Post by eric on Dec 19, 2017 12:31:05 GMT -6
Do Players Chase Triple Doubles?
Yes.
Can We Measure That Phenomenon?
Yes.
Thanks for reading, like and subscribe.
.
Since 1964 there have been 1806 triple doubles of points, rebounds, and assists in the National Basketball Association as recorded on basketball-reference. (The records only go back to 1964. Sue me.) We can also look for games where instead of having at least ten in each stat players obtain at least some other number. That looks like this:
To measure a player chasing a triple-double, we can look at games where a player had at least X in each stat but no more than X+1 in every stat. For example, 58 of the 139 triple-thirteens are also triple-fourteens, so there are 81 triple-thirteens-but-not-fourteens. That graph looks like this:
Shameful. It's easier to see if we divide our triple axis into one and take it to the sixth and a halfth power before comparing it to the relative frequency of each bin, yes this is a real math thing I did.
Shameful!! When one dot is on one side of the line and all the others are on the other, you know something's going down. Look what happens when we just delete that dot and re-regress.
SO much better. UGH. Here's the numerical breakdown of what proportion each bin actually does make up versus how much it should:
So NBA players in general chase triple-doubles, which we already knew. Are there any whose pursuit is especially egregious? Let's look at all 25 players with 50 or more b-r-attested triple eights in terms of how many more or less triple-tens-but-not-elevens-or-mores they have than the model predicts given their triple-eights-or-better as a percentage of their triple-eights-or-better (again, this is totally a real math thing that I did):
Shame! Shame! Shame! Ho boy is there a lot of shame to go around. The top three triple double men to chase at league average or lower are as expected paragons of Playing the Game the Right Way: LeBron James, Larry Bird, Clyde Drexler. Notoriously selfish players Jason Kidd, Rajon Rondo, and Mark Jackson are the chaseiest with Russell Westbrook nipping at their gluttonous heels. Rod Strickland meanwhile apparently went out of his way NOT to get triple doubles. That's the kind of team first playing that gets you on ten teams in your career.
Yes.
Can We Measure That Phenomenon?
Yes.
Thanks for reading, like and subscribe.
.
Since 1964 there have been 1806 triple doubles of points, rebounds, and assists in the National Basketball Association as recorded on basketball-reference. (The records only go back to 1964. Sue me.) We can also look for games where instead of having at least ten in each stat players obtain at least some other number. That looks like this:
triple n
8 7027
9 3412
10 1806
11 718
12 328
13 139
14 58To measure a player chasing a triple-double, we can look at games where a player had at least X in each stat but no more than X+1 in every stat. For example, 58 of the 139 triple-thirteens are also triple-fourteens, so there are 81 triple-thirteens-but-not-fourteens. That graph looks like this:
Shameful. It's easier to see if we divide our triple axis into one and take it to the sixth and a halfth power before comparing it to the relative frequency of each bin, yes this is a real math thing I did.
Shameful!! When one dot is on one side of the line and all the others are on the other, you know something's going down. Look what happens when we just delete that dot and re-regress.
SO much better. UGH. Here's the numerical breakdown of what proportion each bin actually does make up versus how much it should:
triple actual model diff
8 .514 .513 .00
9 .229 .233 .00
10 .155 .113 .04
11 .056 .056 .00
12 .027 .028 .00
13 .012 .012 .00
14 .008 .004 .00So NBA players in general chase triple-doubles, which we already knew. Are there any whose pursuit is especially egregious? Let's look at all 25 players with 50 or more b-r-attested triple eights in terms of how many more or less triple-tens-but-not-elevens-or-mores they have than the model predicts given their triple-eights-or-better as a percentage of their triple-eights-or-better (again, this is totally a real math thing that I did):
3x2x chase% Name
126 10% Magic Johnson*
107 13% Jason Kidd
89 11% Russell Westbrook
78 5% Oscar Robertson*
69 7% Wilt Chamberlain*
61 3% LeBron James
59 4% Larry Bird*
42 6% Fat Lever
33 11% James Harden
30 12% Rajon Rondo
29 9% Grant Hill
28 9% Michael Jordan*
23 3% Clyde Drexler*
21 11% Chris Webber
21 4% Kobe Bryant
20 -1% Charles Barkley*
18 12% Mark Jackson
17 2% Scottie Pippen*
16 2% Kevin Garnett
15 10% Gary Payton*
15 7% Antoine Walker
13 7% Chris Paul
13 8% Alvin Robertson
12 0% Lamar Odom
6 -6% Rod StricklandShame! Shame! Shame! Ho boy is there a lot of shame to go around. The top three triple double men to chase at league average or lower are as expected paragons of Playing the Game the Right Way: LeBron James, Larry Bird, Clyde Drexler. Notoriously selfish players Jason Kidd, Rajon Rondo, and Mark Jackson are the chaseiest with Russell Westbrook nipping at their gluttonous heels. Rod Strickland meanwhile apparently went out of his way NOT to get triple doubles. That's the kind of team first playing that gets you on ten teams in your career.
