Article Length: ~1,100 words or about 6-8 minutes of reading.
My last article on the question of who is the best player remained mostly abstract. While it’s an abstract question, I want to try and provide some concrete answers by applying it to NBA players from the 2018-19 season. This will be extremely difficult to do because everything concrete begins with philosophical abstractions, so you’ll just have to roll with my logic (or disagree with me on Twitter).
If all NBA players were placed in a fantasy draft and I had to start from nothing, who would I select first? The answer to this question is the answer to who the “best, first-best” player in the league is. To me, this comes down to two players: LeBron or Steph. I really really want to pick LeBron for this because I still believe that he can reach transcendent levels better than any other player in the league on both sides of the ball, but the knock against him (like most players) is that he works best in a specific team-build: defensively proficient shooters. If for some reason I’m not able to lock down adequate players like that in subsequent rounds, I’m afraid that he’ll lose value.
On the other hand, Curry, being the “best, second-best player” in the league (we’ll get to this), means that he will retain his value regardless of whom is selected in subsequent rounds; furthermore, we’ve seen Curry literally be the “best, first-best” player on the greatest regular season team in history.
Going back to my first round of the draft, here’s a rough outline the “best, first-best” players in the league (the guys that you would want to draft in the first round of a fantasy draft):
If you don’t have one of these nine players on your team, you cannot win a championship right now. Does that mean that the other twenty-three teams in the league should give up? No. You never know when another player is going to skyrocket into this stratosphere of players (Giannis, Jokic, nor Embiid were at this level last year).
The next tier, the “best, second-best” players become trickier to explain. Players can overlap between different levels of being the “best” player, but not all of them will. Curry will also top this list, but LeBron won’t be on it because if you are making your second round selection and LeBron is somehow still on the board, he would turn into your “best, first-best” player because he would most strongly dictate how your team would play. These are the players who either need a player a tier above them to cover some gap in their own skillset or tier 1 players who can seamlessly provide extraordinary value without the ball (e.g. defense, shooting, passing).
Here are the “best, second-best” players in the league:
- Some order of Vucevic, Kyrie, Simmons, Beal, Paul, Westbrook, and Nurkic
Below this tier are the “best, third-best” players who will generally fill one of two templates: players who excel in a valuable and malleable skills (defense/shooting) while struggling in other areas (think Danny Green and his shooting vs. shot creation and passing) and players who excel in multiple areas that are malleable while not necessarily leading a team through significant scoring or playmaking (think Pascal Siakam and Draymond Green). Here are those players:
- Stephen Curry…yes, I would place him first at literally any round of the draft. I believe he would fit literally in any lineup; therefore, I won’t include him in later lists.
- Pascal Siakam
- Al Horford
- Kyle Lowry
- Marc Gasol
- Khris Middleton
- Malcolm Brogdon
- Josh Richardson
- Jrue Holiday
- Joe Ingles
- Buddy Hield
- Eric Bledsoe
- Myles Turner
- Steven Adams
Are we starting to see a pattern? The most successful teams this year had multiple “best, nth-best” players at different and appropriate tiers. Just from these first three tiers, let’s compare some of the top teams (tier 1 – position; tier 2 – position; tier 3 – position):
- Toronto (1 – 5; 0; 3 – 3, 6, 7)
- Golden State (2 – 1, 3; 0; 2 – 2, 5)
- Milwaukee (1 – 4; 0; 3 – 8, 9, 15)
- Philadelphia (1 – 9; 2 – 9, 13; 0)
- Portland (0; 1 – 8; 3 – 14)
Three of the four conference finalists had a tier one player, and the only one without (Portland) was swept by the Warriors. While neither Milwaukee nor Toronto had a clear-cut “2nd best player,” they both had strong “3rd best players” at multiple positions. After watching how both teams came together, it’s clear that Toronto’s “3rd best players” were simply a better fit in their lineups.
Jacob Goldstein’s PIPM model, which calculates a player’s defensive and offensive impact on a team in a way that takes into account their boxscore production and plus/minus data (for a refresher, check out my article on impact metrics), provides some interesting insights into this “best player” conversation. For 2018-19, the top players are pretty expected: Giannis, Curry, Embiid, etc., but as you go down the list, a few odd names pop up. After LeBron at 14, the following players appear: Vucevic, Brook Lopez, Draymond, Holiday, Drummond, Bledsoe, Danny Green, followed by Kawhi. A simple explanation is that the statistic is flawed because someone like Danny Green (Kawhi’s teammate) couldn’t have had a more positive impact than Kawhi.
The truth is that these players – the Siakams, Greens, Horfords – all can provide more value than a tier 1 player if they are placed in the perfect “nth” role. Having a perfectly fitting “third-best” player can yield more value than a poorly fitting “first-best” player. This explains why someone like Kemba Walker ranks 50th in this statistic: even though he is tremendously talented, he should not be placed in a “first-best” role.
Hopefully, this provides some grounding for my claim that the best player is not always the best player. If I were as in-depth as is necessary, each of these lists would be fluid depending on who was selected in previous rounds. For instance, a player like Lillard would be a better “2nd best player” next to LeBron while Gobert might be “better” next to Harden. I simply don’t have the time to map out an entire fantasy draft matrix contingency.
Once again though, this is why Steph Curry should be rated much higher than he already is. Also, shout out to Robert Covington who is probably the “best, fourth-best player” in the league.
Note: if you want access to the PIPM data, they exist in a few spreadsheets, so searching them out on Twitter is probably your best bet. Jacob Goldstein is the owner.
Addendum: I tweeted a more in-depth explanation of the different tiers, so instead of incorporating those thoughts into this article, I’ve just included the tweets here.