PG - Magic JohnsonConsidered: John Stockton, Oscar Robertson
Magic Johnson is one of the most anomalous players the NBA has ever seen. More so than anyone else, he could (and did) play every position. He could rebound like a center and pass like a point guard. Magic's 52% career shooting average is the highest amongst point guards and his rebounding percentage (11.1%) probably ranks up there as well. His ability to play multiple positions defensively while leading fast breaks is a devastating weapon. To put it in modern terms, Jason Kidd has been an elite (for many years, arguably the best) point guard in the NBA; Magic Johnson does everything Kidd does but significantly better (Kidd's three-point shooting ability towards the end of his career aside).
It's difficult to leave Stockton off this list. He was never an elite scorer, but excelled everywhere a point guard should: good three-point shooter, terrific defender, better at generating steals than Magic, assisted on over half of the possessions where he touched the ball (an unbelievable statistic). In some ways, when building a team, it's better to have a prototypical point guard rather than someone unique like Magic, whose strength comes from his size and rebounding, not traditional point guard qualities. Stockton's ability to space the floor and set up others would doubtless serve a team of superstars well, given that there would be no lack of scoring talent on the floor. In the end though, the tremendous mismatches that Johnson causes (there isn't a point guard in the world who can guard him on the low post) as well as his additional rebounding win out.
The Big O is also difficult to leave off, but not necessarily because he's a comparable talent. Unfortunately, without a three-point line and many statistical categories (steals, turnovers), it's tough to tell just how good Oscar Robertson was relative to Magic. He had the same all-around type of game, with tremendous rebound totals for a point guard, and shot a very good 48.5% from the field while taking 7.5 free throws per 36 minutes (see how close those figures are to Jordan's below). But for what info we do have, he falls behind Magic in many vital categories: rebound, assist, and effective field goal percentages, per-minute assist and rebound totals, win shares. Robertson's star was built off of a few extraordinary early seasons and playing 40+ minutes per night; his career was great but not best-in-class.
SG - Michael JordanConsidered: No one else comes close.
Shooting guard is the only easy choice on this entire list. Jordan is head and shoulders above any other shooting guard to have played the game. Jordan was known for being a dynamic scorer, someone who could create his own shot with ease and take defenders off the dribble. But he had a stunningly complete game: he rebounded better than your average guard, was every bit as amazing a defender as a scorer, generated steals, passed fairly well, and turned the ball over surprisingly rarely given the amount of time he spent handling it (9.3 career turnover percentage). Jordan shot a high percentage for a shooting guard at 49.7 and generated 7.7 free throws per 36 minutes with his aggressive drives. His only weakness is his poor three-point shooting: towards the end of his Chicago days he had a couple good years, but he was a lifetime 32.7% shooter from beyond the arc putting him well below most modern shooting guards.
Some would argue that Kobe Bryant is, if not better than Jordan, at least in the same league. There is no statistical validity to this argument. Bryant is worse is every category I mention above—rebounding, steals, turnovers, shooting efficiency, defensive win shares, free throws attempted per minute. It's barely true that Bryant is a better three-point shooter, but he's still below what you want from a SG at 33.6%. Objectively, he does not belong in the conversation. Julius Erving would be an interesting pick as he is actually better at some things—he rebounded and blocked shots at a SF level—but in the end Jordan is just clearly superior in too many categories to consider anyone else.
SF - LeBron JamesConsidered: Larry Bird
Short forward has surprisingly few candidates for the All-Time Team. Up until about the past decade, when the NBA started to showcase wingmen with tremendous athletic gifts like LeBron James and Kevin Durant, the position was not a premier one. Shooting guards and power forwards accounted for the majority of scoring; SFs were often role players who spaced the floor with shooting and provided defensive versatility, as typified by Bruce Bowen. The few historical exceptions were either high-volume, low-efficiency scorers (Baylor) or rebounding beasts lacking an elite all-around game (Cunningham, DeBusschere, Rodman).
Bird was a significantly better rebounder and distance shooter than James. James is a better passer who also turns the ball over a smaller percentage of the time. James, who seems to continuously improve in terms of scoring efficiency, has pushed both his true shooting and effective shooting percentages higher. He also makes up for his lesser (though dramatically improved) three-point shooting with an uncanny ability finish drives to the rim, which results in him attempting three more free throws per 36 minutes. In my book, free throws are the single most valuable source of points: they come with fouls which get the opponent into trouble and allow your defense to set itself on the next possession. In terms of defense, LeBron is clearly superior. Bird was a crafty and underrated defender but lacked lateral quickness; James not only blocks more shots than Bird (their steals percentages are close) but has more versatility with his quickness. In the end, that's what puts LeBron ahead for me: James has only a slight offensive edge but is a unique defensive talent. His win shares are significantly higher than Bird's which validates my conclusion.
PF - Tim DuncanConsidered: Karl Malone
This was a tough choice between two excellent players. Looking at their careers, a dichotomy becomes clear: Malone was a superior offensive player while Duncan is a better defender. Malone would benefit this team with his uncanny ability to run the floor for a man of his size and his scoring efficiency (57.7% career true shooting as opposed to Duncan's 55.1%). Malone actually wasn't a traditional big man offensively, however: he scored off of pick-and-rolls (playing with Stockton helped a lot here), dribble drives, fast breaks, and the occasional jump shot. He used his quickness as much as his strength to overcome defenders.
Duncan, on the other hand, is more of a traditional back-to-the-basket big man. He scores mostly via post-ups but also pick-and-rolls. While Malone was also an all-NBA defender in his time, Duncan's defense stands head and shoulders above: Duncan's block percentage (4.6 career as opposed to 1.5), defensive rebounding (26.5% career to 23.5%), and defensive rating (an incredible 95 to Malone's respectable 101) are all ample evidence of this. He manages to defend excellently while committing half a foul less than Malone per 36 minutes as well.
In the end, Duncan's defense outweighs the offensive benefit that Malone would bring. Also, while it'd be nice to have Malone's ability to run the floor coupled with the other fast-break superstars on this roster, it's actually even more appealing to have a post-up player in the mix. Duncan's presence down low could give the perimeter players a bit more space to take jump shots and drive into the paint. Since my center pick below isn't going to provide that low-post scoring, it's good to get it out of the power forward. Duncan's not just a good defender for a PF either, he's arguably one of the best defenders the game's ever seen at any position, but he still can't compete with my pick for center.
C - Bill RussellConsidered: Wilt Chamberlain, Shaquille O'Neal, Hakeem Olajuwon, Kareem Abdul-Jabbar
Center is, by far, the most difficult position to make a decision. The NBA has showcased many great centers, all of whom affected the game at both ends of the floor tremendously. Their high shooting percentages, team-leading rebounding, and massive defensive impact has historically made them basketball's premier position.
Bill Russell is also probably the most noticeably flawed player on this list: he's not a good shooter by any means. Russell tended to shoot mid-range jump shots, the NBA's worst shot, resulting in a miserable effective field goal percentage of 44 and a true shooting percentage of 47.1. However, he was an excellent rebounder, above-average passer, and superlative defender. While we don't have block and steal statistics for his era, he lead the league in Defensive Win Shares an unmatched ten years in a row and in eleven of his thirteen years in the NBA. In fact, his Defensive Win Shares are probably the most aberrational statistic in the entire NBA; they're almost 40% greater than the next best player (Duncan). He is the best defensive player ever. He won more championships than anyone else.
While I listed many other centers who rightfully belong in the consideration, the only one who gives me serious doubts is Wilt Chamberlain. Chamberlain and Russell were contemporaries and there is ample evidence that Chamberlain was a superior player. Chamberlain had more win shares, shot an incredible percentage, and was a good (if not at Russell's level) defender. Yes, Russell won many championships, but on a set of deep Celtics teams that featured other superstars. Russell won 5 MVPs to Chamberlain's 4. In the end though, I have to pick Russell. In filling out a roster, you want someone who makes sense given your other players. Russell is a defensive anchor who doesn't need to put up shots offensively. He fits in any lineup. If I could have one player to build a franchise around, it would be Russell and I wouldn't regret it for an instant.
CaveatsTraditional NBA and modern (1980 and on) NBA stats are not comparable because of differences in pace and the three-point line. Older games had more shots, more misses, and stratospheric rebounding totals. New games benefit from the three point shot and new statistics, such as blocks, steals, and plus/minus figures. The three pointer is an ongoing problem; the line keeps getting moved back further. Contemporary players are shooting more difficult threes than Larry Bird did in the 1980s. I tried to correct for the NBA's changing rules but the effort is ultimately futile; we cannot know if the classic greats of the game could compete with even mediocre modern players. My unsupported guess is that athletes have evolved; LeBron James would crush Oscar Robertson if the two competed in their prime.
I do want to take a moment to point out that David Stern and the NBA office clearly have an agenda behind their recent rule changes. They keep pushing the three point line back, a couple of inches every couple years now it feels like. They create new rules (no hand checks, the "no charge" semi-circle, the unspoken ban on travelling) which benefit drives. They are trying to generate dunks with rule changes. The contemporary NBA discourages long jump shots, zone defense, and perimeter play in favor of drives, isolation matchups, and flashy dribbling. Whether that is right in any sense is clearly irrelevant; it's a marketing choice and dunks are exciting. But I do wish somebody (for all the innumerable talking heads, I have yet to hear anyone mention what I consider to be an evident trend) would talk about it.