Key Stats and Trends
- Dominant box and non-box metrics due to versatile skills on offense/defense
- Strong team offenses in Minnesota, all-time level team defenses in Boston
- Elite court awareness on offense (passing) and defense (rotations)
Kevin Garnett was a video-game creation — a quick 7-footer with guard skills and the highest revving motor in NBA history. In his second year, he emerged as an All-Star, flying all over the court on defense while planting the seeds of his scoring arsenal, unveiling spins, fades and stepbacks:
In those early years, he was a lockdown perimeter defender, often assigned to unsuspecting guards who had to face some sort of supersized Sydney Moncrief incarnation that made finding clean looks a nightmare:1
His quickness and length shrunk the court, allowing him to help then recover, which played into one of Garnett’s greatest strengths, his defensive rebounding. He would rotate and roam constantly until a shot went up, then dive toward the rim and inhale the board:
Once he became a full-time power forward, Garnett’s relative defensive rebounding rates were among the all-time leaders, posting five consecutive seasons above the 96th percentile and leading the league twice in the category. His four-year peak of +15.8 percent ranks eighth since the stat was kept in 1974.
On offense, he used quickness to attack bigs, often facing up on the perimeter like a guard and driving. As Garnett moved into phase two of his career, his passing emerged, finding teammates off of penetration like this:
He developed into one of the greatest passing big men in NBA history, dolling out eye-popping interior dishes, touch passes, or point-center dimes from the pivot, as he demonstrates in the first clip below after forcing Michael Jordan’s doppelganger into a turnover:
In tracking over 1,000 of Garnett’s possessions from 1997-2009, his rate of quality passing was near John Stockton’s, and using just his games from 2003 to 2008, his rate of “good” passes was over 4 per 100, comparable to Jason Kidd. While Garnett’s on-ball load dipped in Boston, his passing was better than ever, making life easier for the offensive weapons around him. He was also the most prolific creator of any big man in history (if we don’t count Larry Bird). In tracking Garnett, his creation rates were similar to his estimated non-3 Box Creation marks — in the 7 per 100 range — placing him at the top of the heap since this stat became available in 1978.
Over the years, he developed a mid-post repertoire of fakes, spins and of course, his bread-and-butter fadeaway:
I tracked 151 Garnett on-ball scoring possessions in his sample, and he averaged a stellar 1.08 points per play on such attacks.2 It’s a small sample, but nearly identical to Tim Duncan’s 1.07 points per play in my tracking, mirroring Synergy’s “post up” similarities between both players.3
However, Garnett struggled to generate easy points for himself in the half court. While other greats like Duncan could use power to bang into traffic for free throws, Garnett relied more on quickness and finesse. His free throw attempts weren’t low, per se — between about 8 and 9 per 100 during his best offensive years — but Duncan was at 11 or 12. KG wasn’t quick enough to regularly turn the corner and absorb contact, yet his high center of gravity prevented him from rooting deep in the lane or dislodging defenders for easy put-backs. In that last clip, a stout Chuck Hayes pushes him about five feet off his spot. Notice how he scarcely improves his positioning on this repost against a slimmer Robert Horry:
He did pick up layups in transition, especially when younger; he was always one of the first players down the court, running like an automaton the moment his team possessed the ball. Notice on this play, after one of his typical trampoline closeouts, how he outruns two defenders by instantly bolting when Boston gained possession:
And he used that speed and energy on defense more than most players in history:
Garnett was a diverse defender, owning brown and black belts in multiple disciplines. His hands and feet were always active, sometimes making basic entry passes difficult:
He lacked historically good shot-blocking numbers, but was a strong rim protector, using length and verticality to swat away countless attempts:
Garnett was able to contest like this while rarely fouling, ranking as one of the lowest-fouling bigs on record since 1997. After watching tape on him for a while, it’s not the highlight blocks that stand out, but instead how well positioned he is on nearly every play. He tracks the ball, slides in step with pick-and-rolls and reacts to threats like a basketball T-cell:
To borrow Bill Russell’s vernacular, his “horizontal game” was better than his vertical one. He could switch onto bigs and smalls and was constantly in the right position to deter attacks. Even in his Celtic days, he could totally disrupt wings or help on the biggest of centers:
Overall, Garnett’s athleticism, awareness and motor made him one of the best team defenders ever. In my sampling, from 2003-09, he committed defensive errors at a rate of just 0.7 per 100 (96th percentile). This was nearly identical to his rate during my 2010 and 2011 tracking (which included 2,500 more Garnett possessions). Additionally, his frequency of “good” help plays is second in my historical tracking to only a young Hakeem Olajuwon, at over 7 occurrences per 100, ahead of Tim Duncan’s peak rates of more than 6 per 100. (Olajuwon had far more elite blocks, but Garnett was capable of mind-bending blocks too.) Even in his 16th season, he was still a rotation machine, instantly switching and recovering wherever offensive threats emerged:
Unlike most great players, Garnett’s offensive and defensive peaks nearly overlapped. He likely maxed out on D in 2003 — it was the height of his tracking numbers in my sample and the last year before he started to lose a bit of his invaluable motor — while his offense crested from 2004-06 as his passing continued to grow and his post game added polish.
In 2005, a sore knee inhibited his movement and likely eroded some of his defensive value, and in 2007 he was shut down at the end of the season with a nagging thigh injury. After a major injury ended his 2009 campaign , he returned as a defensive specialist, with a strong argument for DPOY in 2011 at age 34. His size and immense basketball IQ helped him retain defensive value until 2013.
Garnett’s robust skill set produced one of the greatest analytics portfolios in league history. He repeatedly elevated marginal teams in Minnesota, then led some of the best defenses ever in Boston. On the few occasions he was surrounded by viable talent, he guided his teams into the upper-echelon — including a transcendent Boston team — and he did this with nearly unrivaled impact metrics. Only his imperfect scoring prevented him from landing on basketball’s Mt. Rushmore.
Garnett’s development as a player in Minnesota coincided with the Timberwolves steady ascension from expansion-doormat to playoff team. In Year Two of the KG era, the Wolves drafted Stephon Marbury and crossed the 35-win plateau, and in Year Three, Minnesota played like a 45-win team before Tom Gugliotta’s season was ended by injury. Minny then started with a bang in the shortened 1999 season (53-win pace for 18 games) before contentiously trading Marbury for Terrell Brandon.
But then the Wolves spiraled into an historically unmatched stretch of personnel corrosion. First, they were stripped of three first-round draft picks for illegal contract shenanigans with Joe Smith. Then Brandon was forced into early retirement due to knee problems, playing his last full season in 2001. Two years later, Wally Szczerbiak, Minnesota’s only notable addition between 2000 and 2003, missed 30 games and hobbled around for 28 more in ’04. After the ’02 season, the Wolves couldn’t afford to re-sign their best young asset, Chauncey Billups, and even worse, guard Malik Sealy died tragically in a car accident. While many superstars enter their primes with a fully stocked kitchen, Garnett was left with a bare cupboard.
Minnesota’s roster was offensively slanted, with players like Brandon (a borderline All-Star who struggled defensively), Anthony Peeler (a streak shooter who fared well from downtown from ’01 to ’03), Szczerbiak (an excellent shooter who also lacked defensive prowess), and Troy Hudson (a solid pick-and-roll scorer who was a defensive sieve).4 During those years, the Wolves hovered around a 50-win pace, and in 2002 played at an impressive 48-win clip (2.3 SRS) in 51 games without Brandon.
Their modicum of offensive firepower paid off when everyone was healthy, likely boosted by Garnett’s elite passing and strong creation: In 2000, they were 4.5 points better than league average in relative efficiency (rORtg); in 2001, +2.2; in 2002, +4.5 in 40 games without Brandon; and in 2003 an elite +5.4 with Szczerbiak. For perspective, if those were full-season averages, Minnesota’s five-year offensive rating from 2000-04 would be +3.9, nearly identical to the ’83-87 Celtics (+4.0) and the ’00-04 Lakers (+3.9).
Garnett never captained an historically dominant offense, likely caused by his inability to collect easy points and a lack of accompanying firepower. Similarly, he never generated great scoring and efficiency rates in the postseason — his playoff numbers dipped slightly more than expected — although the sample sizes are too small to put much stock in. Over a larger sample, against quality defensive teams, he didn’t appear to decline in any abnormal way. Here’s how he stacks up in the Big 3 box score dimensions among the great modern bigs, where he was the least impressive scorer, but the most prolific creator:5
This kind of profile — lighter on scoring, strong on playmaking — is the inverse of someone like early-’60s Wilt Chamberlain. Wilt could score oodles of moderately efficient points, sometimes single-handedly swinging games, but he couldn’t help his teammates enough to generate efficient offenses in the long haul. On the other hand, Garnett couldn’t ramp up his scoring like some of the behemoths, but when surrounded with weapons, his passing and creation unlocked their potential.
His team’s defense, however, was another story during those years. Minnesota lacked defensive talent, and the Wolves constantly floated around league average with Garnett on the court and then struggled mightily when he was off it. In 2001 and 2002, they were nearly dead average with him in the game and about 4 points worse without him. Those defenses were held back by quirky lineups; in ’01, Garnett played 1,200 minutes at center next to defensive milquetoast LaPhonso Ellis in a small-ball lineup, then in ’02 logged over 1,500 minutes alongside two other bigs in oversized lineups. In 2003, playing more at his natural power forward slot, the Wolves were 1.2 points better than average with him on the court and a disastrous 7.3 points below average with him on the bench.
In 2004, Minnesota upgraded its roster, acquiring Sam Cassel, just the second teammate of Garnett’s from ’98-07 to make the All-Star game.6 The Wolves balanced their lineup with more defensive talent, and while the offense posted a mark comparable to prior years (+3.0 rORtg) it was the defense that finally brought them to the fringe of title contention, posting an rDRtg 3.2 points ahead of the league. This wasn’t a dominant defensive team, but such heights were unlikely under Flip Saunders without another star defender next to KG.7
In 2005, the Wolves roster issues came home to roost. Latrell Sprewell, on his last legs, clanged his way to a dreadful -4 percent relative true shooting percentage (rTS), and the Wolves played at a 37-win pace (-1.4 SRS) in 29 games without Cassell. In those games, the offense regressed to right around average while the defense dipped below-average, the likely penalty for trotting out Hudson, Spree and Michael Olowokandi. With Cassell, they played at a 50-win pace (3.0 SRS), making them arguably the best team in league history to miss the playoffs. In 2006, the band broke up, the roster was scrambled, and they sank into the lottery.
In the first half of the ’06 season, Minnesota hovered around .500 before panic-trading for flotsam, and in 2007, again floated at .500 before a coaching change formally derailed the year. In those two seasons, they were outscored with Garnett off the court by 11.9 points per 48 in 2,021 minutes, approaching some of the extreme lows in history. It was commendable that those teams sniffed .500 given that the roster was replete with replacement players who could barely make the league.8
When Garnett arrived in Boston, his pinpoint rotations and blanket court-coverage spearheaded the second-best relative defense since 1970, 8.6 points better than league average. Boston was strong offensively as well, posting a +3.7 rORtg when healthy. In his first two Celtic seasons, Garnett missed 39 games in which the other starters played, and without him, Boston performed like a 51-win team (3.4 SRS). But with him, they played at a 65-win pace (9.0 SRS), a degree of lift rarely matched in NBA history.9 KG missed another 19 games in 2010 and 2011 and the Boston defense dropped by a substantial 4.9 points without him. Even with Garnett’s missed time, the Celtic’s five-year run of defense ranks second since 1970.
Garnett’s non-box value-signals extend beyond his missed time. He’s arguably the king of adjusted plus-minus (APM), as he’s the only player on record with a top-40 (scaled) four-year peak in both offensive (26th) and defensive APM (first), reflecting the balance he shows on tape. His 2004 season is the best single-year on record and 13 of his seasons as a heavy-minute player fall in the top-200 ever recorded. APM is excellent at describing situational value — who correlates with changes in the scoreboard based on their presence on that particular team — but sometimes huge APM marks are merely the result of a good fit and targeted coaching. Yet Garnett posted three of the 20-best APM seasons ever with different teammates of varying style and skill, demonstrating his incredible portability:10
While many people might balk at ranking him this high, Garnett is the poster child for multiple biases discussed in Thinking Basketball. His strengths — off-ball defense and passing — are historically overlooked by telescoping on isolation scoring and block totals. Before the stigma of the first-round curse and the collapse of the Timberwolves, KG was widely argued as an equal to Duncan and hotly debated as the game’s best player. From 2000 to 2005 he won an astounding nine Player of the Months (POMs), including six of nine starting in early 2003. For comparison, Shaq claimed five POMs in that time span, Kobe four and Duncan won three in his entire career.11 Garnett was second to peak Shaq in MVP voting in 2000, then claimed 73 percent of the share in 2003, narrowly losing out to peak Tim Duncan before grabbing 99 percent of the vote in 2004. Along with Kareem and Karl Malone, he’s a member of the 25-14-5 club and is third all-time in NBA All-Star selections.
Even at eighth, I still question whether I underrate him. He achieved his incredible statistical profile on a broken franchise with mediocre coaches, whereas someone like Duncan played under a metric-enhancer in Gregg Popovich. While Garnett’s scoring limitations and lack of base strength prevented him from a GOAT-level peak, there’s a viable argument, especially from the film, that I’m slightly underselling his offense by taking his small-sampled postseason scoring efficiency at close to face value.12
He and Duncan are the most balanced megastars among the all-time greats, although Garnett leaves a slightly larger impact footprint, trumping Duncan in both (regular season) box and non-box metrics. He was a longevity giant, stringing together 13 All-NBA quality years and eight strong MVP ones by my estimation. Most importantly, his elite passing and high-post offense — which opens the lane for isolationists and allows him to pair with ball-dominant guards — is far more scalable than any other all-time great big. That’s a difference-maker at these heights, and gives Garnett a top-10 peak, just enough to propel him above Wilt for the eighth-best career in NBA history.
- In tracking Garnett, from ’96-02 players shot 11-32 against him in man D or on his closeouts. While this is a super small sample, it’s reflective of how painful it was for wings to find clean looks against him.
- This is different than Synergy’s “isolation” category. In this case, I counted on-ball actions (drive, post, pick-and-roll) that led to a shot, turnover or free throws. Helped offense, off-ball screens, transition scores or put-backs did not qualify.
- Further tracking comparisons between the two: Duncan turned it over a little bit more but Garnett got to the line less. Both of these are reflective of their overall box score data.
- Coach Flip Saunders was also offensively-inclined, using ball movement and spacing principles fairly well for the time. However, he was not a defensive mastermind, to say the least, and sometimes had Minnesota in strange zones with Garnett 20 feet from the basket.
- The chart uses the “Non-3” estimate of creation.
- For comparison, Tim Duncan had four All-Star teammates total six appearances in the same time frame. Hakeem Olajuwon had two All-Star teammates make three games between year two and 11 and Shaquille O’Neal had three All-Star teammates compile nine births just in his seven years in Los Angeles.
- An example of Flip’s offensive preferences: When Saunders took over in Detroit in 2006, the Pistons were coming off of an all-time level ’04 season and an ’05 rDRtg 5.4 points better than league average when healthy. In 2006, with the same starting five all in their prime, that number dropped to 3.1 as the offense exploded.
- As we’ve seen in this series, it’s actually quite common for megastars to suffer consecutive seasons with this kind of nadir. In addition to those listed there, Wilt Chamberlain was on two such teams in three years, including a -4.4 SRS team in 1965 before being traded.
- This is one of only eight WOWY results in my database where a team played below 4-SRS basketball without a player and above 8-SRS with that player (with a sample size of at least 20 games). The other players to do it: Larry Bird, Bill Walton, Jerry West, Tim Duncan, Patrick Ewing, Shaquille O’Neal and Kyrie Irving.
- His 2003 teammates were Szczerbiak (a spot-up shooter), Hudson (an on-ball scoring guard), Nesterovic, Peeler and Gill. His 2004 teammates were Cassell (a pick-and-roll point guard who could post up and was a viable creator), Sprewell, Johnson, Hassell, Hoiberg and Madsen. Then in Boston, he played alongside Pierce (a pinch-post isolation scorer with a more diverse attack), Allen (an off-ball shooter), Rondo, Perkins and Posey.
- Before the conferences handed out separate awards in 2001, Jordan won 17 POMs, Shaq 10, Bird and Karl Malone seven, Hakeem and Magic six.
- Specifically, his 22 points per 75 at league average efficiency from ’02-04 in 27 playoff games. For context, similar scoring seasons include ’79 Erving, ’90 Drexler, and ’98 Iverson.