Backpicks GOAT: #29 Reggie Miller

Key Stats and Trends

  • Historically good combination of scoring volume and efficiency
  • Unrivaled improvement in the postseason, driving elite team offenses
  • Fantastic consistency for more than a decade

Scouting Report

Miller was the progenitor of the modern 3-point centric wing and one of the most economical players in history. He had a below average handle, wasn’t a great leaper and lacked the footwork and panache of most skilled isolation scorers. But his decision-making made up for it, combining pinpoint shooting with deceptive quickness to attack defenses in a handful of ways.

In the half court, Miller was a three-level player, scoring from beyond the arc, the midrange or at the rim. But he went about this in a unique way. Many of his rim attacks involved few dribbles, like this staring contest with Michael Jordan in Miller’s rookie year, a year in which Jordan won Defensive Player of the Year:

In that clip, Miller used a ball screen, but sometimes he isolated on the wing and exploded with a dribble or two:1

At the second level, Miller almost exclusively called upon his leaning floater, shooting this anywhere from seven to 18 feet out:

This shot was unique enough, and Miller quick and aggressive enough, that it led to a sizable chunk of his free throw attempts too. The moment Miller felt contact or noticed his defender reaching, he would lean into the shot and rip his arms through like this:

While this flopping foul-seeking has become commonplace, it was rarely used before Miller. Of course, when we think of Reggie, we think of the third level, above the arc, where he’d fly off a screen and uncork a bomb:

Miller was the best player in history at using screens (rivaled today by Steph Curry), mixing savvy with quickness (and some clandestine pushing) to shake defenders. He had an amazing feel for when to flare, when to pop to the top or when to dive to the rim, as he did in this clip:

Miller’s cuts were a read-and-react game of cat-and-mouse. If the defense played the screen differently, he’d just as quickly fade to the corner. His off-ball movement was so good, and so relentless at times, that he didn’t always need screens. Here, Nick Anderson looks like he’s trying to stay with Barry Sanders:

And really, at the core of it all was the simple idea that Miller could not be left alone to shoot. (He was one of the most accurate volume 3-point shooters in NBA history.) This threat exerted value in two ways. First, it enhanced his creation despite posting low assist numbers — below is the classic example of the off-ball shot created:

Miller never touched the ball, but he vacuumed two Knicks toward him to free up a teammate for a layup, something he would do countless times throughout his career.

Second, Miller’s presence prevented defenders from properly helping off of him. This is the “spacing effect,” where defenders can’t collapse into the lane for fear of leaving a good shooter, but Miller’s effect is magnified because defenders are glued to him everywhere. “Gravity” has been used to describe this phenomenon, but I think in Miller’s case a more appropriate term is “tethering” — he keeps defenders close wherever he goes. They are then so preoccupied with his cuts that they can’t properly help:

At that point in time, most defenders needed to face Miller (not the ball) to chase him, rendering help an afterthought. In sampling Miller’s games over the years, I counted a clear tethering effect on layups like about once every 100 possessions.2 While this might not seem like much, it’s a high-leverage play that adds to impact not measured by assists. There’s the more subtle, classic effect as well — teammates like Rik Smits were often free to work in isolation because doubling off of Miller was a bet most teams would not place. Based on my tracking, Miller’s creation rates were around 4-5 per 100, with some mental curving required for his tethering / spacing effect (captured in his Box Creation estimate).

With that said, Reggie was not a great passer. He was aware and intelligent enough to make good passes — in my tracking, at about half the rate of John Stockton, so not bad — but some of his passes were too slow. He did, however, have the occasional great find in him:

He made up for his passing with excellent decision-making, moving the ball to vulnerable areas or — especially once Larry Bird became coach — finding opportunistic cutters while rarely taking questionable shots himself.

As mentioned, Reggie wasn’t a great ball-handler, and he would often losing it in traffic like this:

I tracked a “fumble” on about 2 possessions per 100, reflective of his struggles with the handle. Many of these dribbling hiccups weren’t turnovers, but they would derail the play.

On the other side of the ball, Miller was a feisty man defender, using his quickness and length effectively. Yes, the man he’s guarding in the next video is peak Michael Jordan:

And later in his career, with slightly less foot speed (and verticality), he bothers Michael into a tough shot:

Miller was vulnerable to penetration at times, and in my sampling he was slightly above average in defensive error rates. Although there were times when his awareness and positioning paid off, like this:

Notice how Reggie wisely cheated toward the hoop to put himself in the right place to react to a Ewing lob. However, there were plenty of moments where Miller took too many gambles or landed out of position — no doubt wanting to leak out in transition — such as this next one. He takes a nonsensical route back to Starks (and isn’t between a Knick and the hoop) instead of racing back toward the rim:

Overall, Miller was a mixed bag on D. He could be above-average on the ball, smart in team defense and had pesky hands, but he wasn’t a great defensive rebounder and took a good number of risks that backfired. But he maintained a solid level of play on defense until about 1997, when he started to slow down physically, and then again around the turn of the century, when his positive defensive attributes dwindled.

Impact Evaluation

Miller’s constant motion and crafty shot selection made him a hyper-efficient scoring weapon. Among players with 25,000 points, Miller is an outlier in efficiency, and among all players is second to only Kareem Adbul-Jabbar in seasons with at least 19 points per 75 and a relative true shooting percentage (rTS) of +8 percent or better.3 His emergence correlated with a decade of solid Pacer offenses, but his postseason brilliance pushed them into the elite.

Reggie joined an average Pacer team from 1987 that proceeded to have a nearly identical, and average, 1988 season. It wasn’t until 1990, when Miller leapt into his prime, that Indiana morphed from a below-average offense into a strong one, 3.4 points better than league average (rORtg). Those early teams were marginally talented but spun on a coaching carousel, churning through five coaches until they landed on Bob Hill for the entire 1992 season, playing at a 46-win pace (1.8 SRS), a mark they would stay above until 2001.

During those years, the supporting pieces were shuffled and upgraded, but only Reggie and Rik Smits remained constants. In ’94, the defensively-inclined Larry Brown took over as coach and the Pacers best on-ball creator (and arguably best isolationist) Detlef Schrempf was traded to Seattle for Derrick McKey in an offense-defense swap. Indiana jumped to a 51-win pace that year, regressing slightly on offense without Schrempf (+1.5 rORtg) and making the first of five Final Fours during the period. And in 1995, they upgraded a flimsy point guard situation by bringing in Mark Jackson, whose passing fit with Miller’s off-ball oscillations.4

Despite the moving pieces, Indiana posted eight rORtgs above +3.1 during Miller’s prime (!), peaking at +6.5.5 Reggie was metronomic, cranking out 11 consecutive seasons with at least 19 points per 75 and +6.5 percent rTS. His moderate creation rates and unexceptional passing prevent him from standing with the offensive giants, but his perpetual motion and additive spacing buoyed a decade of strong team offenses. Here’s how he stacks up against contemporaries in the scaled “Big 4” offensive dimensions of scoring, efficiency, creation and turnovers:

Miller’s three-year peak scoring is a level below Allen Iverson and George Gervin, but their efficiency pales in comparison to Reggie’s.6 Fittingly, Ray Allen, the player most historically linked with Miller, matches Miller’s efficiency with nearly an identical shape.

However, something remarkable happened in the playoffs. In the Second Season, most stars see a slight decline in their numbers, the result of facing harder defenses that game plan for them. But Miller shows (perhaps) the greatest improvement from regular season to postseason of any notable player in history. His scoring spikes with no drop in his efficiency. So, despite more modest regular season numbers, Miller’s prime scoring rates in the playoffs were in the 97th percentile, comparable to rates from prime Kareem Adbul-Jabbar, Julius Erving and Larry Bird, all while maintaining his elite efficiency. Here are the same “Big 4” data points, plotted using each player’s best three-year postseason stretch:

Miller expands in all four dimensions, separating himself from the pack in scoring and approaching Gervin’s rates.7 And while many of these players would decline if we examined other playoff seasons, any three-year Miller stretch from 1990-2002 tells a similar story. For instance, using the previous group, here are the players with consecutive postseasons between the ages of 32 and 34:

Miller’s typical postseason combination of scoring, efficiency and creation produced a similar Big 4 profile to the regular seasons of ’11 Dirk, ’16 Leonard and ’93 Barkley. And Reggie’s insane jump in playoff performance correlates directly with Indiana’s sustained playoff excellence.

Along with Magic and Kobe, Reggie is one of three players in history to play on two separate teams with five-year stretches of +5 playoff offenses.8 Below, I’ve plotted these “unique” five-year stretches by regular and postseason offense:9

During those years, the average defensive efficiency of Indiana’s postseason opponent’s was around 102, and yet the Pacers still scored 109.5 points per 100, near their regular season output. To borrow an economic concept, their Miller-centric offenses were inelastic against stiffer defensive competition.10 Not only did this decade of sustained playoff excellence coincide with Miller’s primacy, but when he missed most of the ’96 postseason with an eye injury, Indiana produced its only below-average postseason offense of the decade.

None of this is to say that Miller should take the lion’s share of credit and be viewed as an offensive megastar. It’s instead a reflection of how effective, and additive, his style was. In many ways, Miller is the poster child for a number of concepts in Thinking Basketball (efficiency, spacing, portability and scorekeeping biases), making life easier for Indiana’s isolation players like Schrempf, Smits and Jalen Rose, who faced fewer doubles because of Miller’s presence. Perhaps more importantly, Miller’s effect would remain on even better teams, as he would, for the same reasons,  make life easier for Jordan or Shaq too.11

Ironically, accolades have historically gone to floor-raisers like Iverson, and it’s worth calling out some of the All-NBA nods that were given to players instead of Reggie. During the ’90s, raw field goal percentage was all the rage, and the effectiveness of off-ball play wasn’t established; Miller’s pedestrian assist and rebounding numbers were viewed as severe limitations, and because of this, comparable (or weaker) offensive players were rewarded at his expense.12 Below is a comparison of all-league candidates, with Miller’s box plus-minus (BPM) in blue circles and his scaled adjusted plus-minus (APM) in blue squares (available since 1994). The red circles and squares are the same data points for the players given an All-NBA nod over Miller:

All of these players struggle when stacked up against Miller in the Big 4 offensive dimensions, and none of them were figureheads on quality offenses. Richmond’s best offense in Sacramento was +1.5. Strickland’s best offense in Washington was +0.1. Dumars at least played on an above-average offense in 1990 (+1.8 rOrtg) and was given preference for his defense. But as you can see, both plus-minus data and BPM often favored Miller, and sometimes considerably so (e.g. 1994 and 1998). Simply put, voters missed the boat.13

Miller’s career reached into the Databall era, and his statistical footprint there is strong but not overwhelming. In scaled adjusted plus-minus, his first seven seasons (1994-2000) are above the 75th percentile, with three seasons between the 93rd and 96th percentile. His game-level plus-minus is steady, a rung below the superstars. So while Reggie lacks the indicators of a monster peak, all signs are that his economical scoring, spacing and moderate creation made him a valuable offensive weapon for a number of years. And we have modern evidence that a lighter version of Miller — a Miller-Lite, if you will — has a significant effect by warping defenses with his tethering and gravity.

While Miller is on many top 50 lists, I’ve never seen him in the 30s on any major GOAT publication. However, this particular exercise reveals the (surprisingly) incredible value he brought to Indiana over the course of his career. Not many players in history played well for so long, which is why he’s 20th in career VORP (since 1974), tucked behind Stockton, and 12th in career Win Shares. Most people have no issue ranking Stockton highly, yet balk at the idea of Miller landing on similar ground, despite neither sniffing a peak worthy of a weak MVP.

All told, I grade Miller with nine to 11 low-level All-NBA years — he should have been a mainstay on that team — and 13 All-Star seasons. If I’m slightly less positive about his defense, he’d still be pushing Bob Pettit for 30th in this exercise, despite a peak that’s outside the top-50. Even when knocking his valuations down a peg to “only” All-Star levels, his sheer longevity still results in a top-40 career. Given the evidence in his favor, I have a hard time scoring his seasons much lower. Either way, it’s safe to say that Miller authored one of the most underrated careers in NBA history.

Backpicks GOAT: #10 Magic Johnson

Key Stats and Trends

  • Led an offensive dynasty, quarterbacking a number of top-50 offenses
  • Elite combination of scoring efficiency and creation
  • Average longevity — took a few years to ramp up to prime levels

Scouting Report

If there’s a single word to describe Magic’s offense, it’s exploitation. He was the best in NBA history at exposing weaknesses, constantly finding a pressure point and attacking it. Sometimes he did this in the half court, sometimes with his own scoring, but most of the time it was on the fast break. Showtime!

In his early days, Magic’s exploitation dial was set to simmer, not boil. He had yet to establish an outside shot or a post threat, and most of his scoring came from opportunities created by teammates, transition scores or the occasional foray to the hoop. When he saw an opening, he’d selectively attack like this:

He was judicious about calling his own number like this, but even in his pre-prime years he could ram the ball down your throat if you didn’t stop his momentum. Here’s a typical high-speed charge of his:

If defenses didn’t stop the ball, he made them pay dearly, advancing at full speed until someone blocked his path:

By his fourth or fifth season, Magic unleashed an uber-aggressive mode, where nearly every time LA found an odd-man advantage he would hit an open slasher for an easy finish. (He even barreled over a few defenders in his day for offensive fouls.) He seemed to always find the right guy for a layup, making the Laker break the most efficient weapon in the game.

Many of young Magic’s passes were too aggressive, leading to deflections or turnovers. He often flirted with danger, looking for small advantages to expose, and his decisions weren’t as crisp as they would become in his prime. Sometimes, a tight pass slipped through, but many of these were turnovers:

As a result, giveaways were a problem. Among high-rate passers,1 Magic’s 1984 adjusted turnover rate ranked in the 5th percentile (only 48 seasons were worse). His turnover rate over his first seven years places him around the 18th percentile, reflecting the risk-taking visible on tape. (His assist to turnover ratio wasn’t much better.) Great passers take these risks (and commit more turnovers) because they see receivers when mere mortals see blockers, but Magic was overly zealous in his early years.

He didn’t have much of an outside shot in those days either, and he certainly didn’t pull the trigger on it often. Here’s the typical amount of space defenses ceded in these sets, especially if Kareem was on the strong side:

But in ’84 and ’85, he started to rain scores from the perimeter. He called his own number a touch more, mixing drives in with his jumpers. He started to back smaller defenders into the post:

This became the foundation of Magic’s scoring arsenal. By ’86 and ’87, he completely leveled up, regularly nailing deep jumpers when his defender sagged off of him. He also used the pick-and-roll more effectively in the half court. But Magic’s ultimate weapon was still that exploitative passing.

As the years went on, he cleaned up the overly risky passes and mastered whipping bullets into spots that others wouldn’t imagine. Here’s an example from 1983, on the doorstep of his prime, where he could size up the defense like a pre-snap quarterback and manufacture easy points:

If you were watching the ball, you probably missed the open receiver. Magic started throwing these kinds of high-leverage assists nightly, catching defenders with their heads turned the wrong way. Layups materialized from the most innocuous moments, almost like…magic.

This isn’t traditional creation, where the defense responds to a scoring threat. This is straight up exploitation — give Magic an inch, he takes a mile. And as a result, his offensive efficiency was always exceptional. Technically, this means he wasn’t great at pulling defenders away from his teammates, although his Box Creation numbers capture the value in finding minute advantages and facilitating fast breaks, which is largely reflective of how easy he made life for his teammates.

On defense, Johnson was fairly mobile when he entered the league, so Los Angeles made him a roamer in their half court trap. Magic used his instincts and length to jump passing lanes and gather loose balls and steals this way. (He led the league in steals per game twice.) Other times, he’d end up in weird positions, looking for someone to harass or guard without knowing where to go, but there was a madness to this method that worked on many possessions:

He almost always guarded the weakest wing player (and the occasional power forward), a trend that would hold throughout his career. On tape, his defensive positioning was fairly sound, although I tallied a moderate number of slow or missed rotations — notice in this clip how he essentially has no reaction to the play unfolding (lower-right corner):

In the next clip, he is caught in a “looking to guard” someone mode, but completely missed helping at the rim as needed:

Even when he did provide help, it wasn’t too effective. Despite standing 6-foot-8, his ability to deter or affect shots around the hoop was largely nonexistent. Some examples:

Those were typical Magic plays near the rim. He essentially never blocked shots in this area; for players 6-foot-8 or taller, Magic’s block numbers from 1985-91 rank in about the 5th percentile.

He also had a habit of gambling for steals in the backcourt (consistent with his aggressive mentality). In tracking about eight games worth of possessions, Magic stole the ball once and whiffed six times. This often led to odd man breaks, as it did below. That blur headed for the front row was Johnson:

However, he reclaimed some defensive value with elite rebounding. Among perimeter guys, Magic was above the 92nd percentile in every season in relative defensive rebounding rate, peaking in the 99th percentile in 1989.2

As his career wore on, Magic abandoned the roamer role and played smarter defense, but his foot speed declined. His offensive aggression ramped up in the mid and late-’80s, and in his last few seasons — still very much in his prime — Magic became a more dangerous half court threat, relying on post play and strong outside shooting.

Impact Evaluation

Magic piloted one of the league’s great offensive dynasties, maintaining elite status despite the decline of Kareem Abdul-Jabbar. But Johnson’s impact can be hard to gauge for two reasons: First, he entered the league early and steadily improved from year-to-year, making it difficult to find a discrete, clear signal from his presence. Second, the Lakers were loaded offensively during his best seasons, so it’s hard to tease out exactly how valuable everyone was. But a closer look at the data suggests that his value increased throughout the ’80s and that he might have peaked as the greatest offensive player ever.

In 1979, LA was anchored by Abdul-Jabbar, an offensive juggernaut at the time. The Lakers were coming off of three consecutive 50-win pace seasons (when healthy), powered by good team offense. Suffice it to say, Magic joined a strong offensive club, especially since rookie coach Paul Westhead was more attack-oriented than predecessor Jerry West. Magic’s role as a rookie was tertiary and, as such, his impact doesn’t jump off the page. LA’s offense did improve to best in the league with Magic, Jim Chones and Michael Cooper on board, but the shift was only a 2-point improvement from the ’79 team. Good, but not great.

In 1981, Magic missed about the half the year with an injury, giving us insight into his early, swiss-army knife value. Without him, LA played at a 47-win pace (2.1 SRS) and with him a 54-win pace (4.7 SRS). Given the large samples, I view this as fairly reflective of his importance to that team — he gave them a solid boost, but wasn’t even LA’s primary point guard yet (the venerable Norm Nixon was). We can use game logs for a more granular view of the team’s changes with and without Magic that year:3

The Lakers were only marginally better on offense and defense without Johnson. (The negative defensive change is good.) Consistent with historical results, the best isolation player — the player least dependent on others for his offense, Jabbar — didn’t need Magic at all. (Kareem’s numbers were actually better without Johnson.) Jamaal Wilkes, another skilled scorer, didn’t miss him either.

After acquiring Bob McAdoo and installing Pat Riley as the headman, the ’82 and ’83 Lakers played at an identical 59-win pace (6.4 SRS) when healthy.4 With Nixon gone in ’84, Magic was (finally) handed the keys to the Showtime car, and was nice enough to miss a chunk of time that year that we can analyze too:

The Lakers actually shot more efficiently without Johnson. This is a small sample, so it shouldn’t be viewed as definitive, but look at what happened to James Worthy. He increased his scoring and efficiency over those 15 games without Magic, which reflects his considerable skill as a scorer. Like Kareem in ’81, Worthy didn’t need Magic for his offense and was capable of ramping up if given more touches. Cooper, the de facto backup point guard, spiked his assists by 50 percent while pinch-hitting for Magic, averaging 9.3 per 36 minutes. (Magic himself averaged 12.3 per 36 that year.)

Once Johnson hit his stride in 1985, the Lakers attack operated in rarified air. For six consecutive seasons, LA posted a top-50 offense of all-time (by relative offensive rating, or rOrtg), finishing with a relative efficiency at least 5.9 points better than average in five of the six years. No other team in history has even had five top-50 seasons in a six-year span, and only the ’90s Jazz and 2000s Mavs had four. The Lakers six-year average rORtg of +6.1 trails only the mid 2000s Mavs and Suns for most dominant offensive stretches in NBA history.5 They were one of the best, if not the best, offensive dynasty in history:

This dynastic offense correlated with Magic’s own increased load, as he bumped his scoring and creation in 1985, an indicator that he was pressuring defenses more. Over the next few years, his scoring continued to rise as his efficiency dropped off, indicating that he called his own number more. His combination of scoring, efficiency and creation paint him as one of the great offensive forces in NBA history:

While some of the players in the upper-right of that plot score far more than Magic, very few ever matched his balance of efficiency and creating. In those two areas, he’s comparable to players like Jordan, LeBron and Bird, although he lacks individual scoring when compared to those offensive engines. However, Johnson makes up ground in most comparisons with a passing edge.

In 1986, he missed nine games — a small sample for highly variable metrics like scoring efficiency or point differential, but large enough to glean something about a more stable metric, assists. As in ’84, Cooper upped his dimes without Magic, from 6.8 to 10.8 per 36, only about 2 off of Johnson’s own season average.6 Cooper was a very good passer, and could set up players in the half court (or even on the break) well enough to support their scoring skill, but there’s little to suggest that he was a hidden game-changer on offense.

Worthy’s efficiency remained the same in those nine games without Magic in ’86, although his assists increased by 30 percent while his scoring volume dropped a touch, implying that he swapped some isolation for playmaking. This pattern repeated itself in 1988, when Magic missed 10 games and Cooper missed seven, leaving LA with replacement parts at point guard (Milt Wagner or Wes Matthews). This time, Worthy upped his scoring (nearly 2 points to 21.3 per game), his efficiency (from 56.7 percent to 59.2 percent true shooting) and his assists (from 3.7 to 4.8 per game). Byron Scott, yet another talented Laker, improved his scoring (up 3 points to 24.4 per game) and increased his assists as well.

These results — LA’s offense was +1.8 in those 19 games, down from +5.9 with Magic and Cooper — imply that Johnson played with a strong offensive cast. When Magic wasn’t there, his ball-dominance was redistributed among threats like Kareem and Worthy — themselves a formidable duo — who could either create their own offense or create for others.

In ’87, Magic authored his magnus opum, leading the same rotation from ’86 to a 66-win pace (9.5 SRS) and a mind-boggling 119.9 offensive rating in the postseason, a record that would stand until Cleveland posted a 120.3 mark in the 2017 playoffs.7 The ’88 and ’89 Lakers regressed slightly and then Kareem retired. With the firepower dwindling, the results still remained — LA maintained a win pace around 60 thanks to its elite offense in ’90 and ’91 —  a testament to Magic’s floor-raising skills.

Injuries and aging complicate any analysis of Magic’s first retirement. Vlade Divac missed half of the 1992 season, and when he returned, Worthy — rapidly declining with age — missed the remainder of the year. The Lakers finished around .500, and in their only full-strength stretch (all of 11 games) they played at a 50-win pace (2.9 SRS). The offense finished right around average. Even five years after HIV abruptly ended his career, Johnson’s presence helped the ’96 team on offense (while hurting the defense): LA posted a +2.3 rORtg (51-win pace) in 38 games with Eddie Jones, and then a +7.4 rORtg (59-win pace) in 32 games when Magic suited up next to Jones.

Collectively, the film and data scream that Johnson was one of the very best offensive players in history. His WOWYR numbers are fantastic, finishing first in the 2016 results, and near the top in all regressed game-level studies. His team’s offenses were even better in the postseason, improving by a weighted average of 2.5 efficiency points. However, Magic’s defensive work dings him somewhat among the other greats, as he was likely a neutral-impact defender in the early part of his career before his defense waned in later seasons. But it’s his longevity that costs him most on this list, as HIV stole valuable prime years for him to climb up the top-10.

All told, I consider Magic’s peak just short of the all-time greats, not only due to his defense, but because his ball-dominance introduces redundancies on good teams. On the flip side, because his offenses were so good, and his style somewhat unique, I can see an outside argument for him reaching Wilt and the next block of players. Either way, what’s clear is that with two more years comparable to his ’91 campaign, he would move up multiple spots, and with three similar seasons would be pushing the top five. Instead, he narrowly edges rival Larry Bird for the 10th-most valuable career in NBA history.

Backpicks GOAT: #11 Larry Bird

Key stats and trends

  • Led some of the top offenses of all time with elite combo of scoring and creation
  • Marginal efficiency in some years, scoring drop in postseason in other years
  • Injuries dented career value, limited longevity

Scouting Report

Bird was the greatest touch-for-touch passer in NBA history. There have been entire highlight videos devoted to this, so I won’t spend too much time on it, other than to illustrate his unparalleled combination of vision, anticipation and aggression. He was the master of the touch pass, a nod to his all-time level court awareness, where he knew what was going to happen long before everyone else. A simple, but typical, example:

Bird was abnormally aggressive with his passing. Because of this, so many of his dishes unearthed high-percentage offense for teammates that few players could ever find, even when nothing seemed to be there. Behold:

He was a master of fitting the ball into these tight windows and possessed this laser vision from Day One. Yet, for a creator of his proficiency, he protected the ball well; his ’84-88 turnover frequency places him in the 78th percentile among comparable creators.1 Here’s another Bird tendency, where he realizes there is an outlet opportunity, and is ready to throw it before he even grabs the ball:

They’ve never tracked “outlet layup assists,” but if they did, I imagine Bird would be near the top of the leader board. This passing, combined with strong scoring, made him a top-notch creator for his time — his Box Creation seasons land him between the 95th and 98th percentile historically, despite individual creation being less prominent in the early ’80s.

Bird was also the best off-ball forward ever, so much so that I’d classify his game as primarily off-ball. Watching him without the rock, particularly in the first (1980-83) and second trimester (1984-88) of his career is a study in advantageous positioning. Here’s a 30-second sample of Bird spinning, cutting, banging, boxing and constantly threatening the defense with his high-motor perpetual motion:

This helped him as a rebounder as well; on possessions that ran through his teammates, Bird often sealed inside position on his man (as he did countless times in the previous clip). Among the 1,000 best scoring rates in history, Bird’s offensive rebounding among players who took at least 100 3s — players who didn’t camp under the hoop — peaked in the 91st percentile.2 All of this movement (and rebounding) created value without the ball.

Boston constantly featured him in a stacked set like this one:

We usually associate this kind of action with Reggie Miller types, but a huge chunk of Bird’s game was flying off screens, even as a 6-foot-9 “power” forward. For many of his early years, he was either in the post, or flaring out/curling off of multiple picks for catch-and-shoots:

And boy were young Bird’s actions decisive. Even when he didn’t catch-and-shoot, he’d use the momentum of the play to transition into a dribble attack:

Although in his early seasons, his quick decisions occasionally led to poor shot selection. There were plenty of these back then:

Bird’s propensity for outside shooting and crafty scoring limited his free throw attempts. While he frequently operated in the mid and low-post, his tendency, even on drives, was to shoot runners, floaters or fadeaways. In this area, his lack of explosive athleticism betrayed him; he wasn’t dangerous enough on rim attacks to constantly draw fouls, instead opting for plenty of these:

His free throw rate (a ratio of shots to free throws attempts) during his best seasons ranks in just the 22nd percentile among 20-point per game players. Free throws spike efficiency, and Bird’s inability to generate them hurt him relative to other great scorers. Despite his shooting skill, his long jumpers yielded only a moderate return on investment in all likelihood. While clearly a great shooter, he wasn’t great enough for his bombs to offset his lack of easy scores. Of course, Bird counteracted this by bending defenses with his movement and post-ups, which only played into his hands as a creator.

Offensively, he peaked during his second trimester, improving his overall shooting, shot selection and incorporating the 3-pointer more. But his defense changed radically over these periods. When he entered the league, he combined that perpetual motion, unparalleled court awareness and large frame to produce a fringe all-defensive player. His rotations and positioning jump off the film, as his 360-degree vision made for quick reactions like these when a teammate needed help:

And in his first few years, Bird had the quickness to contest shots around the rim or even block them. Here, per usual, he instantly moves into position and jabs Kareem’s shot away:

Young Bird was also extremely active around the hoop, challenging or blocking shots in all kinds of ways:

The anticipation that guided his passing made him an excellent disruptor of passing lanes. Here, Kareem thinks he has Magic for a layup because most defenders wouldn’t jump this pass — Bird is “supposed” to flare out with his assignment — but Bird reads the whole thing and makes it look like he was the intended receiver on the play:

This foresight, along with his relentless motor, made him an elite rebounder in those years. He would frequently outmaneuver his man while anticipating the flight of the ball:

Bird’s defensive rebounding rates in his first few years place him near the 80th percentile historically and closer to the 90th percentile in the postseason. But his athleticism and energy started to wane in the second act of his career. He maintained his clairvoyance — reading passing lanes and seemingly always positioning himself in the right place — but his foot speed and leaping faded, and so did some his defensive efficacy. Here’s a play from the ’84 Finals where he switched to deny Kareem, then considered Magic, then recovered to play Rambis well.

I wonder if a younger Bird would have been quick enough to outright steal that pass. At that point, his foot speed impacted his change of directions, leading to off-balance plays that prevented a good challenge, even when he managed to maintain guarding position:

Under the right conditions — in help defense or against slower post players — plays like this were still common in the mid-’80s:

As Bird grew older, he became more vulnerable to quicker perimeter opponents. Fortunately, he played most of his career as a defensive power forward, and even in later years, when Boston went with its big frontline of McHale and Parish, Bird checked the opposing big forward on most occasions. But in those final three seasons, all that was left was anticipation, good hands, and sound positioning.3

Impact Evaluation

Before losing his mobility, Bird was a key cog on premium defenses for much of his prime. His scoring and table-setting powered some of the best offenses the league had ever seen, although Boston’s postseason attacks lagged while Bird struggled to score at high efficiency (or stay healthy) during the playoffs. At the same time, the Celtics were impervious to key injuries over the years as long as Bird played.

He entered the league as a polished, 22-year old rookie, spearheading one of the biggest turnarounds ever (a 32-win improvement). It wasn’t all Larry — Boston brought in a new coach (Bill Fitch), Tiny Archibald’s health improved and poor-rep players like Marvin Barnes and Bob McAdoo were replaced on the bench.4 But it all centered around Bird. He took 19 percent of the team’s scoring attempts, the exact same number as MVP Kareem Abdul-Jabbar in LA.5 He handled the brunt of the creation and the result was an offense 4.2 points better than league average (rORtg), the 15th-best ever at that point in time.

Dave Cowens — not the same player from his MVP-candidate years — missed 15 games during the season. With Cowens and a full-strength lineup, Boston played like a 60-win team (7.1 SRS). Without him, they were even better, clocking along at a spectacular 67-win pace (10.0 SRS). Archibald played well at the point, and Cedric Maxwell added a hyper-efficient post up game.6

Bird’s passing was so good out of the gate that Rick Barry — a great distributor himself — called him the best passer in the league during one telecast. Longtime AP basketball columnist Alex Sachare wrote this at the end of the ’80 season:

“In Boston, the key man in the Celtics’ remarkable turnaround has been Larry Bird. His passing game helped the Celtics more than double last year’s 29 victories…While his scoring and rebounding certainly helped, the attitude he instilled was the most vital factor in the Celtics’ renaissance.

[An MVP vote] for either Erving or Bird would certainly be reasonable.”

Pressuring the defense starts with the threat to score, and Bird certainly did that. However, his scoring itself was not elite in his first few years, occasionally rushing shots and avoiding contact (per the scouting report). It wasn’t until he cleaned up his shot selection and used the 3-pointer more that he became topflight in this area:

Still, all-time level passing is perhaps the most additive and scalable skill in basketball. The worst offense Bird ever played on was about 2 points better than league average. After 1984, all of his teams were at least 4 points better than average on offense with him in the lineup (87th percentile in the 3-point era). He led the NBA in estimated creation in 19807 and his combination of scoring and playmaking pegs him as one of the greatest offensive weapons ever. Below are all of the 22 point per 75 players between 1978 and 1990, plotted by efficiency and creation:

Bird’s trifecta of volume, efficiency and creation was only matched a handful of times in the first 10 years of the 3-point line; by Magic (although his volume was lower, dropping him off this chart in some seasons) and by Jordan (with higher volume). Bird provided additional benefits over both of them as an off-ball threat and court-spacer — he was a big forward who didn’t need to occupy the post and could capitalize on his teammates’ creation with outside shooting.

As the team around Bird changed, his impact seemed to remain. In 1981, the Celtics played at a 58-win pace (6.1 SRS) en route to the title. In 1983, Archibald missed 16 games and Boston played at a 56-win pace (5.2 SRS) with and without Tiny. And in 1986, McHale missed 14 games and the team played at a 61-win pace (7.3 SRS) without him.

The ’82 and ’83 Celtics were playoff disappoints, transitioning from Bird’s early years into a team with Dennis Johnson at the point and Kevin McHale playing a more prominent role. Coach Bill Fitch, who largely eschewed the 3-point shot, was replaced by KC Jones.8

Meanwhile, Boston’s offense improved, commensurate with Bird’s individual growth. The ’85 and ’86 teams were just under 5 points better than league average offensively.9 In ’87 and ’88, Boston posted a staggering +7.3 rORtg in 148 games with Bird and McHale. Bird was the cornerstone of two of the 25-best offenses in league history, including the 1988 squad, the best of all time to that point, surpassing even the ’87 Lakers in regular season efficiency. Boston’s true shooting percentage was 58.8 percent that year, a record that would stand until the 2016 Warriors shot 59 percent in their 73-win season.

When Bird missed most of 1989, the Celtics fell back to earth. After five consecutive years playing at a 60-win pace or better, Boston dropped to a 45-win clip (1.3 SRS) under new coach Jimmy Rodgers, although there were other notable lineup changes as well that season:

Bird’s injuries were the black mark on his career, preventing him from ascending even higher on this list. There were a few small hiccups in the early days — an elbow injury and a mistimed flu in the 1983 playoffs — but Bird dented some of his best seasons by himself. In 1985, he was in a bar fight in the middle of the postseason, and his performance declined after his hand injury. That summer, he injured his back repaving his mother’s driveway in Indiana. In 1988, shin splints took their toll as the playoffs progressed and Bird struggled in the Eastern Finals against Detroit’s swarming defense. This ate at some of Bird’s value and led to an overall decline in playoff scoring. And while Boston’s postseason offenses were excellent, they fell further behind LA’s in most years.10

We can further examine the impact of Bird’s injury on his teammates, who played 40 full-strength games without him in 1989 before Danny Ainge was traded. Reggie Lewis, a fringe All-Star caliber player that year, essentially “replaced” Bird in the Celtics lineup, Robert Parish upped his scoring 4 points per game, but the offense still tumbled from historically good to slightly above average. The chart below reflects Bird’s effect as a creator and passer — every single teammate declined substantially in scoring efficiency without him, particularly McHale:

The back injury also limited Bird in the ’91 playoffs and render him nearly useless in the ’92 postseason. These were critical years for him to provide value to his team that were either completely lost or marginalized thanks to his breaking body. Although, even in those last few seasons, he provided an enormous lift to the Celtics, another indicator of how valuable his passing, spacing and shooting were.11 And despite his limited mobility (per the scouting report), the only season in Bird’s entire career in which he played on a below average defense was in 1988 (with nearly every season from ’80-86 landing near the top of the league).

If there’s a knock on his statistical resume, it was a lack of high-end efficiency in the postseason. Below, I’ve scaled the “Big 4” offensive box components of scoring, efficiency, creation and turnovers for him and some contemporaries; each axis represents the span of the metric across all post-merger seasons, so Magic’s scoring rate of 20.1 per 75 is about 42 percent of the way from the qualifying low (10.5) to Jordan’s best (33.2). Here’s how Bird compares:12

Given Boston’s balanced roster and Bird’s lack of ball-dominance, lower scoring rates weren’t an issue, per se, especially on such efficacious teams. However, his failure to spike efficiency while reducing volume was likely a byproduct of his low free throw rate and inability to burn opponents as a slasher. This limitation is picking nits — those postseason offenses were, after all, some of the best on record — but it’s the difference between Bird peaking as one of the five-best offensive players in history and the best ever.

While Bird’s prime WOWYR score (which includes 1990-92) is a good distance from the all-time best, other historical regressions suggest that he was likely having superstar impact.13 The version of that regression that I ran up until 1983 actually had Bird No. 1, inline with many of the impressive WOWY results he registered. These studies are consistent with the seismic shift in Boston when he joined in 1980 and the drop-off without him in 1989.

In total, Bird has the imprint of a GOAT-level offensive player with a clear case as a strong defensive factor early in his career. However, like so many other superstars, his offensive peak didn’t coincide with his best defensive seasons, making it less likely that he was a top-5 peak player ever. By my valuations, Bird’s first eight years are MVP-worthy, including one of the best two-year apexes in NBA history with an argument for the best post-merger rookie season. I do wonder how much credit to give his non-traditional offensive traits, but either way, minor adjustments up or down don’t move Bird much on the list. With a phenomenal peak and problematic longevity, he lands here, barely behind No. 10.


Backpicks GOAT: #24 Moses Malone

Key Stats and Trends

  • Offensive value from dominant rebounding and scoring, not creation
  • Poor impact numbers suggest non-elite peak
  • Average meaningful longevity despite huge counting stats

Scouting Report

Moses wasn’t the Chairman of the Boards for nothing. He made a living from offensive putbacks, parking himself near the baseline to RSVP rebounding position, ready to pounce on interior misses. Unlike Dennis Rodman, another titan of rebounding, Moses wouldn’t chase long balls or tip misses out to himself at the 3-point line. Instead, he used his rebounding strategy as a gateway to get buckets, receiving bricks as if they were meant to be passes. In the following highlight, you can see this tactic and Moses’ quick leaping, along with some agile post moves and a dribble-drive:

Malone dominated the offensive glass unlike any other scorer in history. This rebounding-centric game kept Moses banging near the rim and generated huge free throws numbers. Among volume scorers, Malone’s ratio of free throw attempts to shot attempts for the majority of his prime seasons ranks above the 97th percentile.1

After a few years in the league, Moses added a face-up jumper to balance his arsenal of power moves. The following highlight from the 1981 NBA Finals demonstrates Malone’s offensive tendencies well:

Missing from that highlight — and most of Moses’ career — was strong passing. He lacked court vision and, as a result, was largely nonexistent as a creator, peaking in only the 3rd percentile in Box Creation among 24 point per 75 scorers.

Malone’s all-time offensive rebounding and limited passing forged a unique package.2 He couldn’t create shots for weaker players and the offense ran through him less than his scoring would suggest, but his dominant rebounding added off-ball value that fit almost anywhere.

Defensively, Moses was strong but not elite. He ended possessions with his rebounding and was a solid shot-blocker — his best season in Philadelphia ranked in the 63rd percentile among bigs3 — but his reactions seem a little slow on film, his coverage mediocre. Below is a cut of a 1984 game between Moses and Artis Gilmore that demonstrates his defensive strengths and some laboring movements. His midrange shot and rebounding, of course, are on display, but declining athleticism clearly chipped away at the quickness in his attacks:

Malone faded out of his prime after the 1985 season (and lost the ’86 postseason to an eye injury), but he continued to rebound well on both ends and produced four respectable post-prime seasons banging around the hoop.

Impact Evaluation

Despite three MVP awards, there’s limited evidence that Malone was a high-peak player. He was an impact-rebounder and viable isolationist, but his presence rarely correlated with meaningful team changes (likely caused by the aforementioned passing deficiencies and questionable defense). More detailed value-measurements are even less kind to him.

Moses entered the pros straight from Petersburg high school in Virginia, playing two seasons in the ABA before the leagues merged in 1977. He missed half of the ’76 season with the Spirits of St. Louis, and the team performed nearly identically without him.4 The following year, a 21-year old Malone averaged 31 minutes per game and posted a career-best 19.8 percent offensive rebounding rate, transforming Houston from a below-average offensive rebounding club to best in the league, a trend that would continue throughout his prime.

Since 1974, only 59 teams have posted an offensive rebounding rate at least 4 percent above league average (relative offensive rebounding rate), or about one in 20 teams. Malone was the rebounding force on six such squads, more than any other player by far. 5 Below, you can see that Malone’s presence correlated with massive jumps in his team’s rates (gray circles are the year before/after Moses, whites are the rest of the league):

He ultimately landed in Houston, playing 31 minutes per game in 1977 before taking giant strides in ’78. Malone missed 23 games that year, and without him the “healthy” Rockets played at a 21-win pace (-7.3 SRS), but with him only a 29-win pace (-4.1 SRS).6 Although he was just 22 and rapidly developing, that’s a less-than-desirable result for any star-level player.

In ’79, Houston’s performance improved in conjunction with Moses’ statistical growth as he entered the heart of his prime. That Rocket team, featuring jitterbug scorer Calvin Murphy, Rudy Tomjanovich and a long-toothed Rick Barry, produced a formidable offense (+4.9 rORtg) while playing cringeworthy defense. The net result was barely above neutral (0.9 SRS) and an MVP nod for Moses that looks stranger and stranger over time.7 Until 1982, the Rockets continued to spin on the treadmill of mediocrity, swapping coaches and tweaking the rotation with little effect before filling the ’83 team with replacement parts in a tankathon.

When Moses arrived in Philadelphia in 1983, he joined an upper-crust club that had reached the Finals two of the three prior seasons. The 76ers crushed the league that year, finishing with a defense 3.8 points better than league average and clocking along at a 64-win pace at full-strength (8.8 SRS). But that excellence evaporated in 1984, despite no notable roster changes and a core with five players between 26 and 28 years old. There were grumblings of disappointment from ownership about a lack of effort and the Sixers sputtered to a 52-win pace (3.7 SRS).8

In ’85, Philly bounced back, playing at a 58-win pace (6.0 SRS) with the addition of rookie Charles Barkley. But again, the team regressed in ’86 (50-win pace, or 3.1 SRS) under new coach Matt Guokas as Moses and Erving aged and scoring dynamo Andrew Toney missed most of the season with stress fractures in his feet. Malone was traded to Washington for the ’87 season, where an overhauled Bullets squad played at a pace nearly identical to their ’86 team (38-win pace or -1.0 SRS). Meanwhile, back in Philadelphia, with Moses and Bobby Jones gone and Erving in his final year, the 76ers treaded water as an average team (0.8 SRS when healthy).

Fortunately, we have Harvey Pollack’s plus-minus data for Moses’ four seasons in Philadelphia to help evaluate his impact. His AuPM oscillates between strong (but not transcendent) and pedestrian in those four seasons, with ’83 and ’85 looking like typical top-20 seasons and ’84 and ’86 lacking impact. Similarly, Moses’ regressed game-level data tells us that he made a difference, but that his impact was far short of a Grade-A superstar’s.

While his rebounding and physicality likely made him a positive defender — I certainly view him as such — the case for any kind of considerable defensive impact is lacking. Malone played on six below-average defensive teams in Houston, some of which were dreadful, and the Rocket D didn’t collapse after he left. In Philadelphia, the Sixers generated two strong defensive years in his first two seasons (3.8 and 3.0 points better than average, respectively). However, with the core of the team intact (save for Caldwell Jones), Philly’s three-year defensive efficiency peaked in 1981 and ’82 before slowly dropping off in the Malone years.9

Moses’ rebounding does scale well because it’s off-ball — good shooting teams would be even better with more chances — but the rest of his isolation scoring game does not make him a desirable offensive centerpiece; he was a finisher, not a creator. His best years were brief (’79-85), and he missed the 1986 postseason, chipping away value from his last prime year after feuding with coach Guokas. So, despite tallying a ton of games, I only credit him with seven All-NBA seasons, which, combined with a peak outside the top-25, prevents Malone from serious top-20 candidacy. There’s an argument based on his statistical portfolio that he belongs closer to 30th, but that involves dinging his defense slightly beyond my comfort zone. As such, he falls comfortably between 23rd and 26th, earning the 24th spot.

Backpicks GOAT: #16 Julius Erving

Key stats and trends

  • Excellent scorer and offensive centerpiece, but game doesn’t scale well
  • Part of strong defensive teams but had questionable defensive habits
  • Inconsistent — represented poorly by impact metrics

Scouting Report

Erving was a prototypical “scorer.” Despite marginal shooting accuracy, his explosiveness and size unlocked angles and pathways most could never access. The Doctor operated in three main phases: the low-post, the mid-post and the fast break.

In the post, Erving excelled at finding deep positions and sealing opponents, where he could then leap and twist with enough athleticism to finish around the rim:

Second, especially in Philadelphia, Erving operated out of the high/pinch-post area. He would catch and face from this spot, looking to drive (especially to his right) or take a jumper:

He had an arsenal of finishing moves like this on his forays to the rim (thus the “scorer” moniker). He would sometimes counter back to his left to create space for the jumper, or simply launch it as he did in the first clip. It was an effective shot, although based on his shooting percentages and my assessment of the film, it was somewhat inconsistent.

Erving’s third (and probably most dangerous) mode of scoring was in transition. Games were faster then, there were more opportunities, and Erving displayed many of the same superpowers that make LeBron James a one-man efficiency-spiker in the open court.

The speed stands out, especially in the second clip, where he’s able to accelerate past everyone and finish with one dribble from the 3-point line. The open court weaponized his athleticism:

Erving was also a decent creator for his time. He lacked great court vision, but his passes were often aggressive, high-reward dishes, and sometimes they payed off. Here, he (almost) instantly recognizes the double and creates an open jumper:

And another opportunity created, at the rim this time:

Erving was quite good at finding proximal passes — tight passes to the same side of the court, often to the immediate outlet valve or a flashing cutter:

But that was the extent of his vision, and he rarely found anything valuable beyond this “first level.” In all the Erving games I’ve seen from his prime, he probably threw one skip pass to the opposite side of the court, and that was a loping, awkward duck that looked like the quarterbacks arm was hit while he was throwing.

Still, in these clips, it’s clear that Erving’s scoring — particular in the post — strained defenses, as they reacted quickly and he was able to (occasionally) make them pay by finding an open teammate. But spacing wasn’t great in the NBA then, and opponents could throw “free double-teams” at him:

These packed lanes were a common countermeasure against Philadelphia in the late ’70s. Sports Illustrated described it like this:

“In the NBA, however, everybody doubles up on him, which is natural, but teams also pack defenders down low, clog the lanes and (sh, keep this a secret now) zone the bejeezus out of the Doctor. This makes it practically impossible for Erving to consistently drive to the hole for the swoop baskets by means of which he developed his Dr. J reputation.” (March, 1979)

Between his occasional habit of shooting over double-teams, limitations in his vision and his team’s problematic spacing, his creation rates were still low compared to the offensive engines of today. The non-3 version of Box Creation pegs his creation rates at around 5 to 7 per 100. I’ve hand-tracked 12 of his games from the early ’80s, when his creation looked better on film, and in those games he was slightly over 4 per 100. For that era, that’s somewhere between a top-10 to top-30 creator each season — very good, but not elite.

All that athleticism allowed him to protect the rim on defense, and he was a good shot-blocker for a wing. At his best, he was around 2 blocks per game, and a decent percentage came from high-leverage plays like this:

His defensive rebounding rates were above average for a non-big and his “stocks” (blocks and steals) were elite for a non-big, tallying over 5 per 100 in multiple seasons. Since 1978, only 18 non-bigs have averaged 5 stocks per 100 (minimum 1500 minutes) and only seven players did it for multiple seasons.1

He wasn’t without warts though. For most of his prime, he wanted to leak out on the fast break in every game I’ve seen. There were numerous examples, like the one below, where Erving never boxed out (a poor habit of his) and instead inched toward the other end. When this didn’t work, it often ended poorly:

He also lacked foot speed when guarding the ball. Here’s some matador defense against the smaller Louie Dampier on the perimeter:

And if you’re thinking, “that’s a guard, what about a bigger player?” wings like David Thompson (below) and even bigs blew by him:

In the games I’ve sampled from the early ’80s, Erving’s defensive error rates are moderate, coming in at about 1.5 per 100. However, his earlier year rates appear higher, bordering on problematic. Although, all players make errors, and these habits were largely offset by Erving’s strengths.

He maintained his athleticism into the mid ’80s before starting to slip, chipping away at his scoring game and reducing his defensive effectiveness in his final seasons before retiring in 1987.

Impact Evaluation

Erving played his first five seasons in the ABA, which, despite its lack of historical prominence, was what the AFL was to the NFL. At the time, the NBA’s marketing efforts tried to depreciate the ABA, branding it as a defense-free, lesser alternative. However, the ABA continued to pilfer talent from the NBA and as the years went on (and the NBA rapidly expanded, despite losing so much talent), the leagues grew comparable in quality. Here’s an attempt to quantify this by Mike Goodman at APBR:

In the last few years of its existence, the talent gap between the leagues was small, although there were some differences that impacted Erving. While it was known as a no-defense league, many of the best defenders, like Artis Gilmore and Bobby Jones, played in the ABA, and the top defensive team the year of the merger was an ABA import, the Nuggets. But ABA offensive ratings were higher because the rules of the league made offense easier.

Much like the relaxed enforcement of palming and improved spacing helped improve NBA efficiency in the ’70s, the ABA’s 3-point line and skilled dribblers made defending a harder task. ABA turnover rates were significantly lower than the NBA’s, as its overall efficiency, not surprisingly, mirrored typical NBA seasons from the 3-point era. Erving himself likely benefited from the more spread out, free-flowing game.

In his 1972 rookie season, Erving’s Virginia Squires were a .500 team, average on offense and defense.2 In 1973, Erving shifted into his prime and his scoring rate and efficiency spiked to a level that he would maintain throughout his ABA career. When compared to NBA stars, there were only about 30 player-seasons before 1982 that were comparable to Erving’s best scoring-efficiency combination. He then regressed in his first few NBA years:

Erving’s turnovers also increased in the NBA, and for many years were in the 30th or 40th percentile for scorers of his ilk (i.e. a 20 point and 4 assist per game player). We’d expect a player like this — efficient volume scoring and moderate creation — to leave some footprint on the offensive end. However, given his lack of outside shooting and his mediocre passing attack, we’d also expect that signal to be stronger on weaker teams.3 Which seems to be what happened.

Perhaps the first good glimpse of his impact can be found in 1974 when he moved to New York. The Nets jumped to the top of the league, although they massively upgraded the entire roster (with All-Star scorer Larry Kenon and John Williamson). New York actually won the title with the best defense in the league (and an average offense). In ’75, they again posted a slightly better offense than league average, but were second in defensive efficiency. So with skilled offensive players next to Erving, they produced a good-but-not-great offense, which makes what happened next such an interesting data point.

In 1976, Kenon, All-Star Billy (the Whopper) Paultz and Mike Gale all left New York. Erving upped his scoring rate by 8 percent and the Nets offense regressed…by all of 1.3 efficiency points. In the playoffs, Erving cranked the volume up to peak Wilt Chamberlain levels on improved efficiency en route to another title. His high-volume carry jobs yielded similar results to his more talented offensive teams, the classic ability of an isolation scorer to raise the floor of a struggling offense but not the ceiling of an adequate one.

As for Erving’s defensive impact, I’m mixed on the numbers. On one hand, his ABA defense is worse on film than his NBA years. However, he was at his athletic peak then and often played at the big forward position. As such, I land somewhere in the middle, giving him credit for his defensive rebounding, rim protection and impressive team results, assuming that against certain competition his propensity to go for steals yielded positive results.

Thanks to the wonderful work of Harvey Pollack and others, Erving’s transition into the NBA is like hopping into a time machine. Pollack tracked plus-minus data for Philadelphia, and as a result we can use Augmented Plus-Minus (AuPM) to evaluate his impact on those teams. Below I’ve plotted the percentiles of his AuPM values and their typical ranking in a given season:

As you can see, Erving’s results were subpar for a perceived superstar. He generated only two top-50 seasons (per this metric) and never showed up as a top-20 player in a given year. Whether it was shaky knees or an adjustment to the NBA — Philadelphia was criticized for its poor fit, with George McGinnis and Erving functionally redundant — the results in his first few 76er seasons were disappointing. Expectations in Philly were sky high, but the team ignominiously underachieved, as described by Sports Illustrated in 1977:

“Erving and McGinnis went together like cream cheese and scrapple; they could not get along, much less play alongside each other. Neither man could coexist with Free, who monopolized the ball and was known to start shooting before the concluding notes of the national anthem.”

Erving’s disappointing plus-minus and Philadelphia’s deflating results aren’t too surprising in retrospect: Julius was surrounded with redundant talent (McGinnis) and his offense wasn’t built to scale since he lacked outside shooting to exploit sagging defenses that were rarer in the more spacious ABA. Philadelphia still posted the best offense in the NBA in 1978 under new coach Billy Cunningham, a stellar 4.1 points better than league average. However, I see a loaded offensive roster and believe a transcendent star would have elevated them to greater heights.

Philadelphia shipped McGinnis out in 1979 and Doug Collins suffered a crippling injury. Philly played 31 full-strength games without Collins and didn’t miss a beat, performing at exactly the same 48-win pace (2.3 SRS), another testament to Erving’s floor-raising. In 1980, young Maurice Cheeks sprang to life and the Sixers rode the stifling presence of Bobby and Caldwell Jones to the league’s best defense (4.3 points better than average). However, they posted a slightly below-average offense. Again, Erving played on a middling attack that was elite defensively.4

1981 and ’82 were his brightest years (by narrative) in the NBA. The 76ers broke through to a 63-win pace when healthy (8.0 SRS) and Erving claimed the MVP. His estimated creation numbers went up with the presence of the 3-point line (which makes sense if it helped spacing improve), in turn making Julius harder to defend. His 1982 AuPM finally paints him as an All-NBA level player.5 Here’s an overview of his Philadelphia teams for those years:

Despite playing long before the Databall era, we have an amazing amount of information on Julius. In addition to AuPM, WOWYR suggests he had star-level impact, and that’s based on NBA data only. However, another game-level historical regression (GPM) casts doubt on that.6 His defensive footprint is confusing — he was given 1976 all-defense honors in the ABA, posting strong block and steal numbers, but he did this in prior seasons as well and wasn’t given a nod. Blocks and steals are memory-defining highlights, and as a result such players are historically awarded with all-defense because of it. But Erving wasn’t, despite logging major minutes on strong defensive clubs.

Still, I credit him with having decent positive impact on defense in his best seasons, although he’s far from the top perimeter defenders in history. It’s possible that in the years he displayed marginal impact, his defensive performance waned. I’m not sure that a fully engaged Erving could be a defensive superstar given his lack of foot speed, but again, some positive trend in that direction can make a huge difference.

Overall, Julius had outstanding longevity, especially for starting in the early ’70s. By my valuations, he tallied 12 consecutive All-NBA seasons with an MVP-level peak. I think the argument for placing him higher rests on larger defensive impact, and it’s hard for me to buy that too much. On the other hand, placing him lower means viewing him as a neutral-impact defender, and that seems unlikely too. Ultimately, the duration of his career lifts him into this cluster of players in the 12-16 range, however, his lack of an all-time peak, caused by limitations in his offensive game, land him at No 16.

Backpicks GOAT: #28 Rick Barry

Key Stats and Trends

  • Elite scoring rates but only moderate efficiency for era
  • Presence repeatedly correlated with moderate improvements in team’s offense

Scouting Report

Rick Barry was a gunslinger, pinpoint passer and defensive pest. He also shot free throws like your grandmother.

That unapologetic style made Barry a 90 percent career free thrower, and in his final eight years in the NBA, Barry converted at 91.4 percent from the line, which would top the all-time career mark of 90.5 percent, currently held by Mahmoud Abdul-Rauf.1

Barry launched merciless midrange shots over defenders, using his 6-foot-7 frame and high-release point to bludgeon opponents, one long 2-pointer after another.

While many of these pullups evoke Jerry West, Barry didn’t attack the hoop like The Logo. Upon returning to the NBA in 1973 (after four seasons in the ABA), Barry shot a free throw for every 4.5 field goals he attempted, a rate far below West’s, who was closer to a free throw for every 2 field goals attempted during the heart of his career.

Barry’s iconoclastic foul shots overstated his skill as a shooter. After all, if he were a deadeye assassin, why didn’t he use his normal stroke from the line? In his final two seasons in the ABA, Barry made 92 of 323 3-point attempts (29 percent). While deep bombs weren’t as common back then, the best triplers of the time still shot in the mid 30s, and Julius Erving, a career 26 percent 3-point shooter in the NBA, managed to hit 34 percent of his 233 attempts in his final three ABA seasons. So while Barry lived off his midrange pullup, its accuracy was reflected in his typical 2-point field goal percentages, which hovered between 44 and 46 percent.

But Barry’s best attribute was his passing, slinging darts to cutters like few forwards in NBA history. Below is a collection of Barry’s passing from a single game, in which he hit a cutter for a layup, created two jumpers by drawing defensive attention, displayed elite court awareness after a steal and flicked a ridiculous touch pass. It’s the portfolio of an elite passer:

He only missed one high-quality pass in that game, and so many of Barry’s dishes ended in layups or free throw attempts. The second pass in that video demonstrates the meta-game value of his midrange jump-shooting. While those shots lack the efficiency of Steph Curry 3-pointers, they occupy the defense and can be leveraged for creation, which is exactly what Barry did.

He didn’t ramp up his passing frequency until returning to the NBA, balancing his quick-trigger scoring attempts with improved distribution and creation. His fantastic motor — he stressed defenses with constant movement and back cuts — and elite free throw shooting brought Barry just above league-average in efficiency on high-volume scoring.

Defensively, he had amazing hands and a knack for jumping passing lanes. These tendencies helped him lead the league in steals in 1975. His size, above-average rebounding and strong court awareness can be seen in highlights from this 1975 Finals game:

Barry evolved as a player, entering the league in the late ’60s as a gunner before rounding out his game in the early ’70s while improving his defense, per Sports Illustrated in 1972:

“The Warriors, who remember Barry as a nonstop offensive player who always left the other end to Nate, have been surprised to find that he apparently learned some defense in the ABA.” (November, 1972)

He maintained a high-level of play until his final season in Oakland before winding down his career in two forgettable seasons with the Rockets.2

Impact Evaluation

Barry changed teams during his prime more than any other player on this list, largely due to legal disputes. He fled San Francisco for the ABA after the ’67 season but was forced to sit out a year for contractual reasons, then played on two ABA teams before returning to the Warriors in 1973. During his comings and goings, Barry left a compelling case as an impact player.

With Barry aboard in 1966, the Warriors offense improved 3.5 points per 100, from 5.9 points below league average (rORtg) to a more respectable -2.4 rORtg. The venerable Alex Hannum still commanded the team, and veteran holdovers from the Wilt era, like Guy Rodgers, Tom Meschery and Al Attles, rounded out a roster centered around Barry’s scoring and Nate Thurmond’s defense.

Most players exhibit natural growth in their sophomore campaign (or around Barry’s age that year, 23), and the ’67 Warriors indeed graduated to the next level. When healthy, San Francisco played at a 56-win pace (5.3 SRS) en route to the finals.3 Barry authored the fifth-highest scoring rate of the ’60s — ahead of all but Wilt’s three most prolific years and Elgin Baylor’s abbreviated 1962 season — averaging 24.8 points per 75 possessions on +3.8 percent scoring efficiency.4 The Warrior offense improved to around league-average.

Barry played for the ABA’s Oakland Oaks in the 1969 season, missing 43 games and giving us some insight into the value of his volume scoring. With Barry in the lineup, the Oaks ticked along at a dominant 67-win pace (10.4 SRS). Without him, they were a more mortal 56-win team (5.3 SRS). This is the kind of impact we see from superstars who make good teams great.

However, that stretch was not all roses for Barry’s data footprint. The Warriors remained strong in 1968 without him, led by Thurmond, the emergence of another scorer (Jeff Mullins) and newcomers Rudy LaRusso and Clyde Lee. In 1970 (with Oakland now in Washington), the Capitols were slightly better without Barry for 15 games (2.3 SRS compared to 0.3).5 And as impressive as the split-season result was in ’69, the early ABA wasn’t a strong league yet. Although, when Barry left the New York Nets in 1973, they collapsed from .500 to a 25-win pace (-5.8 SRS) with little roster movement.

Barry returned to the Bay in 1973, joined a nearly identical team from the year before and again spearheaded an offensive improvement. Golden State jumped from a -2.8 rORtg to -0.2 and played at a 50-win clip when healthy, a threshold they would eclipse for five consecutive years. Despite weaker scoring numbers than his first stint with the team, Barry’s passing and creating likely made a difference for the marginally talented Warriors.

In 1974, the Warriors produced the first of four straight laudable offenses, climbing to a 55-win pace (4.3 SRS) when healthy and a +3.1 rORtg for the season. The following year, they shuffled the roster, surrounding Barry with a crop of young talent, yet the results were a similar offensive rating and another 50-win pace before catching fire in the playoffs and winning the championship. Teams were evenly matched across the league that year, so Golden State’s performance was slightly more impressive than the raw numbers suggest.

In 1976, some of that young talent blossomed — specifically Phil Smith and new acquisition Gus Williams — and Golden State played at a 58-win pace (6.2 SRS), Barry’s best team. It wasn’t until ’78 that the offense would dip below average again and the Warriors returned to .500, a decline that coincided with Barry’s drop-off as a player due to age.6

In the mid ’70s, Barry’s repertoire made him a premier creator, capable of bringing offenses to moderately strong heights. His multiple correlations with moderate offensive advancements are consistent with a high-volume gunner whose scoring style doesn’t scale too well but whose elite passing does. Regressed game-by-game results portray Barry as a star, although there is some variability in his data, likely caused by his lack of a consistent 10-year stretch in the NBA. Otherwise, the aggregate of his missed time (WOWY) and his overall statistical footprint reflect a general positive trend that support the case-by-case points examined above.

All told, I give him 11 All-Star level seasons and a weak-MVP peak, helping Barry pass a number of lesser-peak players in the 15 spots behind him. Depending on how much credit one gives his defense, he could move down a few slots, although it’s hard to see him leapfrogging any of the players ahead of him. He left two prime seasons on the table with his choice to sit out the 1968 season and his left knee injury in 1969, but still amassed sterling longevity for a player who (amazingly) debuted in the fall of 1965 and ended his career in the Bird and Magic era.


Backpicks GOAT: #12 Oscar Robertson

Key stats and trends

  • Arguably the strongest statistical footprint before the Databall era
  • Anchored elite offenses for entire career while playing on subpar defenses
  • Excellent combination of scoring, efficiency and playmaking

Scouting Report

Oscar was the first ball-dominant quarterback, his offensive game a fusion of Chris Paul and Dirk Nowitzki. He liked control, was deliberate and didn’t try many aggressive, high-risk plays. His bread-and-butter was the ability to score (seemingly) at will in the midrange with his pull-up, often executing it going to his right like this:

He had a size advantage over opposing guards, but he used his hips and shoulders to create space and find an open release point. This even worked against larger defenders like John Havlicek (shown above). Here he comes back to the left with pump fakes on the same move:

And he could do it going left. Notice the high release point as he (somehow) scores this over Russell.

The “Dirk” part of his game was the slow, deliberate artistry with which he could carve out his own shot in the mid and high-post area. This is some of the best old-man game you will ever see — he sticks his backside into his defender and probes for a look.

Much of Oscar’s effectiveness stemmed from this kind of isolation. What made him an all-time offensive talent was that he would also pass out of these situations as well, making the “right” play when he spotted mismatches. Below, he recognizes the switch on the pick-and-roll and feeds big-man Connie Dierking for an easy score.

Oscar made plays like this regularly and was quick to exploit these kinds of opportunities. He was also stellar in pick-and-roll action in general, particular for the time. Below, he slips a great pocket pass to Wayne Embry the second he feels Embry clear both defenders:

As the years wore on, spacing and pick-and-roll action advanced league-wide. By the early ’70s offenses showed more fluidity and, in ’71, teams like Milwaukee ran more modern action. 1 A better-spaced court weaponized a quarterback like Oscar, opening up his driving and passing lanes, allowing him to create easy shots for teammates:

These kinds of players were less common back then, but my bet is that Oscar set up his teammates more than any other player in the ’60s. Perhaps his riskiest (and some of his best) passes were in transition, where he always seemed to know who was going to be open before they were open.

He had great vision, but lacked some of the same aggression as the all-time great passers. This is subtle, but it’s the difference between a great offense a transcendent one.

Windows like these are small, and exploiting them requires hutzpah, but the few times they materialized on film Oscar didn’t go after them. Ironically, he would occasionally toss a lazy outlet pass for a turnover.

Defensively, he appears fairly neutral from the available games we have. When he’s engaged, he moves his feet fairly well and uses his size and good hands. Here he is disrupting great offensive players late in his career:

However, he did have sub-optimal tendencies, like floating out of position occasionally. In the first clip below (guarding the inbounder), he sort of overreacts to the cutter, and in the second, demonstrates some curious pick-and-roll defense.

Finally, he gambled for steals a number of times. It’s unclear whether this was strategy or him going rogue, but it certainly played into the identity of the ’60s Royals, who would score efficiently but also were easy to score on.

Oscar aged well, as we’d expect from someone with Paul-like control and Nowitzki’s YMCA tricks. His physical condition started to fade in 1972, when he broke down during the playoffs with a “deep muscle pull in his stomach.”  In ’73, he labored through problems with his toe, neck, shoulder and hamstring before retiring in 1974.

Impact Evaluation

Management and team construction can saddle even the greatest of players, and Oscar’s Royals were the original exemplar of that. When Robertson arrived in 1961, the Royals offense immediately spiked. The ’60 team had finished dead last defensively and middle of the pack offensively, resulting in a 24-win pace (-6 SRS). With Oscar aboard, the defense remained porous but the offense jumped to best in the league.

The Royals were an undersized team filled with solid offensive talent, and Oscar’s passing and command catapulted them forward. But during those years, Cincinnati’s lineups lacked size, rarely featuring big men over 6-foot-8, and the ’64 and ’65 clubs were the only teams of the decade not to have a player taller than that log a single minute! Below is a year-by-year plot of the number of minutes per game occupied by 6-foot-9 players or taller on teams around the league:

The ’60 Royals were young, with no one over the age of 26. They were led by scorer Jack Twyman (31.2 ppg, +2.3 percent relative true shooting, or rTS) and a platoon of supporting players who averaged under 28 minutes per game. But they were a solid offensive group, and alongside Oscar, Twyman’s efficiency shot up in 1961 (+6.4 percent rTS). Robertson made an immediate impact, elevating Cincinnati from +0.2 (efficiency relative to league average or rORtg) on offense to +3.5, a number they would eclipse in all but two years of the 1960s.

Oscar missed nine games that year as a rookie, and the Royals played disastrously without him, falling from a 36-win pace to a 9-win pace. The same team, along with Adrian Smith, came back a year older in ’62, and the offense inched-up to elite (+4.7 rOrtg) while the defense improved slightly. When healthy, the ’62 Royals played like a 45-win team. They repeated the pattern in 1963 — a 46-win pace, the league’s best offense (+3.5) and one of its worst defenses. By ’63, their tallest player was Hub Reed at 6-foot-9 (16 mpg), and their starting bigs were Bob Boozer (6-foot-8) and Wayne Embry (6-foot-8). Embry was long and built like a tank, but he wasn’t a rim protector nor a disruptive team defender.

1964 was Cincinnati’s year. They finished first in offense (+4.3) and balanced it with an average D. Twyman missed 12 games —  the Royals dropped to a 40-win pace without him — and with him played at a noteworthy 55-win clip. Oscar claimed the MVP, and if Bill Russell decided to play baseball that year, the Royals would have been strong title contenders. The addition of rookie Jerry Lucas (a stretch big) likely helped; despite defensive shortcomings, Lucas was an excellent rebounder and cleared possessions next to Embry. He added efficient scoring, averaging 17.7 points per game on an awesome +9.3 percent rTS.

In ’65, despite returning the same young core, the Royals fall back to their familiar 46-win pace (1.8 SRS). Lucas missed time that year, but Cincy was actually better without him.2 This was a pattern with Lucas, who posted commendable stats but seemed to barely move the needle; he has one of the worst WOWY scores on record (-1.3) and WOWYR finds him similarly ineffective (ranking 559th as a neutral-impacter player). Odds are, his defensive deficiencies limited his value, and while the Royals maintained the same offensive heights as the ’64 team, they regressed defensively in ’65.

Up until that point, Cincinnati had the best offense of the decade, consistently finishing at the top of heap every season. In ’66, like clockwork, they were again a 46-win team, although this time the offense dropped a few points and the defense picked up the slack. (They had the same top-6 in their rotation, save for Twyman, who was essentially replaced by Happy Hairston.) The following year was a rerun of ’66 but the team was slightly worse (42-win pace). As SI put it, they “developed some sloppy habits, especially on defense.” In other words, Oscar was in basketball purgatory.

Robertson missed 10 games in 1968 before Hairston was traded, and Cincinnati collapsed without him, cascading from a 46-win pace (+1.8) down to a 17-win one (-9.3) in his absence.3 In 1970, he sat for 12 games, and an otherwise healthy Royals dropped from a 42-win pace with him (+0.3 SRS) to an 18-win pace without him (-8.7 SRS). Oscar’s missed time came in small samples, but the results hinted at his extreme value.

Unlike the modern ball-dominant quarterbacks, Oscar wasn’t spearheading attacks by relentlessly creating opportunities for teammates — such plays weren’t common for much of the 1960s. Oscar led the league in assists in most years, but even then assist rates were far below what they would become after the merger in 1977.4 Oscar’s assists per 75 possessions were regularly between 5.9 and 7.5. For comparison, John Stockton has the highest rate ever at 13.6, while Magic and Steve Nash peaked around 12. But the best mark before the merger was Kevin Porter’s 8.5.

Thus, Oscar wasn’t making life way easier for his teammates the way creators like Nash and LeBron did. Instead, he was a great facilitator. His more conservative passes put players in the right position to score. He could find easy offense in transition and his great feel for mismatches helped team efficiency too. But an enormous chunk of his global impact came from his own isolation scoring, which was orders better than anyone that decade not named Jerry West:

Oscar took about 27 scoring attempts (TSA) per game for many years in Cincinnati, and the difference between average efficiency and Oscar’s +8.5 percent rTS (his ’60s average) was roughly four points per game for his team’s net efficiency. Coincidentally, that mirrors the Royals offensive advantage over the league for much of the decade. It’s never that simple — the game is far more interactive — but it provides perspective on how valuable that kind of efficiency can be while taking less than 25 percent of a team’s scoring attempts. And while West’s scoring was even better, he lagged behind Oscar as a playmaker; West peaked at 5.7 assists per 75 during those years, with a number of seasons in the 4s.5

In 1970, Bob Cousy replaced longtime coach Jack McMahon and the wheels started to fall off. Lucas was traded at the start of the season for Jim King and (bootstrap) Bill Turner. Cousy clashed with Robertson’s style, wanting to reduce his ball-dominance and up the tempo, and at one point Cousy even made an ill-fated comeback attempt that lasted 34 minutes over seven games. Their offensive rating dropped almost 6 points (to -1.0) but the defense improved by 4 from the previous year, to 1.4 points better than average. Shifts like this are often a reflection of strategic changes, where neither side of the ball changed that much, but instead lineups and transition tactics shifted. Cousy wanted to fast break, no doubt an attempt to rekindle the Celtic glory years, and that might have led to hurried possessions. The loss of Lucas may have also played a role in the offense’s decline and the defense’s improvement.

Mercilessly, in 1971, Oscar was traded to the Bucks for peanuts: Charlie Paulk (18 mpg in Cincy) and Flynn Robinson (34 mpg in Milwaukee and then 19 mpg in Cincy). Without Oscar, the Royals played like a 33-win team (-3.0 SRS) in ’71. In Milwaukee, the Bucks strung together one of the most dominant seasons in NBA history en route to a title. They would maintain elite status until his final season in 1974.

Like West, Oscar leaves an impressive statistical footprint. He doesn’t miss as many games, but when he did miss time Cincinnati fell ill. His WOWY sample sizes aren’t too large — 69 missed games during his prime — but regressed game-level data reinforces that he was a huge-impact player. Corroboratively, they suggest elite value and a strong MVP-level peak.

I’m not overly excited about the scalability of most ball-dominant players. They certainly can scale — Oscar, after all, played on a dominant team — but it’s hard for others to make life easier for them. For instance, when Oscar left Cincinnati, his efficiency and scoring declined, despite playing alongside Kareem and the loaded ’71 Bucks.6 Kareem created easy shots for the shooters around him, but not many easy ones for Oscar because he wasn’t a cutter or spot-up shooter.

Overall, Oscar has the portfolio of an all-time great offensive star, but his defensive impact is a question mark. His team defenses were porous — his presence never correlated with much on that end — yet numerous accounts praised his ability to bother opponents with his size in Milwaukee. Without video, it’s difficult to hone in on this area; I hedge my bets and view him as average during his prime.

His health was good, for the most part, although he lost value in the ’72 playoffs to injury. Otherwise, he was an MVP-level candidate for his first 11 years in the league and the original offensive quarterback. If I curb his peak slightly — a reasonable stance — he falls back two spots to No. 14. On the other hand, I can’t see the evidence to boost him into the first 11. And while the group of legends between Oscar and West are all within an MVP season of each other, Robertson edges them out for the 12th-most valuable career in NBA history.


Backpicks GOAT: #17 Jerry West

Key Stats and Trends

  • One of the largest statistical footprints before the Databall era
  • Historical combination of scoring volume and efficiency
  • Led some of the greatest offenses ever before the 3-point line

Scouting Report

Jerry West was the original perimeter vortex, pummeling defenses with futuristic scoring and adroit passing. His hair-trigger release and long arms — he was said to have a 6-foot-9 wingspan — allowed him to create his own shot from nearly anywhere. Here’s a quintessential West pull-up:

Along with his quick pull-up, West employed a more deliberate jumper, using his body to create space, then hitting defenders with head fakes before launching over them:

When evaluating any ’60s guard it’s important to remember that dribbling rules were enforced quite differently then, and players could barely turn their wrists without being whistled for a palming violation. Thus, guards like West dribbled closer to the floor and lacked the arrhythmic cadence of modern crossovers. Still, West could drive and finish well around the rim with a wide range of shots. He described his newly developed handle for the 1962 season:

“I can do a lot more with the ball, too. I was strictly a right-handed shot and I didn’t drive much, so the defense was playing me a whole step to the right and in tight. Now I can go to my left and shoot with my left hand, and I’m driving a lot.” (November 20, 1961 in Sports Illustrated.)

Incorporating these drives nearly doubled West’s free throw attempts in ’62, a key pillar of his efficient attack. Below, he goes to the bank on a spin move, flips a finger roll over the venerable Bill Russell and scoops in a hoop after buckling his defender with an inside-out dribble:

Based on the available film, these were typical West attacks. He was also a noteworthy passer with good court awareness. In the first clip below, he showcases his vision by dropping a perfect dime in transition. In the second, he displays a sound ability to hit open men on the break.

He had a good feel for pick-and-roll action in the half court, although from the available footage this wasn’t used as much in the earlier part of his career:

In the next clip, he creates offense for Wilt Chamberlain by drawing a help defender and slipping him the ball for an easy finish.1

West’s assists jumped at the end of his career during a league-wide trend, peaking at 8.1 per 75 possessions, slightly higher than Oscar Robertson’s best season.

All told, West produced 25 to 32 points per night on extraordinarily high efficiency for the times. His ability to draw fouls resulted in 11 consecutive seasons at 8 free throw attempts per game or better. As a career 81 percent free thrower, this carried him to the top of the league in efficiency twice (as a volume scorer!), and from ’64-71 he was 7.7 percent above league average in true shooting percentage (rTS).

West was also a noteworthy defender, using quick hands and long arms to generate blocks and steals. He was often disruptive, slapping at balls, which led to plays like this:

And his wingspan helped him block a number of shots for a guard — in this vein, he’s similar in stature and “true height” to a modern player like Dwyane Wade, who was also an exceptional shot blocker as a guard. Below, he sends back a shot defending a two-on-one fast break and then another in the half court.

While West was good right away, it wasn’t until his third year in 1963 that he really hit his stride. He maintained his skills during the heart of his prime until the early ’70s, when he tapered off before retiring at the end of the 1974 season.

Impact Evaluation

West left one of the largest impact footprints in NBA history and comes away looking like an all-time great. It was West, not Elgin Baylor, who guided Los Angeles to a decade of offensive excellence, and it’s West who has a viable claim as the best offensive player before the 3-point era.

In 1959, the Lakers added Baylor, but were still a below-average team, finishing a with a -1.4 SRS (or a 37-win pace over an 82-game season). Larry Foust (seven-time all-star) and Vern Mikkelsen (six-time all-star) were aging holdovers from the Mikan years, and Foust was traded in 1960 while Mikkelsen retired. That ’60 team, the last in Minneapolis, finished with a -4 SRS and the worst offense in the league. So while Baylor racked up worthy stats (29 points and 4 assists per game), he wasn’t able to do much with spare parts.2

In 1961, with basically the same rotation back and a year older, the Lakers moved to LA and added a rookie West. Baylor took another step forward, but West wasn’t the assassin he would soon become, scoring 17.6 points per game on below league-average efficiency and shooting just 67 percent from the free throw line as a rookie. The Lakers were a .500 team again, and the offense was merely the second-worst in the league; Baylor averaged 35 points and 5 assists per contest. It marked the last time the Lakers would field a below-average offense until West’s final year in 1974.

Per the scouting report, West improved in ’62, although, based on his free throw accuracy, his shooting didn’t reach peak levels until 1964. Baylor missed most of the second half of the ’62 season (military service), and in his absence, the Lakers still played at a respectable 37-win pace. With Baylor, the Lakers played like a 55-win team and finished the year with a positive offensive rating (1.4 points better than league average, or rORtg). But it was West’s growth in the middle of the decade that coincided with a Laker offensive boom.

In ’62, West called his own number a good amount, but Baylor still took about 40 percent more scoring attempts than him, despite Elgin’s lesser efficiently. This is not a total indictment of Baylor — he was a fairly good passer himself and his attack in those years was effective, as evidenced by the team’s improvement in ’62 with him — but given the lack of cohesive offensive structures at the time, it is a red flag that Elgin was eating up too many possessions for himself. More on this in a moment.

Below is a plot of scoring volume (x-axis) and efficiency (y-axis) for West and Baylor; West’s combination of volume and efficiency was unmatched for the period:

In 1963, West missed his first major chunk of time, portending an injury-riddled career. The Lakers played at a 55-win pace with West that year (4.9 SRS), but dropped to a 35-win pace in 26 games without him (-2.2 SRS). The Lakers were so hot at one point that Sports Illustrated called them one of the greatest teams ever before West’s injury:

“On February 3, 1963 one of the best basketball teams ever assembled stopped being that. It was on that day that Laker All-Star Jerry West pulled a muscle, and the team that had just won 42 of 50 was thenceforth to be no more than a win-one, lose-one powderpuff in powder blue. It is going to be a lot of seasons before anybody wins 42 of 50 in the NBA again.” (October, 1963)

In ’64, LA returned a nearly identical rotation, West’s efficiency jumped to peak levels, but they only played at a 47-win pace when healthy. The offense was good again, but the defense completely dropped off. Notably, this was the year in which Elgin Baylor’s knee problems began, yet his scoring attempts were still nearly identical to West’s. Baylor’s rebounding dropped off, perhaps a reflection of lesser athleticism.3 The following year (1965) was a near copy of the results, and Baylor suffered a major knee injury that postseason.

In ’66, the Lakers finished first in offensive efficiency (+3.4 rOrtg) as Baylor took on a secondary role post-injury. In 12 games without Elgin that year, LA played exactly as they did with him — a 47-win pace. West’s bump in assists while maintaining his scoring volume indicate an increased offensive load. After a difficult ’67 season for the team (West missed the playoffs) in which Baylor’s scoring attempts shot back up (despite subpar efficiency) the plateauing Lakers moved on from longtime coach Fred Schaus and brought in Butch Van Breda Kolff.

The geometry of the NBA was different in the ’60s. Fluid ball-movement and spacing were non-existent in most half-court sets. The area around the hoop was clogged like the pile up in front of a hockey net. As a result, wings who were good enough to drive couldn’t easily maneuver to the hoop, often met by a wall of players cluttered in their path. Defenses sagged back in the lane and willingly surrendered outside shots that were still only worth 2 points.

But in 1968, Butch Van Bredda Kolff implemented a Princeton-based system in LA, which naturally emphasized spacing and a clear lane. The limited footage reveals a stark contrast between other years, where the Lakers offense had space to operate, was more fluid, and as a result their dominant wings could drive to the basket more frequently. In short, it looked more modern.

This made West particularly deadly, as he had by far his best season from the field, shooting 51.4 percent from the floor. His true shooting was the highest in the first 22 years of the NBA for a guard, nearly 10 percent above league average.4 The improved spacing amplified both West’s ability to score and to create for his teammates, and with West the Lakers played at a 62-win pace. In 27 games without him, they regressed all the way down to .500 ball. Despite his missed games, they still finished with the highest offensive rating ever posted at the time (101.7). Here’s how West’s ’68 Lakers compare to the top teams before the 3-point line by relative shooting efficiency (rTS):5

With West in the lineup, the ’68 Lakers were the first offense in NBA history to hit +4 percent rTS and the second-best relative offense before the 3-point era. The 1967 76ers were the only other team of the ’60s to even eclipse +3 percent rTS (3 percent equates to about 6 extra points every 100 possessions.) LA’s raw shooting efficiency was only topped by nine teams in the following decade, despite a large uptick in scoring at the start of the ’70s. Van Breda Kolff’s schematic shift was nearly to the ’60s what Mike D’Antoni’s Seven Seconds or Less was to the aughts.

But the Laker chemistry was disrupted by the arrival of Wilt in 1969.6 There’s a play in the scouting report above from 1970, where West goes to drive but Wilt’s defender and Wilt occupy the left block, clogging his path. Any offense that spread the court, opened the lane and asked bigs to stay above the free throw line was not conducive to Wilt’s tendencies (or strengths), and despite the promise of being the original super team, the Lakers were worse in 1969. This is the first major example in NBA history where a lack of portability rendered a contending team less than the sum of its parts. And the coach was scapegoated for it.

However, West was still additive. In 20 games without him, LA was a .500 team, but with him, they played at a 57-win pace (still worse than the healthy ’68 squad). In 1970, after the Happy Hairston trade7, the Lakers played 32 games with West and without Wilt, and in those games, LA played at a 54-win pace (4.7 SRS). With Wilt in the lineup, the Lakers were again slightly worse. West and company even logged 16 of those 32 Wilt-less games without Baylor, maintaining a nearly identical pace without Elgin too.

In 1971, the last year of West’s stretch of elite efficiency, LA played like a 55-win team with him (5.0 SRS) and a 37-win team without him (-1.0 SRS) for 18 games. Without West in the playoffs, the Lakers scored at 0.98 points per scoring attempt (PPA, or true shooting times two), down from 1.04 in the regular season, hinting that West’s presence was felt almost entirely on offense. Complete shooting records are spotty during West’s era, but there are two large available sets from his WOWY career, in 1963 (19 of his 26 missed games) and 1968 (24 of his 27 missed games) that demonstrate his massive offensive value. Below I’ve plotted those two teams with scoring efficiency on the x-axis (TS) and the change in efficiency with West in the lineup on the y-axis (PPA):

In both cases, the Lakers were drastically improved with West. The ’63 team posted a 50.5 percent true shooting mark with West in those 19 games we can access, up from a slightly below-average 48.5 percent in all other games. The ’68 team improved by 3.4 percent in rTS (6.8 PPA), the equivalent of taking an average offense to the top of the league.8 LA’s bigs experienced the largest shift with West in ’63, but in ’68 he made life easier for just about everyone. Also, we expect West to have a smaller influence on Baylor precisely because Elgin self-created well and wasn’t a good outside shooter who could capitalize on West stressing the defense.

West exited his prime on two dominant teams — the ’72 champion and the ’73 runner-up. Both clubs played at a 69-win pace when healthy, although the ’73 team was closer to .500 in 12 games without West.9 Here’s a summary of West’s teams throughout his career, with and without him (WOWY):

Overall, West posts one of the highest WOWY scores on record, and regressed data supports that whether at the game level or with WOWYR. In most of those studies, he’s a hair behind Oscar Robertson, however I give West a slight boost in portability, as he achieved his results alongside multiple stars, whereas Oscar was always ball-dominant (despite jelling with Kareem in Milwaukee). West was also one of the few superstars ever to improve his scoring in the postseason.10

It’s hard to make an argument for his defense making too much of a dent, although I give him solid marks; West was a perennial all-league defender, reflecting some of the skills he shows on tape. Based on the totality of the evidence and his remarkable offensive apex, I think he flirted with an all-time (top-15) peak. His longevity and sustained prime are good for his era, although West’s injuries robbed him of two prime postseasons (and his final playoffs in 1974). Without those lost years, West would likely be 10th on this list. Instead, he barely edges out No. 18 for this spot.

Backpicks GOAT: #30 Bob Pettit

Key Stats and Trends

  • Never played on a dominant team
  • Despite strong box stats, limited evidence for elite peak

Scouting Report

There’s almost no video of Bob Pettit – the closest thing we have to a continuous reel of game tape is probably the 1962 All-Star game – so this will be the briefest scouting report in this series. It’s clear from the limited evidence that Pettit was a fluid athlete who had a good first step and an effective outside jumper. (He hit two shots near 3-point range in the first half of that ’62 ASG.) He could drive and finish around the hoop, was an active offensive rebounder and seemed to constantly probe for better position off the ball. Pettit himself felt his offensive rebounding was his best attribute, discussed below in this wonderful video on his career:

In the limited archives, there aren’t many instances of Pettit finding a great pass. However, there are some clips of decent assists or outright creation, setting up teammates after drawing defensive attention. Combined with his typical assist per game figures (often in the 3s) it’s likely that Pettit was a moderate creator for his time.

On film, his defense looks like a mixed bag. He occasionally reached when guarding the ball, but otherwise constantly swiveled his head to check his positioning. His recovery and shot-blocking don’t pop in any available footage, and he wasn’t known for verticality. However, it appears he was a strong defensive rebounder, but not quite elite in that realm.

Using estimates of rebounding, it’s likely that he was around 17 percent in total rebounding rate during his best seasons, comparable to modern bigs like Anthony Davis or Pau Gasol. In the first five seasons rebounding percentage were officially tallied — when defensive rebounding rates were chronologically closest to the ’60s — Pettit’s numbers would have ranked about 10th in a given season, or around the 80th percentile among big men.

As his career unfolded, Pettit’s physical condition changed dramatically. According to his account, he was a slender 210 pounds when he entered the league. After taking punishment in the paint, including 140 career stitches in his face and a broken hand that forced him to wear a cast at times in 1957 and ’58, he added 35 pounds with weight training, bulking to 245 pounds (at 6-feet-8 in socks). Pettit retired at 32, tearing a ligament in his knee in his final season in 1965.

Impact Evaluation

The shot-clock was to the NBA what the Cambrian explosion was to biology. Before Danny Biasone’s timekeeping innovation, the league was in a dull place, contracting a team in 1953 (Indianapolis) before another disbanded in 1954 (Baltimore). In 1951, there was even a 19-18 game. With the clock’s implementation in the 1955 season, the league entered a period of exponential growth in which racial barriers eroded, rules evolved and money poured in, all of which attracted a larger talent pool. The game grew so fast (pun intended) that there were conversations about banning tall players.

One measurement of this growth is the prominence of new players, and as you can see below, an influx of rookies played larger roles at the outset of this period:

In the last 65 years, there have been only five seasons where rookies topped 13 percent of the 1500-minute players, and all five were between 1955 and 1963. The league was immature then, and the teams tightly packed; the hardest period in history to create any separation was in the late ’50s and early ’60s. So while parity prevented a dominant team until the Celtics empire, some of those 50-win teams were quite impressive.

Pettit entered the league in ’55 and immediately assumed a leading role, nearly doubling his second-best teammate in scoring. Despite frequent roster turnover and coaching turmoil during his first few years, the Hawks gradually improved, climbing from an also-ran to a .500 team, adding notables like Slater Martin and Ed Macauley. And a .500 team was good enough to win back then, as St. Louis took the ’58 title with a quotidian SRS of 0.8.

Pettit was the first great scorer of the shot-clock era and claimed two scoring titles in the ’50s. Thanks to his outside touch (visible on film), his efficiency was bested by only a handful of players during the post shot-clock explosion. Here’s how he stacked up in the first 15 years of the clock:

The Hawks peaked in 1959, playing at a 50-win pace (prorated to an 82-game schedule). Macauley moved to coaching and All-Star center Clyde Lovellette joined the team. More importantly, Pettit, free of his cast, spiked in scoring and efficiency while his assists ticked back up. Commensurate with Pettit’s individual improvement, the St. Louis attack finished first in the league in relative offensive rating in ’59. After two average seasons of offense, they posted +2.9 rORtg in ’59, a near identical number to their 1960 mark of +3.0. So while the defense remained steady, the offense turned them into potential challengers to Boston in those years.1

With rookie and future Hall-of-Famer Lenny Wilkens aboard in 1961, the Hawks produced another 50-win pace season. But the ’62 team fell apart, despite Pettit and Hagan logging big minutes. The defense betrayed St. Louis, dropping from well above average to well below it, losing 7.4 points in relative efficiency overnight.2 Lovellette was injured for half of the season, but the team wasn’t so hot with him either. Wilkens also missed most of the year for military service, and in the 20 games he played, St. Louis looked average (+0.6 SRS). Another key factor, along with any regression from aging, was St. Louis’s coaching carousel; the Hawks trotted out three coaches that season, including Pettit himself for the final six games! (He was the eighth Hawks coach in six seasons.)

After that, St. Louis strung together a few more runs behind Pettit (the player), Zelmo Beaty and Wilkens, playing at a 45 to 49 win pace for Pettit’s final three seasons while returning to defensive performances that were comparable to their pre-’62 numbers.

Unfortunately, we have limited data from those years to gauge Pettit’s impact. If we examine his missed time, his WOWY score in 35 missed games during his prime is unimpressive (+0.9), although some scaling of those numbers is required given how tightly compacted the league was then. Using a more robust method like WOWYR demonstrates decent positive impact, but his numbers are closer to Sam Jones than the giants of the era. Given his injuries, It’s likely these studies understate his peak play, although I do think they accurately reflect a lack of dominance compared to that period’s transcendent stars.

I could easily see Pettit a slot or two lower on this list. However, it’s harder for me to see him much higher. This is largely due to a lack of information and rapid change during the era; Pettit is really the earliest star of the shot-clock period, and because of that, some curving is required to account for the influx of talent that would hit the league in the ’60s. Still, I give him nine All-NBA type seasons with a peak that barely touched MVP status, good enough for the 30th most valuable career since 1955.


The Backpicks GOAT: The 40 Best Careers in NBA History

Welcome to the Backpicks GOAT, a list seven years in the making! You may have seen ESPN, Slam, Elliot Kalb and Bill Simmons take a crack at the top basketball players ever. Maybe you have your own list of the NBA greats. Or maybe you just like reading lists. Either way, this particular one is a little different.

This is less about The List and more about the exercise of player evaluation. It’s intended to be an historical reference, organized by player, that (hopefully) adds to the understanding and appreciation of players, coaches and teams over the years. If you like videos, charts and graphs, you’ve come to the right list.

What This List Is Not

This list will not make traditional “arguments” for players. I won’t attempt to balance Kobe’s championships without Shaq, nor do I care about accolades like All-Star teams or the number of Hall of Fame teammates someone played with. I also don’t care how many rings a player won; the very thing I’m trying to tease out is who provided the most lift. Sometimes that lift is good enough to win, sometimes it’s not.

There are no time machines either — it’s not about how players would do today if transported into the past or future. It’s about the impact each had in his own time over the course of a career.

What This List Is

This list also goes far beyond the box score — indeed, the box score is merely a reference for quantifying tendencies — so if you’re used to citing points per game and Win Shares, this will be a bit different.

Instead, this is a career-value, or CORP list; it ranks the players who have provided the largest increase in the odds of a team winning championships over the course of their careers. This means that having great Finals moments or winning the hearts of fans with innovative passes is irrelevant. You can make a great list with those criteria, but that’s not what this exercise is intended to be.

This list is really about evaluating players based on “goodness,” not merely situational value. (If David Robinson backed up the two best centers ever, he wouldn’t be very valuable, but he’d still be very good.) Players do not earn credit for potential — Michael Jordan helped no one in 1994.

All told, in the last seven years I’ve evaluated over 1,500 player-seasons to compile this list.

Thinking Basketball

As you read player profiles, you will notice little mention of playoff performances or game-winning shots. That’s because sample-sizes are incredibly small; instead, playoffs are included as part of an entire evaluation. I’ll only call out the playoffs if they reflect something larger about a player. If you’re struck by the lack of discussion around clutch play or why “losing” players are ranked highly, all of these topics and more are explained in detail in Thinking Basketball. The book also examines critical components of team building (portability) and individual scoring that are foundational to these rankings. (Buying the book also supports the blog and is greatly appreciated!)

List Criteria

The first step is to evaluate a player season. My practice starts with film study in order to understand context.  Perhaps the most beautiful thing about basketball is that there are so many ways to skin the proverbial cat; 20 points per game for one player is not the same as 20 for another. Of course, some skills are more valuable than others. Here’s a guide to the major ones:

On defense, quality of rotations, court coverage, rim protection and length are all countermeasures to the above offensive criteria. (Rebounding counts too, separately for offense and defense.) I tracked these, shot selection, and passing habits in over 100 hours of video study specifically for this series. (To avoid winning bias, I watched segments of games from random quarters.)

After establishing the skill set and tendencies of a player (“Scouting Report”), I then leverage data to quantify the effect of these tendencies (“Impact Evaluation”). All of this ultimately leads to a numerical valuation that allows me to compare the impact of different seasons. The high-level criteria for determining “best career of all-time:”

  1. Evaluate how much a player impacts different lineups (Global offense and defense)
  2. Calculate the probability change in championships based on his health
  3. Add all his seasons together to determine CORP
  4. Adjust for longevity based on era
  5. Compare who has the highest impact

While the first step is my assessment of a player’s seasons, the next four steps are an attempt at an objective measure of career value using those assessments. To do this, I rely on a championship odds calculator I’ve developed over the years so I don’t have to worry about arbitrarily balancing “longevity” and “peak.” I then make an adjustment for era-based longevity, and typically sort out any close calls by defaulting to the player with the better peak or stronger era.

To simplify things, each player-season can be slotted into different tiers:

  • GOAT Season (30 percent or more chance of a title on a random team, or about +7 points per game on an average team)
  • All-Time Season (23-30 percent or +6)
  • MVP Season (17-23 percent or +5)
  • Weak MVP Season (12-17 percent or +4)
  • All-NBA Season (8-12 percent or +2.5)
  • All-Star Season (5-8 percent or +1)
  • Strong Role Player (3-5 percent or 0)
  • Role Player (1-3 percent or -2 to -0.5)

This allows for easy comparisons between multiple seasons; we can see if two MVP-level Bill Walton seasons are more valuable than, say, five All-NBA seasons from John Stockton.

Ranges, Not Absolutes

This is still only one person’s opinion. A “better” list would come from a group of diverse and highly knowledgable evaluators, like realgm’s top 100 list. I see my value here as a video and data curator and as an analyst of that data; obviously, mileage may vary on the rankings, especially depending on criteria.

With that said, I will try and highlight where there’s wiggle room and the ranges that I believe players fall into, but the final order is based on the most likely answers to me (i.e. gun to my head, how good I think a career was).

Stats Glossary

Throughout this list, I’ll use the following metrics regularly:

  • Efficiency (for individual players) – This is measured in true shooting percentage (TS), or occasionally points per scoring attempt (PPA). In the simplest terms, PPA estimates how many “attempts” were actually two-shot fouls, and takes the total number of points scored from 3-pointers, 2-pointers and free throws divided by attempts. True shooting divides PPA by two. In order to compare efficacy across years, this is almost always cited as relative to the league average (rTS). NB: Postseason rTS values are relative to the league (not the opponent) unless otherwise specified.
  • Efficiency (for teams)
    • Offense  This is an estimate of points scored per 100 possessions, or the team’s offensive efficiency. It is often cited as relative to the league average or “relative offensive rating” (rORtg). For the playoffs, rORtg is the difference between the team’s raw offensive rating and the opponent’s regular season defensive rating.
    • Defense – This is an estimate of points allowed per 100 possessions, or the team’s defensive efficiency. It is often cited as relative to the league average or “relative defensive rating” (rDRtg). For the playoffs, rDRtg is the difference between the team’s raw defensive rating and the opponent’s regular season offensive rating.
  • Creation – This is an estimate of how many shots a player created for his teammates per 100 possessions played. It’s also sometimes referred to as a percentage.
  • SRS – The “Simple Rating System,” it is a measurement of point differential for teams, adjusted for schedule strength. SRS is highly predictive of regular season wins and more predictive of games and playoff series than win percentage alone. For this series, a teams “win-pace” is based on its SRS.
  • The Big 3 / Big 4 – These are the three primary offensive dimensions of the advanced box score: Scoring rate (points per 75 possessions), efficiency (rTS) and creation. A fourth dimension — “The Big 4” — includes turnovers (modified for the presence of creation). “Scaled” graphics (sometimes titled “Normalized”) shrink each dimension on an axis of the same length for an equal comparison between them.
  • WOWY / APM – These are the non-box score, scoreboard-based family of plus-minus metrics and some of the most important measuring tools we have in basketball. Most of the references to these are summarized in the historical WOWYR series and this post on the historical compilation of plus-minus metrics.

Who Am I?

The Backpicks Top 40

The list will snake around a bit until the final eight players are revealed in order. The series is intended to be read in the order the profiles are released, which is noted next to each player. Players 31-40 are profiled in small blurbs, most players from 21-30 have limited video-based scouting reports, and all profiles in the top-20 feature full video-based scouting reports.

*Limited video-based scouting report 

  1. March 8
  2. March 8
  3. March 8
  4. March 8
  5. March 8
  6. March 12
  7. March 12
  8. March 12
  9. March 12
  10. March 12
  11. Bob Pettit* (2)
  12. Reggie Miller (10)
  13. Rick Barry(5)
  14. Patrick Ewing(11)
  15. March 5*
  16. John Stockton (12)
  17. Moses Malone(7)
  18. Scottie Pippen(13)
  19. Dwyane Wade(17)
  20. Chris Paul(18)
  21. Charles Barkley (14)
  22. Steve Nash (19)
  23. February 26
  24. Jerry West (3)
  25. Julius Erving (6)
  26. David Robinson (15)
  27. March 1
  28. Karl Malone (16)
  29. Oscar Robertson (4)
  30. Larry Bird (8)
  31. Magic Johnson (9)
  32. Wilt Chamberlain (1)
  33. March 19
  34. March 22
  35. March 25
  36. March 28
  37. April 2
  38. April 5
  39. April 9
  40. April 12