The NBA Playoffs are contested under a bright spotlight that can enhance or diminish a player’s historical standing. For every Hakeem Olajuwon or Isiah Thomas — players who augmented their already-impressive resumes with memorable postseasons — there’s a Karl Malone or James Harden — players whose reputations took a hit due to their playoff shortcomings.

I wanted to come up with a way to identify which individuals consistently rose up (or fell back) in the playoffs, so my first step was to select a group of players to include in the study. I ended up restricting my attention to players who:

Debuted in the 1951-52 season (when the NBA began tracking minutes played) or later.

Played at least 1,500 career minutes in the postseason.

This gave me a sample of 425 players. I fit a simple linear regression model using postseason Player Efficiency Rating (PER) as my dependent variable (y) and regular season PER as my independent variable (x).

Now, on my first attempt I realized I had a problem with players like Nate “Tiny” Archibald. If you look at Archibald’s top six seasons by PER, you’ll find he only made one playoff appearance. Most of Archibald’s postseason minutes came late in his career, when he was no longer the same player. In other words, Archibald’s postseason expectation is inflated by his non-playoff seasons. Thus, I decided to eliminate any season in which the player did not appear in both the regular season and postseason (i.e., the player’s regular season PER will only include seasons in which he also appeared in the postseason).

With that change in mind, here’s the resulting regression equation:

`PS_PER = 0.9306 * RS_PER`

Note that these players lost, on average, about 6.94% of their regular season efficiency in the postseason.

Going back to Archibald, his regular season PER in his playoff seasons is 16.1 (as opposed to 18.0 if you include all of his seasons), so his postseason expectation is:

`PS_PER = 0.9306 * 16.1 = 15.0`

Archibald’s actual postseason PER is 12.7, so his residual is:

`residual = 12.7 - 15.0 = -2.3`

In other words, Archibald’s actual postseason PER is 2.3 below what would have been predicted using the regression model.

I calculated residuals for the entire sample, then standardized them so the differences are expressed as standard deviations above or below average. Archibald’s standardized residual is –1.71, meaning his difference is 1.71 standard deviations below average.

Here’s a histogram of the standardized residuals for all 425 players:

Now that all of the details are out of the way, let’s get to the goof stuff. Do you see that bar with only one observation on the far left-hand side of the histogram? That’s none other than last season’s MVP Award winner, Joel Embiid.

Embiid’s regular season PER in his playoff seasons is 28.1, but that falls to 21.6 in the postseason. Even after adjusting for the expected drop in postseason efficiency, Embiid still falls well below his predicted value (4.55 below, to be exact). His standardized residual is –3.45, by far the lowest such value in the pool:

–3.45 — Joel Embiid

–2.64 — Bill Cartwright

–2.62 — Lindsey Hunter

–2.48 — Tree Rollins

–2.29 — Clifford Robinson

Embiid’s marksmanship suffers considerably in the playoffs. In the sampled seasons, Embiid shot 50.3% from the field and 33.4% from 3-point range during the regular season, but just 46.1% and 28.0%, respectively, in the playoffs. That’s a postseason effective field goal percentage of 49.1%, 3.5 percentage points below the league average for those playoffs.

Clifford Robinson’s scoring efficiency also plummeted in the playoffs. His regular season shooting percentages are 43.8/35.6/68.9, but in the postseason those figures fall to 39.3/29.8/62.9.

Robinson is one of 183 players to score at least 1,000 postseason points in the 3-point era. He ranks last in that group with a true shooting percentage of 46.2%, 1.3 percentage points lower than the next-closest player (Robert Reid, 47.5%).

On the flip side, here are the top five overachievers:

4.45 — Jamal Murray

3.71 — Baron Davis

2.63 — Hakeem Olajuwon

2.43 — Tim Thomas

2.26 — Gus Williams

Two players in particular stand out when it comes to exceeding playoff expectations: Jamal Murray and Baron Davis.

Murray shouldn’t come as a shock, as his performances in the 2020 NBA Bubble and in the Denver Nuggets’ 2023 championship run have received plenty of attention. Murray has made three postseason appearances, recording PERs of 19.4 (2019), 23.7 (2020), and 21.6 (2023). Those are considerable improvements over his regular season figures of 16.0, 17.7, and 18.0, respectively.

Davis may be a surprise, but he frequently stepped up his game in the postseason, most notably with the “We Believe” Golden State Warriors in 2006-07. Those Warriors upset the 67-win Dallas Mavericks in the first round of the NBA Playoffs, as Davis averaged 25.0 PPG, 6.2 RPG, and 5.7 APG with 54.0/45.5/77.1 shooting splits for the series.

Olajuwon’s reputation as a player who excelled in the postseason is well known, but I found it eye opening how much he separated himself from the other stars in this analysis.

What’s a star? There are any number of ways you could define this, but let’s say a “star” is a player with at least 10 career All-Star selections. There are 39 players in this sample who meet that definition, and Olajuwon’s standardized residual is far and away the highest:

2.63 — Hakeem Olajuwon

1.78 — Elvin Hayes

1.70 — Isiah Thomas

1.36 — Bill Russell

1.36 — LeBron James

1.34 — Jerry West

1.34 — Tim Duncan

I’ll have to admit I was surprised to see Elvin Hayes’ name in the second spot, but in the sampled seasons his PER was 18.0 in the regular season and 19.1 in the playoffs. Hayes was especially good for the Washington Bullets in their 1977-78 title season, recording a PER of 20.3 in the playoffs, a significant improvement over his regular season mark of 17.1.

Honorable mention goes to Rick Barry, an eight-time NBA All-Star who surely would have been a 10-time selection if he hadn’t missed five seasons after jumping to the ABA. Barry’s standardized residual is 1.48, which would have put him fourth on the list above.

For completeness, here are the “stars” with the lowest standardized residuals:

–0.96 — Bob Pettit

–0.93 — David Robinson

–0.87 — Wilt Chamberlain

–0.85 — Karl Malone

–0.61 — Chris Bosh

–0.52 — Patrick Ewing

–0.48 — James Harden

Bob Pettit does not carry the label of postseason underachiever that several of the other players above do. In fact, his 1957-58 St. Louis Hawks are the only team to defeat a Bill Russell-led squad in the NBA Finals. That said, his career PER of 22.6 in the playoffs is 1.27 below what would have been expected based on his regular season performance.

When people assess David Robinson’s place in history, in particular his ranking relative to Olajuwon, inevitably the 1995 Western Conference Finals come up. Robinson was significantly outplayed by Olajuwon as his San Antonio Spurs lost to the Houston Rockets 4-2:

Olajuwon — 35.3 PPG, 12.5 RPG, 5.0 APG, 4.2 BPG, 56.0 FG%

Robinson — 23.8 PPG, 11.3 RPG, 2.7 APG, 2.2 BPG, 44.9 FG%

To be honest, Robinson did not play *that* badly; Olajuwon was simply superhuman.

Let me wrap this up with two lists. The standardized residuals above are similar to a rate statistic (e.g., points per game) rather than a total statistic (e.g., points). We can create a “total” for each player by multiplying his standardized residual by his postseason minutes played. Here’s the top 10 when you do that:

15,805 — LeBron James

15,103 — Hakeem Olajuwon

12,545 — Tim Duncan

10,417 — Robert Horry

10,231 — Bill Russell

9,674 — Draymond Green

9,327 — Kawhi Leonard

9,232 — Rajon Rondo

9,172 — Jamal Murray

9,034 — Reggie Miller

And here’s the bottom 10:

–9,218 — Bill Cartwright

–8,890 — Clifford Robinson

–7,580 — Robert Parish

–7,360 — Kyle Lowry

–7,248 — Gary Payton

–6,707 — Karl Malone

–6,636 — JJ Redick

–6,607 — Wilt Chamberlain

–6,534 — Klay Thompson

–6,336 — Joel Embiid

## Playoff Booms and Busts

Justin, I think this is great! For me, the key is comparing apples to apples: RS seasons to the same PO season, which is what you do here. Too many commentators just compare overall RS to overall PO. I love what you did here!

However, I'd take it a step farther and compare RS to PO in two ways: First, compare PO performance to a standard FOR EACH SEASON. For example, I used playoff standards of 22.7 PER, .180 WS/48, AND 6.5. Then, find how often a player exceeded those standards. For example, Jordan in 13 playoffs exceeded those standards 36 of 39 POs (92%), which was the highest percentage of the top 25 players I examined. After MJ:

87% LeBron

77% West

72% Magic

65% Wilt

63% Durant

63% Kawhi

63% CP3

58% Robinson

57% Duncan

Second--and this is more directly connected to your article--I compared playoff PER, WS/48, BPM to regular season PER, WS/48, BPM (adjusting for the difference in RS and PO; for example, I found PER in PO averaged 0.94 of RS PER; I think this is the same adjustment that you found). Then, I examined how often PO exceeded PO. Jordan's PO (adjusted) exceeded his (raw) RS in PER, WS/48, PER in 32 of 39 times (82%), which was the highest. Next highest were:

81% Kawhi

77% West

70% Isiah

69% Russell

69% Olajuwon

67% Dirk

63% Durant

60% LeBron

60% Oscar

58% Wilt

Jordan and West rank high in both exceeding the standard and also their own PO > RS. Isiah and Russell were each low in meeting the standard but high in their own PO > RS. Lowest in both meeting the standard and PO > RS were Karl Malone (no surprise) and ... Larry Bird!!! Kobe was also near the bottom in both!

He's only at 1049 minutes, but could you calculate Luka's figures for me? Playoff Luka is a thing, right? 🔥✌🏼