Your favorite rankings of NBA players* are turning 11 this season, and it's a good time to revisit what we've learned from the first decade of ESPN's NBArank.
(*Might not actually be everyone's favorite.)
Like all great offseason content, NBArank was conceived out of necessity during the summer the NBA locked out its players, a stalemate that wasn't settled until the day after Thanksgiving, with a shortened 66-game season starting on Christmas Day.
With neither games nor transactions during the extended NBA summer, NBArank wasn't the only player rankings to debut in 2011, but it was the most comprehensive -- both in terms of the amount of players rated and the large panel of voters that benefited from the wisdom-of-the-crowd approach -- and influential.
As a result, the rankings have long outlived lockout boredom.
Over the past 10 years, NBArank has inevitably evolved, both because of some high-profile misses and despite some controversial placements that turned out to be pretty obvious in hindsight. The scope of the rankings has changed, as has the method used to generate them.
One thing has remained constant: People love debating exactly where their favorite players belong in the NBA's universe of stars.
Note: NBArank 2021 begins Tuesday with the unveiling of players Nos. 100-51.
We're gonna rank ... 500 players!?
With plenty of time to fill in 2011, NBArank started out going through 500 players, enough to cover just about everyone likely to be on a roster that season. Eventually, that scale proved unsustainable, in both practical terms -- how many people have informed opinions on the 450th-best player in the league? -- and real-life ones, given how fit becomes increasingly important relative to talent for role players.
The highest-profile early NBArank results were predictably from the 400s as players surprised us with their play. In the original 2011 incarnation, Jeremy Lin -- who'd played sparingly in garbage time for the Golden State Warriors as a rookie -- was ranked 467th. He was one spot behind current Brooklyn Nets GM Sean Marks, who retired before the 2011-12 season started.
Less than four months (and two trips through waivers) later, Lin would become a phenomenon by averaging 24.6 PPG and 9.2 APG over a 10-game stretch remembered as "Linsanity." By 2012, Lin would jump all the way up to 78th after signing with the Houston Rockets.
Like Lin, Kent Bazemore was a little-known member of the Warriors when he ranked 499th out of 500 players in 2012 ahead of his rookie season, trailing three players who would never appear in the NBA. Bazemore famously had "499" stitched into his shoes as motivation. Although he was still a few years away from developing into the quality 3-and-D contributor he has become, Bazemore brought enough perimeter defense to play 61 games for the Warriors as a rookie and nine more in the playoffs.
By 2015, we'd cut the list down to 400 players. It went to 200 the following year and has focused solely on the league's top 100 since 2017, a more manageable task.
Too early or too late?
Much of the fun of NBArank is that we're not evaluating players on how good they were the previous season, something that's relatively easy to quantify. Instead, we want to know which direction players are trending.
Sometimes, that means betting on breakouts too early. After a strong playoff run as a rookie, Jayson Tatum of the Boston Celtics ranked 25th in 2018. He wasn't yet ready to challenge for an All-Star appearance at age 20 and actually slid 10 spots to 35th the next year, when he did develop into an All-Star.
At other times, NBArank has been ahead of the curve. Nikola Jokic jumped all the way to 42nd after his rookie season, which saw him average just 10.0 PPG, on the strength of his impressive advanced stats. We moved Jokic to 16th in 2017, before he'd made an All-Star Game. (He wouldn't get that accolade until 2018-19.) It took awhile for the rest of the league's observers to catch up, but they eventually got on board with the 2020-21 MVP.
It wasn't always this way. Originally, NBArank was framed as more of a current ranking rather than a forward-looking one, but that presented challenges for players at the beginning and end of their careers. That changed in 2013 with an explicit focus toward predicting the future.
Injuries matter
If we're looking back at the times NBArank was most correct relative to online consensus, I suspect many of them would involve players coming off injuries. Because NBArank is asking the question of how much value a player will provide in the upcoming season, health matters -- a lot.
Most famously, fans of the late Kobe Bryant were outraged when he ranked 25th in 2013, months after suffering an Achilles rupture that ended his 2012-13 season. Certainly, Kobe's previous play had merited a ranking in the top 10, but he didn't return from the injury until December. What nobody could have predicted was that Bryant would suffer a season-ending tibial plateau fracture six games later.
More recently, the NBArank panel had to figure out what to do with Kristaps Porzingis coming off an ACL tear that ended his 2017-18 campaign. Porzingis ranked 58th in 2018, seemingly low for a 23-year-old rising star who had just made his All-Star debut. As it turned out, Porzingis didn't play at all in 2018-19 as he rehabbed the injury.
Parsing is harder at the top
The original NBArank method was that every voter -- up to over 100 at times -- rated every player on a scale of 0 to 10 and players were ranked by the average score. The challenge with that system was that among players given perfect 10 scores by a majority of voters, a handful of 9s ended up making the difference. That didn't stop LeBron James from being atop NBArank each of its first eight seasons, before his injury-marred 2018-19 campaign dropped him to third, but did produce some other slightly wonky results.
In 2015, that changed to a system where voters are shown two players and asked to choose which will have the better season based on quality and quantity of play. That makes it easier to directly compare the top players to each other and determine where they relatively rank.
Regular season or playoffs?
Perhaps the toughest aspect of NBArank now is determining how to weigh regular-season and playoff performance, a choice left up to individual voters.
Particularly as star players sit out an increasing number of games in the regular season for rest or injury, it's tough to know how to handle a leading MVP candidate who's less effective in the postseason and a veteran star who coasts through the 82 games before turning it on in the playoffs.
This issue came to the fore in 2018, when Giannis Antetokounmpo finished tied for fourth with two-time reigning Finals MVP Kevin Durant behind LeBron, Stephen Curry and James Harden. At that point, Giannis had yet to see the second round of the playoffs. I argued at the time that putting him in the top five was reasonable given the possibility he emerged as MVP -- which is precisely what happened, as the Milwaukee Bucks posted the league's best record. Still, it was Durant who looked like the league's best player in the 2019 postseason before he was sidelined by a calf strain.
After claiming the No. 1 spot in 2019 -- Durant was not eligible due to his Achilles rupture -- Antetokounmpo dropped to No. 3 after a playoff flameout in the bubble. On the strength of his postseason performance in the Lakers' title run, LeBron grabbed the top spot for the ninth time in 10 years. Durant, Giannis -- who has now matched his two regular-season MVPs with one from the Finals -- and LeBron again figure to be in the mix for this year's No. 1 player.
We've learned a lot in the first decade of NBArank. Yet there are still plenty of lessons sure to come -- some the hard way -- as we continue the challenging task of ranking NBA players.