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
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.
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.
- Those players: Andrei Kirilenko (5x), Alvin Roberston (4x), Erving (4x), Michael Jordan (3x), Nate McMillan (3x), Scottie Pippen (2x) and Dudley Bradley (2x).
- Erving almost ended up in the NBA in 1973, playing the preseason with the Hawks as the Bucks owned his draft rights.
- For Thinking Basketball readers, think portability.
- Bobby Jones, who received first-place MVP votes in 1975 and 1976 in the ABA, was ostensibly the defensive star of those Sixer teams. Cheeks and Caldwell were also notable defenders.
- Removing McGinnis clearly wasn’t a panacea, as these improvements came two years after George left.
- WOWYR values him around +6 points per game, a tier behind the eight or 10 players who fare best in the metric. Conversely, GPM pegs him closer to a +3 player, a level below an MVP peak.