Some Concerns with Giannis
For those voting for Giannis: what are your criteria?
I have a hard time voting him above Barkley or Nash, much less Durant or Erving, just given the lack of cumulative career value. He only really has 5 MVP level seasons, he got injured in 3/5 of those postseasons, and his team got upset by an SRS underdog in 3/5 years (including one where he was healthy).
Don’t get me wrong: His peak is up there (I’d definitely take his peak over Barkley or Nash) and I weigh peaks heavily. His goodness when healthy is up there, and I tend to care less about health concerns. But is his peak really so much higher as to take him just yet?
Concern 1: A lack of career value (from a lack of prime longevity)
Let’s check PIPM as a ballpark estimate for total career value. It’s just one stat, it’s not perfect, but it does a fairly good job at capturing value (it’s like luck-adjusted RAPM), it includes playoffs, and it has one of the better box estimates on the market going back to the NBA Merger.
PIPM Career Value (1977–2020):
Spoiler:
And it makes sense. PIPM grades Giannis as having 5 MVP years, 2 all NBA years through 2023. Compare that to Erving (likely 2 more MVP years, 1 more all NBA year including ABA), Durant (likely 5 more all-nba years through 2023), or Barkley (5 more all-nba years).
What about Moonbeam’s RWOWY? WOWY metrics have super wide uncertainty ranges, but they’re based on actual impact, and we have all the years for everyone.
Spoiler:
-Durant (4 more samples above 90th percentile, 6 more above 75th)
-Nash (5 more above 90th, 8 more above 75th)
-Barkley (4 more above 90th, 13 more above 75th),
Moses (2 less above 90th sample, but 3 more above 75th percentile
So… Giannis seems pretty significantly below most of these guys in career value. I would suspect Giannis will shoot up in career value the next time we do this project — 3 more seasons at this level do a lot for a career — but I’m also not sure we can credit him for those seasons just yet.
Concern 2: Health and a lack of playoff resilience
When a player doesn’t have the prime length or any longevity, they likely need to make up the value elsewhere. Since Giannis doesn't have the career value (yet!), is he a sufficient playoff riser to surpass these other players?
Not exactly, although some of the decline comes from frequent health problems. A quick overview of health and team level performance:
2018: 1st Round loss to Celtics without Kyrie.
2019: Upset, with major decline in performance. Beat by team with 2.55 worse SRS.
2020: Upset, with a major decline in performance. Beat by team with 6.82 worse SRS. Giannis was injured, but Bucks were down 0-3 even before Giannis’ injury. Although the Bubble makes this a very unusual circumstance.
2021: Championship! But Giannis injured, misses 2 games, and the Bucks were a KD shoe size away from being upset again by a team with 1.33 worse SRS. Still, great performance post injury in the finals.
2022: 2nd Round loss. Giannis shoots 10.2% worse (!) in relative True Shooting than his Regular season average, although his defense absolutely picks up against a strong opponent.
2023: Upset, one of the biggest upsets of the modern era as a 1st seed losing in the 1st Round. Beat by a team with 3.74 worse SRS. Giannis was injured and missed 2.5 games and returned unhealthy, although the Bucks played better without him.
So… not exactly the kind of stalwart playoff performance to make up for the longevity disadvantage. Now Giannis' team results may look better with more granular analysis, like playoff SRS or net rating. But Giannis does have poor playoff health (injured in 50% of his prime playoffs!), and his teams have had multiple playoff upsets to weaker SRS teams (significant SRS upsets by teams 2.55, 3.74, and 6.82 SRS worse in 50% of his prime playoffs!). Although to his credit, the defense is absolutely resilient when he's healthy.
What if we check more granular team data? Let's look at playoff Relative Net Rating, to capture both the fantastic defense and the concerning offense. Here are the best 3-year runs for each of the players:
-Giannis: +8.55 (2019-2021, 80th all time)
-Durant , +14.9 (2016–2018, would be 1st all time); if discounting GSW, +8.03 (2013–2014/2016)
-Nash: +8.26 (2008–2010)
-Moses: +7.07 (1983–1985)
This is obviously a very crude metric, highly dependent on teammates. But if you're looking for evidence that Giannis' team significantly outperform these rivals in the playoffs to make up for Giannis' lack of longevity and prime years, you're going to have to keep looking. Durant looks significantly better in the superteam years, and the other players aren't far behind Giannis at all. In fact, Giannis' team's all-time postseason ranking here (80th all time) is actually lower than their best regular season ranking, again suggesting there's a playoff decline.
This decline shows up in the individual data too. In Augmented Plus Minus, Giannis declines by -8% in the playoffs, which would be the 2nd biggest decline on record to be voted in. For context, Chris Paul declines -4% and Durant declines -1%, and both have overall more postseason value. Note that this data is only through 2021: he improved slightly in 2022 and likely declines in 2023, so the true average may look slightly better, but likely not enough to take him over Durant or make up for the longevity disadvantage.
Concern 3: His peak isn’t far enough ahead to make up for Concern 1–2.
We’ve voted for players with shorter longevity and poor playoff health and/or resilience. Curry was 11th and has far fewer impactful non-prime years… yet he still has ~50+% more prime years than Giannis, a higher peak, and less playoff impact decline (perhaps because better health). We just voted Chris Paul and Karl Malone, who have larger playoff health concerns or playoff decline than most… but they have an even greater prime length / longevity advantage over Giannis. As do Durant and Nash and Barkley and others, as I’ve shown above. So for us to pick Giannis, he has to have such a better peak or prime that it makes up for the longevity, poor playoff health, and larger playoff decline than most.
EPM is generally the best stat on the market for measuring current goodness/value. Here are the EPM ranks for Giannis:
2019: 6th
2020: 1st (2nd in total season value)
2021: ? (I don’t have a subscription, but Jokic was 1st)
2022: 3rd
2023: 8th
… which is great stuff! Absolutely one of the top players in the world right now. But not exactly domination by such an extreme extent that it makes sense to take Giannis over guys with 50% longer primes and careers than him. Giannis' best years are better than Durant per 100 possessions, but Durant’s minute/game advantage in 2014 actually gives him the best overall season. Durant is still ranked 5th in 2022 (only 2 spots behind) and 9th in 2023 (1 spot behind) while being 6 years older. Harden has actually has more total season value in his best years than Giannis in his best too.
Now Giannis likely does have a better peak than Moses or Barkley or Nash, but again they have significant prime length and longevity advantage. Does Giannis really have enough of a peak advantage to make up for the other disadvantages?
…
In sum, I’m not saying you can’t pick Giannis. There’s definitely certain criteria where he rises up… extremely peak heavy, less of a focus on playoff health or playoff decline, caring more about whether a player reached “that level” of goodness at some point even if it hasn’t been that long (yet!), caring much less about career value or prime length or longevity. But, me personally, given how many player have a career value advantage over Giannis (Concern 1), and given he doesn’t show enough playoff resilience/health (Concern 2) or have a significant peak advantage over the best of these players (Concern 3), I’m not quite ready to vote for him just yet.
But anyhow.. looking through data, more box-oriented and even box/on-off hybrids are the ones that grade Durant rather highly and certainly more-so than metrics focusing on pure +/-. Am curious why this could be the case, and think there's at-least some situational context needed here (eg. Westbrook controlling more through playmaking , GSW's 'system' built around Steph). 
). 
Compare that to their 3-year playoff relative Net Rating, which ranks 80th all time! That's a drop in the playoffs. Still good stuff, still able to win a championship, but still a drop. 
