James Harden

[Mike G]

"... I understand the concern about adding Harden to a team with NBA Finals aspirations given his track record of playoff underperformance."
https://www.espn.com/nba/story/_/id/478 ... whats-next

Kevin Pelton article at ESPN on the Harden-Garland trade.
Is anyone else wondering what KP is referencing with "playoff underperformance".

His regular season career BPM and playoff BPM both are 6.2
Season and playoff BPM in each stop:

Code: Select all

tm    RS   poG   PO
OKC   2.3   43   4.5
Hou   8.1   85   7.6
Brk   5.7    9   9.1
Phl   5.0   23   2.9
LAC   4.1   13   5.8

https://www.basketball-reference.com/pl ... l_advanced
(middle column is 'playoff games')
No big underperformance as an individual. How have his teams fared? Separated by whether his team was favored (by SRS) or not:

Code: Select all

SRS:   Favored       Underdog
tm   games series  series games
OKC  13-8   5-1     2-2   11-11
Hou  25-15  5-2     2-6   17-28
Brk   4-1   1-0     0-1    2-2
Phl   8-2   2-0     0-2    5-8
LAC   5-8   0-2     ...    ...
tot  55-34 11-5     4-11  35-49

He's averaged 1.5 more minutes in playoff games -- normal that his effectiveness would diminish.
Not counting 4-0 and 4-1 winning series, he's gone an extra 2.5 mpg.

His two Clippers' playoffs, 1st round exits, have been at the hands of the Mavs and Nuggets.
2024: Luka, Kyrie, Gafford, and Lively totaled about 50% more mpg in that series than they had in the RS.
Kawhi played a lot less both years.

[Crow]

In his last 7 playoff series, his efg% has been 51% or below 5 times. His 3pt fg% has been 35% or below 5 times. But the ts% has always been 57+%. It could be influenced by appearance of raw fg% of 40-42% 5 times.

Struggle might refer to specific games or moments. 10 of his last 24 playoffs had bad or very bad shooting overall and only 2 outstanding. 10 times below 30% from 3. 8 times with 10+ assists but 11 times with 4+ turnovers.

[DQuinn1575]

Ben Taylor pointed out Harden's numbers went down a lot in playoff series as the games progressed. Per B-Ref here is BPM by game

1 31 g /7.3
2 30g/5.6
3 30g/8.0
4 30g/6.7
5 29g/4.5
6 16g/4.9
7 7g/4.6

If I do an estimated average just using games, he goes from 6.9 in games 1-4 to a 4.6 in games 5-7. with a sample size of 42 games. It's really his big drop in the last games of a series that have been the issue.

[Mike G]

BPM is the one stat that takes into account the exact opponent strength -- not just the opponent regular season team averages, but the actual players who have been in the games.
It may over- or under-rate certain stats or positions, but it should be consistent about a given player's rise or fall, relative to his normal self.

Of course in playoffs you're almost entirely going against above-avg and elite opposition; your shooting%, reb%, etc are all likely to fall. Add in the extra minutes you are expected to go, and you have to hope the other team is just as tired as you are.

Are games 5-7 more important than 1-4? In retrospect they seem to be. But if you win the first few, the later ones matter less.
It's like all shots made or missed in the 1st quarter of a game may turn out to be plays that won or lost the game. In the 4th, it's only sometimes.

Harden has made the playoffs in all 16 of his seasons. In 16 losing playoff series, the team that eliminated his went on to win the championship 7 times: LAL (2010, 2020), Dal 2011, MIa 2012, GSW (2015, 2018), Mil 2021.

ALSO: currently in the West, Harden is 15th in WS and 13th in VORP.
In the East, he ranks 12th in WS and 8th in VORP.

[DQuinn1575]

BPM is the one stat that takes into account the exact opponent strength -- not just the opponent regular season team averages, but the actual players who have been in the games.
It may over- or under-rate certain stats or positions, but it should be consistent about a given player's rise or fall, relative to his normal self.

Of course in playoffs you're almost entirely going against above-avg and elite opposition; your shooting%, reb%, etc are all likely to fall. Add in the extra minutes you are expected to go, and you have to hope the other team is just as tired as you are.

Are games 5-7 more important than 1-4? In retrospect they seem to be. But if you win the first few, the later ones matter less.
It's like all shots made or missed in the 1st quarter of a game may turn out to be plays that won or lost the game. In the 4th, it's only sometimes.

Harden has made the playoffs in all 16 of his seasons. In 16 losing playoff series, the team that eliminated his went on to win the championship 7 times: LAL (2010, 2020), Dal 2011, MIa 2012, GSW (2015, 2018), Mil 2021.

ALSO: currently in the West, Harden is 15th in WS and 13th in VORP.
In the East, he ranks 12th in WS and 8th in VORP.

Well if I counted it right he is 4-11 in Game 6's - while being 86-72 in the other games. Obviously you can say all games count the same, buf pretty sure you'd rather have a guy be at a 6.0 level for all games, rather than a game be a 7.0 for 4 games and a 4.5 the last 3.

[Mike G]

Good question. If you finish the other team off in short order, you get more rest before the next series; so there's that.
So we have Harden dropping about 2.3 of his BPM after game 4; and if that is for 3/4 of a game (36 min), about 1.7 ppg
A team that loses 1.7 pts in scoring differential loses about .056 in Win%
If two teams are otherwise equally matched, but your team has this guy who makes your team .028 better in some games and .028 worse in others, how does this affect your odds in a best-of-7 series?

Team James Harden vs team Rock Steady:

Code: Select all

WIn  RS     JH
4   .063   .078
5   .125   .131
6   .156   .149
7   .156   .149
any .500   .507

No disadvantage to be seen here. Indeed, he's 10-5 in 4- and 5-game series.
It's possible that Harden has gone more minutes in Games 5-7, and it looks like that has been bad for his BPM.
His Houston regular seasons:

Code: Select all

year    mpg   BPM
2013   38.3   5.6
2014   38.0   5.6
2015   36.8   8.8
2016   38.1   6.6
2017   36.4   8.7
2018   35.4   9.9
2019   36.8  11.0
2020   36.5   9.6

Correlation -.77

[Mike G]

Harden game 1 thru 7 average rates in his playoff career.
Bottom line is correlation to his minutes per game for each column.

Code: Select all

Gm   mpg    TS%   TRb%  Ast%  St+Bk   TO%   Usg%    ORtg   DRtg    BPM  bpmG
1   34.9   .594   7.7   32.1   3.1   13.7   29.9   117.7   110.3   7.0   7.7
2   35.8   .595   9.2   29.9   3.6   14.2   28.5   119.2   110.1   5.8   6.9
3   36.1   .583   8.9   33.4   2.8   11.9   29.2   119.7   110.9   8.0   8.6
4   37.8   .597   9.4   29.9   4.8   16.9   27.5   115.7   109.3   6.3   7.7
5   37.1   .577   8.5   28.4   3.6   17.1   26.4   111.8   109.1   4.6   6.2
6   38.5   .566   8.0   33.1   3.6   17.0   25.6   111.1   110.8   4.8   6.6
7   40.5   .501   7.9   31.9   6.6   17.5   25.1   103.7   108.6   5.0   7.2
mpg corr:  -.85  -.26    .08   .86    .77   -.91    -.91   -.56   -.62  -.23

Note: I didn't sum his shots and makes, just took the avg of his TS% for each game; same with other stats, not minutes weighted; so there may be oddities.
Final column is bpmG = (BPM+3.5)*mpg/48 -- 3.5 is about halfway between "replacement level" and "net zero".
In game 7's, his steal% (3.5) would rank behind only Cason Wallace this season; his block% would rank behind only Derrick White, among guards.