Photo credit: Keith Allison from Hanover, MD, USA / CC BY-SA (https://creativecommons.org/licenses/by-sa/2.0)
Let’s get this out there from the jump: Giannis Antetokounmpo is the NBA MVP.
In a piece covering the restart and the impact on the NBA’s awards, Tom Ziller argued that with the bubble games non-eligible for award recognition, Giannis should easily repeat as the league’s Most Valuable Player.
We’ve known for about five months that Giannis Antetokounmpo would win his second consecutive NBA MVP award — all four months of the NBA shutdown, and at least a month prior to that as No. 34 dominated and his Milwaukee Bucks took control of the Eastern Conference. There was some intrigue after LeBron James and the Lakers beat the Bucks on March 6, and given LeBron’s leadership and performance for the top team in the West. But this was mere intrigue: few suggested by March 11 when the NBA went on hiatus that LeBron could truly catch Giannis in the hearts and minds of enough voters.
If these Orlando games count only for playoff considerations, the time’s finally arrived to run the 2019-20 data and pick a winner.
My scoring model takes in a variety of advanced and per game statistics and compares each player’s output to his peers. I compiled this from every season since 2013-14 from Basketball Reference. Each data point turns into a z-score, or the count of standard deviations away from the mean NBA player’s result for that stat.
Let’s use a practical example to visualize. In 2018-19, James Harden scored 36.2 points per game. He leads this entire dataset and produced the 7th highest season in NBA history. Here’s how he looks compared to the rest of the players’ scoring marks. The 25th through 75th percentiles for points per game fall in the Rockets red box. Non-outliers fit between the two brackets.
The bearded man himself tops the rest of the outliers at the very peak of this graph. That very large distance should be rewarded and tracked in any attempts to holistically score a season. Harden’s points per game translates to a 4.31 z-score, meaning he’s over four full standard deviations above the average NBA player’s scoring output.
All of the tracked stats in this analysis become similar z-scores. I weight slightly by adding emphasis to the advanced metrics that correlate nicely with what we’d consider calculations of value. These include numbers like VORP (value over replacement player) and WS/48 (win shares driven per 48 minutes, an estimate of wins driven by the player in the length of one game).
I changed the model’s scoring system this year to eliminate total win shares. This stat, while incredibly important, penalizes players from this abbreviated season who couldn’t play a full slate of games.
Here’s how the total weighted score looks for every season tracked in the database. The outliers on the right represent the most impressive MVP candidates.
This framework lets us isolate MVP candidates and crown a winner.
Back-to-back mvp for giannis
Last year, this model picked Giannis as a clear-cut MVP, and he’s again the easy selection. Giannis beats out Anthony Davis and James Harden to capture a second-straight, highly-coveted Warm the Bench MVP vote.
Kawhi Leonard and LeBron James round out the top five for this MVP ballot. None came close to matching Antetokounmpo.
Last year, Giannis eked out a victory over James Harden by the slimmest of margins. He produced a weighted score 0.67 above the Beard. This year? He’s ahead of Anthony Davis by over 3.50.
I updated a chart pulled from his last MVP win to showcase Giannis’s improvement. Giannis transformed from unrefined Milwaukee prospect to league-destroying wrecking ball.
While he’s again top of the league, his total weighted MVP score dipped slightly in this season. This decline deserves a deeper dive.
Year over year change for giannis as nba mvp
Giannis’s total weighted score dropped from 30.45 to 29.36 in this season. The interesting takeaway there is less the aggregate decline and more the composition of those different scores.
On a per-game basis, Giannis increased his scoring and rebounding outputs, while declining in assists, steals, and blocks. He turned the ball over less this year, too.
Per Basketball Reference, Giannis led the league in PER, WS/48, BPM, and Defensive Rating.
Picking the tiniest of nits, he’s shown diminishing results for true shooting and offensive rating. His efficiency dip compared to last season can be partly explained by a league-highest usage rate and uptick in three pointers taken. Giannis is still an efficient and prolific scorer, but an altered shot profile and higher workload on the offensive end brought down his true shooting slightly. He made up for any declines by posting a career-best 29.6 points per game.
I’m more fascinated by that NBA-best defensive rating. I dove into that through some research on Cleaning the Glass. This data frames Giannis as the spearhead of an insanely stingy defensive Bucks team. With Giannis on the court, the Bucks held opponents to 97.7 points per 100 possessions and a pitiful 47.3% effective field goal rate, good for the 99th and 98th percentile respectively.
Giannis is functionally incapable of small steps—the man can go from halfcourt to the rim in what seems like two strides—but this defensive brilliance vastly outweighs any slight stumble he took backwards on the offensive end.
Milwaukee performed at a +16.1 point differential with him on the court. This translates to a 73 win season and would tie the Warriors for the most wins in a season.
What about LeBron?
A good rule of thumb for NBA analysis? Always agree with Zach Lowe, and he’s on board with Giannis repeating.
Yet, while arguing his case on ESPN’s Get Up, he encountered the typical retort of narrative that always plagues MVP discourse. Giannis’ near-complete statistical superiority faces the ill-defined, qualitative reasons for a LeBron ballot. LeBron’s success for the Lakers should be commended, but in this scoring, he’s a distant fifth.
Let’s take a look at how his stats stack up against Giannis for this season. On a per-game basis, LeBron won big in assists and edged Giannis slightly in steals, but he’s outpaced by a large margin in rebounds, blocks, points, and turnovers.
The gap widened immensely for the advanced statistics. Here, Giannis beat LeBron in every metric other than offensive rating, where LeBron picked up a very small victory.
Giannis led the entire league in PER, box score plus/minus, and defensive rating. For defensive rating, he’s at least one full z-score above LeBron. He scored more points at a more efficient clip for the best team. Add in the fact that Giannis spearheaded the most fearsome defense on the league’s best team, and there isn’t a data-driven argument for taking LeBron over him.
Giannis unquestionably wins the league’s Most Valuable Player, and perhaps the Defensive Player of the Year, too.
Even though the Orlando bubble will not count toward MVP consideration, these games offer Giannis a national showcase of his gamebreaking skills. With a deep playoff run, or a Bucks championship, he’ll silence any would-be dissenters and take full ownership of the crown.
Giannis Antetokounmpo, your NBA MVP.
Thanks for reading! Feel free to argue your favorite MVP candidate to me at mike@warmthebench. Stay safe and healthy!