: FiveThirtyEight has “inexplicable crushes” (their words) on some players. One of whom is Elfrid Payton. Despite a down season last year, Payton is still projected to play at an all-star level for years to come, although the projections are not as lofty as they were last offseason. There are perhaps two reasons Payton might eventually reward the Magic’s faith in him: A) his historical comparable players, which include some pretty notable players like Norm Nixon, Rajon Rondo, Mike Bibby, Deron Williams, Baron Davis and Mike Conley, and B) his high steal percentage (at least one FiveThirtyEight analysis posits steals are more predictive of player impact than any other box score statistic). Whether Payton can reach the heights FiveThirtyEight predicts will go a long way toward revealing how far the Magic can go, and he will probably still have quite a bit of leash given that backups D.J. Augustin and C.J. Watson are not projected to be any more than replacement-level players.
  • Nikola Vucevic: Despite his much-maligned defense, Nikola Vucevic was not only the Magic’s best player last year, he surprisingly performed better on defense (+1.1 points per 100 possessions relative to an average player) than on offense (+0.7), according to FiveThirtyEight Box Plus-Minus. But offseason acquisitions have clouded Vucevic’s role this coming season. The trade for Serge Ibaka seemed to be an effort to cover Vucevic’s rim protection deficiencies, implying Vucevic will be sticking around, but the signing of Bismack Biyombo raises the question of whether head coach Frank Vogel might be envisioning Biyombo as a defensive anchor for the Magic (much like the role Roy Hibbert played for the Indiana Pacers a few years ago), relegating Vucevic to the bench. Vucevic does have a decent WAR projection for next year (3.7), but that projection is based on FiveThirtyEight’s assumption he will be playing even more than he did last year, when Biyombo was not around. How much Vucevic actually contributes next year will be largely dictated by how he is used (e.g., starter, closer, bench-unit focal point, post-up vs. pick-and-pop), and no one knows for sure how he will be used.
  • Bismack Biyombo: The Magic were one of quite a few teams that fell in love with Biyombo based on his performance in the Eastern Conference Semifinals last year when he took over for an injured Jonas Valanciunas and wreaked havoc on the boards against the Miami Heat — without Hassan Whiteside. It is safe to say Biyombo’s playoff WAR was probably higher than his actual 2016 regular season WAR (1.9). For his work in the playoffs, he earned a contract that will make him the highest-paid player on the Magic next year (tied with Evan Fournier at $17 million). Biyombo’s WAR projection for next year might be high or low depending on whether you believe his playoff breakout was a harbinger of things to come and/or how Vogel decides to manage the relationship between Vucevic and Biyombo, but FiveThirtyEight believes he will play closer to the level he was at during the regular season last year.
  • Serge Ibaka: FiveThirtyEight projects Ibaka to continue a downward trend in performance next year (2014: 5.8 WAR, 2015: 3.3, ’2016: 2.6, projected 2017: 2.3), but there are reasons to think Ibaka might be ready for a bounce-back season. Reports surfaced shortly after the Oladipo-Ibaka trade that Ibaka had been unhappy about his role in Oklahoma City, particularly on offense (here is a story from March). Given that he is just 27 years old, his decreases in WAR likely stem not from an erosion of skills, but perhaps because he was not particularly motivated knowing he was not a high priority for getting the ball. Ibaka should be one of the Magic’s focal points this coming season (because the Magic have no one on the level of Kevin Durant or Russell Westbrook to feed), and Vogel would be smart to give him offensive touches even if it is just to keep him engaged/motivated on the defensive end (similar to what the Houston Rockets should have done with Dwight Howard). A fresh start might be just what Ibaka needed to recover his old form, but the Magic need to do their part too.
  • Evan Fournier: After Aaron Gordon, Evan Fournier was the player who most exceeded expectations in 2016, according to FiveThirtyEight. The Magic appear to believe Fournier’s 2016 season was legitimate and an indication of good things to come, demonstrated by their doubling down on him by trading Victor Oladipo, the player most similar to him on the roster. But FiveThirtyEight’s system projects Fournier’s true talent level to be a bit lower than he displayed last season, and the system expects some form of regression to that level (even if his overall career outlook is a bit sunnier after last season).

Next: The options left for the Orlando Magic this offseason

  • Mario Hezonja: Many Magic fans are high on Hezonja, but FiveThirtyEight is much less so. But given Hezonja is so young and early in his career, there is not a lot of information to base projections from (FiveThirtyEight did not even offer a projection last year for him), and future projections are going to be heavily influenced by what happens this coming season. The “project” description is apt, with free throw frequency and turnovers being the two areas holding him back the most.