Evan Fournier never found his consistency in 2019

Orlando Magic fans know the narrative surrounding Evan Fournier.

He is up and down. He is inconsistent. If the Magic are to go anywhere, they need an upgrade at his position.

Evan Fournier has not been many of those things throughout his career with the Orlando Magic. He was solid and a reliable shooter for much of his tenure with the team. Perhaps the Magic asked him to create too much on the perimeter, but he was the team’s best 3-point shooter.

In his first four seasons with the Magic, he averaged 15.6 points per game, shooting 37.9 percent from beyond the arc and a 52.7 percent effective field goal percentage overall. Those are not All-Star-level numbers, but hardly inconsistent.

Unfortunately, in 2019 this perceived narrative about Fournier became true.

He has his worst season in a Magic uniform, averaging 15.1 points per game, the worst since he first joined the team in the 2015 season. Fournier shot a career-worst 34.0 percent from beyond the arc. His 50.9 percent effective field goal percentage was one of the worst marks of his career.

Fournier improved in several other areas. He averaged a career-high 3.6 assists per game. And his 2.5 win shares and -1.1 defensive box plus-minus were by far the best marks of his career. That is not a lot to write home about, but he showed dramatic improvements in those areas.

But ultimately, his main contributions to the team came on offense. He is the team’s most potent 3-point shooter and that part of his game was non-existent — especially in the playoffs when he made only 8 of 34 (23.5 percent) 3-pointers in the five-game series.

His struggles defied explanation, even from him. Fournier tried to remain confident but could not get his shot to go down.

Many NBA players receive the “inconsistent” label. And Fournier certainly received that label fairly.

But how inconsistent was he, really?

Looking closer at his 2019 season, the answers were damning for the 2-guard but there remains reasons for optimism.

Hollinger’s Game Score Metric: A Quick Intro

Over at Basketball Reference, John Hollinger devised a “Game Score” metric, which roughly tracks a player’s productivity over the course of a game and doles out a “score” using the following formula:

=(Points)+0.4*(Field Goals Made)+0.7*(Offensive Rebounds)+0.3*(Defensive rebounds)+(Steals)+0.7*( Assists)+0.7*(Blocked Shots)- 0.7*(Field Goal Attempts)-0.4*(Free Throws Missed) – 0.4*(Personal Fouls)-(Turnovers).

A Game Score of 10 is average, while something in the neighborhood of 40 is outstanding.

It is not without its flaws.

For example, Game Score does not account for pace as PER does. Also, in combing through all this data, there were some cringe-worthy standouts. Draymond Green was NOT an average player during the 2019 regular season.

Nevertheless, the metric still provides a rough estimate for a player’s game-by-game productivity on the floor. For what it is worth, the all-time high Game Score is Michael Jordan‘s 69 point performance against the Cleveland Cavaliers in 1990, followed by Kobe Bryant‘s 81-point performance against the Toronto Raptors in January 2006.

So, despite its imperfections, it is not way off. And in looking at Fournier’s 2019 season, it will help us defined and quantify his inconsistency all season.

For purposes of this post, we measured consistency by taking the standard deviation of the Game Scores of several players during the 2019 season.

Daily Knicks

Standard deviation measures how spread out numbers are. Thus, the standard deviation of a metric like Game Scores is one way to see which players have more or less variance in their performances game to game.

But that is not enough. The standard deviation of Giannis Antetokounmpo‘s Game Scores for 2019 is higher than Emmanuel Mudiay‘s. Is Mudiay more consistent than Antetokounmpo?

That answer should be self-evident.

Better players typically average higher Game Scores. The standard deviation of Game Scores does not differentiate where a Game Score slides of, say, 15 to 5 from a slide of 35 to 25.

As a result, the players’ average Game Scores are included.

From there, I included the coefficient to demonstrate the Game Score standard deviation relative to a player’s average (mean) Game Score. The lower the coefficient, the more consistent a player is relative to his average Game Score.

The Data We Collected

We looked into the Game Scores from Basketball-Reference and standard deviation of those scores of the major minutes-earners at the shooting guard position throughout the league.

It is true that much of the NBA trends toward “positionless basketball”, but presenting all data for players that roughly occupy the same position as Fournier — i.e., wing players –would be intractable.

So, we stuck with each team’s primary 2-guards.

Some teams had funky shooting guard rotations (the Cleveland Cavaliers and Detroit Pistons stand out). For those squads, we simply included all players that fit the bill.

With all that table setting, how did Fournier come out? Is the inconsistency narrative overblown?

The answer is an emphatic no.

What It Means

Out of the 44 players evaluated, Evan Fournier is 25th in this measure of consistency. This suggests he hovers around the middle — we knew that, right?

So why do we need the table and all this data?

A closer look tells a more nuanced story.

Fournier’s average Game Score is 9.706 while the standard deviation is 6.724, for a coefficient of 0.693.

Looking at the players further down the list, all but one (Terrence Ross himself) has a lower average Game Score.

With an average Game Score that high, one might expect the standard deviation to be lower. That is, for there to be less variance and more predictability of what Fournier could provide on any given game night.

Instead, many of the players in the ballpark of Evan Fournier’s average Game Score, like Joe Harris (9.851), Dwyane Wade (10.367) and Tim Hardaway, Jr. (10.871), had smaller deviations from the mean or made up for a larger deviation with improved overall play.

Similarly, with a standard deviation of 6.724, it might be easy to hope for a higher average Game Score that would leave added room for such variance in order to maintain a sliver of predictability or reliability, like the cases of Buddy Hield (standard deviation of 6.754 but with an average Game Score of 13.990) or Zach LaVine (6.998 standard deviation against an average Game Score 15.775).

Those figures highlight the importance of the players’ coefficients and also show what an aberration Evan Fournier is (sadly, so is his teammate Terrence Ross).

Ultimately, the data suggests the Magic really did not know what they would get out of Fournier night in and night out.

Even though his average Game Score suggests he would put up an average performance on any given night, his performances varied too wildly from game to game and swallowed whole any ability to anticipate what he could bring.

Caveats to the Results

The Game Score data shows Evan Fournier was glaringly inconsistent in 2019, but there are a few grains of salt available for taking.

First, it is important to look at Fournier as a player historically.

Since coming to Orlando, inconsistency has not defined him as a player using the same methodology as above:

The data suggests Fournier is capable of greater consistency.

In terms of average Game Scores, 2019 was his most inconsistent since his first in Orlando. In each of the three seasons from 2015 through 2018, Fournier showed more steadiness at the 2-guard position for Orlando.

So, despite how maddening the Game Score data suggests 2019 was for Magic fans watching Fournier, it is not the case that all we can wish for is more wildly scattered game performances. Fournier can do better.

We should also consider that 2019 was Fournier’s first under coach Steve Clifford, who is well known for his emphasis on defense.

He asked Fournier to step up his defensive game, and Fournier responded. He posted his best defensive rating since 2016, and career bests in defensive win shares and defensive box plus-minus.

Those types of results would not be reflected in Hollinger’s formula, which relies on weighted assignments to standard counting stats.

Key Takeaways

It does not take a wildly observant fan to know Evan Fournier was inconsistent in 2019.

But the Game Score data suggests Fournier was in rarefied “Inconsistency Air.” He could truly give you nothing one night, then be an above-average shooting guard the next.

But Fournier’s defensive success, coupled with the limits to the data — some believe Hollinger’s Game Scores rely too much on scoring— pose a decent argument for ignoring the Game Score data altogether.

That goes too far. Even if Game Scores focuses too much on scoring, Fournier’s offensive performances apparently varied so wildly that he sticks out like a sore thumb when looking at the table above.

It is great Fournier improved defensively, but that cannot send his offensive reliability so far askew.

This offseason, the Magic doubled down on the roster that brought the organization its first playoff berth since 2012. Nikola Vucevic and Terrence Ross got paid. The team drafted another seemingly long-term project in Auburn Tigers forward Chuma Okeke. Management avoided adding point guard help but did bring in Al-Farouq Aminu to help shore up the bench.

At shooting guard, then, any improvement will have to come from within, barring a trade. It follows that another season like 2019 from Fournier will pose a setback for a team hoping to continue its upward momentum.

Hollinger’s Game Score data suggests the team simply could not rely on him, at least offensively, because on any given night, the range of his potential production was so wide.

But with Fournier proving capable of consistency in prior seasons of his career and clearly improving on defense, the Magic faithful have reasons for optimism despite Fournier’s down 2019 season offensively.

Fournier knows he can do better and if he combines his defensive improvements with offensive production of years past, it will bode well for a Magic team looking to be even noisier in 2020.