You Are Here > Basketball-Reference.com > BBR Blog > NBA and College Basketball Analysis

SITE NEWS: We are moving all of our site and company news into a single blog for Sports-Reference.com. We'll tag all Basketball-Reference content, so you can quickly and easily find the content you want.

Also, our existing Basketball-Reference blog rss feed will be redirected to the new site's feed.

Basketball-Reference.com // Sports Reference

For more from Neil, check out his new work at BasketballProspectus.com.

Percentage of Team Shot Attempts “Created” While On the Floor

Posted by Neil Paine on February 23, 2011

With noted shot-creator Carmelo Anthony on the move, I was wondering which players have a hand in "creating" the highest percentage of their team's shot attempts when on the floor.

If we're just looking at a player's own shooting attempts, this is pretty easy. You can look at the percentage of team FGA a player takes when on the court (all leaders minimum 1,366 minutes):

Player Pos Tm G Min FGA FTA %FGA
Kobe Bryant SG LAL 58 1959 1121 415 33.6%
Carmelo Anthony SF DEN 50 1774 967 417 32.7%
Derrick Rose PG CHI 53 2012 1068 328 31.8%
Kevin Durant SF OKC 51 2011 1037 457 31.3%
LeBron James SF MIA 55 2100 1034 479 30.7%
Dwyane Wade SG MIA 53 1964 957 467 30.4%
Amare Stoudemire C/PF NYK 53 1949 1028 422 30.3%
Andrea Bargnani PF/C TOR 51 1828 935 254 29.7%
Michael Beasley SF/PF MIN 48 1566 828 196 29.7%
Russell Westbrook PG OKC 55 1953 932 446 29.0%
Monta Ellis SG GSW 56 2298 1156 336 28.6%
Joe Johnson SG/SF ATL 47 1688 788 182 28.5%
Dirk Nowitzki PF DAL 47 1611 737 274 28.1%
Antawn Jamison PF/SF CLE 53 1737 822 214 28.0%
Kevin Martin SG HOU 56 1745 848 476 27.8%
Eric Gordon SG LAC 41 1550 712 296 27.6%
Brook Lopez C NJN 57 1959 870 351 27.4%
Stephen Jackson SG/SF CHA 55 2025 884 260 27.3%
Luis Scola PF HOU 58 1939 920 251 27.2%
Blake Griffin PF LAC 57 2156 971 492 27.1%

You could also take it a step further and factor in free throws as well, calculating the percentage of each team's True Shooting Attempts (FGA + .44 * FTA) each player takes while on the floor:

Player Pos Tm G Min FGA FTA TSA %TSA
Kobe Bryant SG LAL 58 1959 1121 415 1303.6 34.5%
Carmelo Anthony SF DEN 50 1774 967 417 1150.5 33.3%
Kevin Durant SF OKC 51 2011 1037 457 1238.1 32.1%
Derrick Rose PG CHI 53 2012 1068 328 1212.3 31.9%
LeBron James SF MIA 55 2100 1034 479 1244.8 31.9%
Dwyane Wade SG MIA 53 1964 957 467 1162.5 31.8%
Amare Stoudemire C/PF NYK 53 1949 1028 422 1213.7 31.7%
Kevin Martin SG HOU 56 1745 848 476 1057.4 30.6%
Russell Westbrook PG OKC 55 1953 932 446 1128.2 30.1%
Andrea Bargnani PF/C TOR 51 1828 935 254 1046.8 29.6%
Monta Ellis SG GSW 56 2298 1156 336 1303.8 29.1%
Michael Beasley PF MIN 48 1566 828 196 914.2 29.1%
Dirk Nowitzki PF DAL 47 1611 737 274 857.6 29.1%
Blake Griffin PF LAC 57 2156 971 492 1187.5 28.8%
Brook Lopez C NJN 57 1959 870 351 1024.4 28.5%
Eric Gordon SG LAC 41 1550 712 296 842.2 28.4%
Joe Johnson SG/SF ATL 47 1688 788 182 868.1 28.0%
Antawn Jamison PF/SF CLE 53 1737 822 214 916.2 27.5%
LaMarcus Aldridge PF/C POR 56 2206 989 332 1135.1 27.3%
DeMarcus Cousins C SAC 53 1443 662 248 771.1 27.2%

But I think you need to take assists into account as well.

For the sake of argument, let's say a player's teammates convert his "potential assists" into made field goals at a FG% rate of (team FG - player FG) / (team FGA - player FGA). Also, let's assume that the player's rate of potential assists per team FGA while on the floor also holds for TSA from free throw attempts (.44 * FTA). Given those assumptions, you can estimate how many shots a player had his hand in creating via potential assists as well as shooting attempts.

Take for instance Kobe Bryant... He's taken 1,121 FGA and 415 FTA for 1303.6 True Shooting Attempts. He also has 280 assists, and the Lakers are shooting (2223 - 516) / (4747 - 1121) = 47.1% from the field if you exclude his totals. From this we can infer that he has participated in (280 / 0.471) = 595 shots through his passing as well. Finally, if Kobe's potential assist rate of (595 / 3334) on FGA while in the game also holds for the Lakers' (1002 * 0.44) = 441 TSA from free throw attempts, we can estimate that Kobe had a hand in creating 179 teammate FTA via passing.

Adding it all up, that's 1121 + 0.44 * 415 + 595 + 0.44 * 179 = 1,977 shots which Kobe had a hand in creating, or 52.4% of the Lakers' TSA while he was on the court. Here's how that stacks up against the rest of the league leaders:

Rank Player Pos Tm G Min FGA FTA ast_FGA ast_FTA TSA_created %Team
1 Steve Nash PG PHO 52 1732 599 205 1269 373 2122 63.9%
2 Russell Westbrook PG OKC 55 1953 932 446 1002 378 2297 61.3%
3 Derrick Rose PG CHI 53 2012 1068 328 934 281 2270 59.7%
4 Deron Williams PG UTA 53 2009 806 354 1108 351 2225 58.6%
5 Chris Paul PG NOH 58 2070 649 297 1228 372 2171 57.5%
6 Rajon Rondo PG BOS 44 1657 420 88 1094 325 1696 57.1%
7 Devin Harris PG NJN 54 1724 612 307 933 281 1804 57.0%
8 LeBron James SF MIA 55 2100 1034 479 841 305 2220 56.8%
9 John Wall PG WAS 43 1588 567 212 853 242 1620 53.1%
10 Kobe Bryant SG LAL 58 1959 1121 415 595 179 1977 52.4%
11 Jose Calderon PG TOR 50 1550 443 75 965 274 1562 52.0%
12 Tony Parker PG SAS 56 1826 757 211 811 245 1769 51.0%
13 Raymond Felton PG NYK 54 2074 801 181 1044 310 2061 50.5%
Rank Player Pos Tm G Min FGA FTA ast_FGA ast_FTA TSA_created %Team
14 Ramon Sessions PG CLE 55 1421 485 269 682 208 1376 50.5%
15 Andre Miller PG POR 55 1803 592 206 917 257 1713 50.5%
16 Jameer Nelson PG ORL 52 1539 543 108 699 222 1387 48.1%
17 Joe Johnson SG/SF ATL 47 1688 788 182 545 150 1479 47.7%
18 Dwyane Wade SG MIA 53 1964 957 467 496 180 1737 47.6%
19 Gilbert Arenas PG TOT 51 1376 585 118 518 156 1223 46.8%
20 D.J. Augustin PG CHA 57 1907 637 199 774 259 1613 46.1%
21 Manu Ginobili SG SAS 56 1738 726 306 567 171 1503 45.5%
22 Stephen Curry PG/SG GSW 48 1605 669 158 618 153 1424 45.5%
23 Monta Ellis SG GSW 56 2298 1156 336 654 162 2029 45.3%
24 Rodney Stuckey PG DET 49 1473 572 275 479 136 1232 44.9%
25 Jason Kidd PG DAL 56 1888 457 49 946 266 1542 44.6%
26 Eric Gordon SG LAC 41 1550 712 296 410 140 1314 44.4%
Rank Player Pos Tm G Min FGA FTA ast_FGA ast_FTA TSA_created %Team
27 Tyreke Evans SG/PG SAC 46 1740 781 215 561 161 1508 44.1%
28 Darren Collison PG IND 52 1580 607 151 600 171 1349 44.0%
29 Carmelo Anthony SF DEN 50 1774 967 417 292 111 1491 43.2%
30 Jason Terry SG/PG DAL 56 1803 780 145 508 143 1414 42.8%
31 Jrue Holiday PG PHI 56 1956 668 153 748 218 1580 42.7%
32 Blake Griffin PF LAC 57 2156 971 492 463 158 1720 41.8%
33 Mike Conley PG MEM 58 2075 669 181 816 243 1672 41.7%
34 Kevin Durant SF/SG OKC 51 2011 1037 457 307 116 1596 41.4%
35 Kyle Lowry PG HOU 54 1785 528 175 755 230 1461 41.3%
36 Chauncey Billups PG DEN 51 1646 536 287 566 215 1323 41.3%
37 Amare Stoudemire C/PF NYK 53 1949 1028 422 319 95 1574 41.1%
38 Stephen Jackson SG/SF CHA 55 2025 884 260 454 152 1520 40.9%
39 Luke Ridnour PG MIN 46 1373 432 94 580 167 1127 40.9%
Rank Player Pos Tm G Min FGA FTA ast_FGA ast_FTA TSA_created %Team
40 Andre Iguodala SG/SF PHI 44 1631 497 212 580 169 1244 40.4%
41 Kevin Martin SG HOU 56 1745 848 476 296 90 1393 40.3%
42 Dirk Nowitzki PF DAL 47 1611 737 274 263 74 1153 39.1%
43 Jamal Crawford SG/PG ATL 50 1532 595 205 363 100 1092 38.8%
44 Michael Beasley PF MIN 48 1566 828 196 240 69 1185 37.7%
45 Josh Smith PF/SF ATL 56 1954 755 230 418 115 1324 36.9%
46 Danny Granger SF IND 54 1954 893 301 326 93 1393 36.8%
47 Paul Pierce SF/SG BOS 55 1917 718 285 363 108 1254 36.5%
48 Tim Duncan PF/C SAS 56 1609 642 178 348 105 1115 36.5%
49 Luis Scola PF HOU 58 1939 920 251 322 98 1396 36.3%
50 Kirk Hinrich PG/SG WAS 48 1471 438 105 474 134 1017 36.0%

With point guards dominating this list, you could certainly argue that it's more difficult to create your own TSA than to create them for others with passes. But the take-home point is that any measure of "shot creation" should take into account not only the ability to hoist shots yourself, but also your passing tendencies. Someone like Carmelo Anthony is good at the former, but not the latter, perhaps overrating him in the public's perception when it comes to great shot-creators.

ShareThis

30 Responses to “Percentage of Team Shot Attempts “Created” While On the Floor”

  1. Neil Paine Says:

    This is a very "dumbed-down" version of the FGA, FTA, & AST aspects in Dean Oliver's individual possession rates, btw. The point was easier to illustrate this way, but Dean's work is about a billion times more rigorous.

  2. Alex Sonty Says:

    @Neil (RE: "Dumbed down" Oliver) --

    I wouldn't call this "dumbed down," as much as a simple (in a non-pejorative sense) of gauging relative volume. It's a more in-your-face reflection of through whom offenses are run. Great post.

  3. Neil Paine Says:

    Thanks. Dean's is great because it actually attempts to weigh TSA creation via shooting vs passing, such that everything adds up to the team total. Here, I'm double counting every assisted TSA and giving the passer and shooter equal credit. But it's still useful to see how guys like Rose, Lebron, Kobe, etc. have a level of "creating" beyond what a one-dimensional pure scorer like Melo brings to the table.

  4. AYC Says:

    Don't players hit a higher % of FGA that come off of passes? When the subject of clutch scoring stats came up awhile ago, the Kobe-lovers pointed to his low % of shots assisted to explain away his low shooting %. Kevin Martin, on the other hand, had a very high % of his FG assisted. So using team FG% less the players contribution might be overstating "potential" assists. Is there a way to quantify how much better players shoot off a pass, rather than creating on their own?

  5. yariv Says:

    Neil, could you give some list using weights for assisted shots? I guess using 50-50 or even all for the passer (since we're talking about shot creation, not credit for scoring)? It is not deductible from the numbers you already put up, because you'll have to reduce some of the credits for actual shots.

  6. P Middy Says:

    A valiant effort. Of course, it's impossible to take into account possession that create shots but don't directly lead to a stat.

    Example, the ball is sent down low to Howard. He makes a move and gets a double team, he kicks it out. Defense begins it's recovery rotations. Two perimeter swings later, Anderson hits a three. Is that possession more about responding to the double team correctly, or about those perimeter swing passes?

    Example two, Kevin Garnett sets a (moving) screen on Allen's man allowing him to swing from one side to the other. Rondo hits Allen with the easy pass for a jumper. Credit goes to Allen and Rondo, Garnett's (moving) screen is lamentably lost.

  7. Neil Paine Says:

    #4 - You're right about that -- in fact, 82games studied the matter a few years ago. I might be able to estimate that someday.

    #5 - I can try to take a look at that at some point.

    #6 - You've just described every pro-plus/minus argument, like, ever.

  8. k Says:

    I'd love to see an all-time breakdown. Particularly curious as to how high volume scorers like Jordan and Malone would be versus super-facilitators like Magic and Stock.

  9. Patrick Hodgdon Says:

    Very cool Neil.

    I especially like seeing that Blake Griffin is the highest rated forward on this list! Says a lot about his passing skills and ability to draw fouls. Can't wait to see what happens when the 17 foot bank shot is even more automatic and he adds another 3-4 feet to hit 3 pointers.

    Will be linking to this later today over at QuakeGriffin.com

  10. Walter Says:

    Very cool Neil.

    I took the information in this post along with the "Black Holes by Position" post and took it one step further and tried to answer the question of who generates the most shots (either for themselves of their teammates) per touch.

    I took the Touches per Minutes statistic in the black hole thread and multiplied it by the minutes in this thread to get an estimate for the number of touches. I then took the TSA-Created and divided by the touches to get the final ratio.

    The results are pretty interesting. I will leave it to you to quickly perform the same thing and post it here if you would like.

    As always though... great stuff!!

  11. P Middy Says:

    #7 - I did? That sucks. Because I kind of hate plus/minus.

  12. Neil Paine Says:

    #11 - Yeah, maximizing how well your team does when you're on the court is stupid.

  13. ElGee Says:

    Hi Neil - I actually hand-track something like this. LIterally, the goal is to see who is creating open shots for teammates:

    http://elgee35.wordpress.com/2011/02/22/opportunities-created-oc-leaders-2010-playoffs/

  14. Greyberger Says:

    I'm surprised to see the Spurs' big three all in the top fifty of this list. I shouldn't be, they make it onto every list and usually hang out closer together.

    Other teams with three players above 36% or more - Houston with Lowry, Martin and Scola. Dallas with Kidd, Terry and Nowitski. Atlanta with Crawford Smith and Johnson.

    In Atlanta and Dallas, there's a sixth-man setup that helps explain how three players can be so involved in the team's offense. San Antonio used a scheme like that for Ginobili for years.

  15. Jason J Says:

    #12 - He probably hates the 350% standard margin of error...

  16. P Middy Says:

    Too many times I've gone to the boxscore for the game I just watched, and the +/- is garbage.

  17. Kelly Says:

    This is what I've been arguing all along for Rose inclusion in the MVP conversation and all the efficiency geek. Thee's one name in the top five of both lists. That sounds like MVP to me.

  18. Neil Paine Says:

    #15, 16 - Right, the margin of error between "true skill" and observed performance is huge:

    http://www.basketball-reference.com/blog/?p=8772

    But at the same time, raw on/off +/- tells you something that actually happened, that you could actually have seen with your own eyes. If a player was +9, his team was literally 9 points better when he was in the game than when he wasn't. People always say "scoreboard!"... well, this is that, in stat form. I would think it would appeal to old-school fans more than all of the advanced stats in the world.

  19. Neil Paine Says:

    #13 - Great stuff. In fact, good work with your whole blog; I added a link to our sidebar.

  20. Jason J Says:

    #18 - Actually when I first saw the adjusted +/- breakdowns on 82games years ago, I thought they'd given us the Ark of the Covenant of stats... My face melted! It was a let down when I realized that a guy like Ben Gordon can have top 5 defensive +/- at his position under coach Skiles and then fall off the map under Del Negro. Ironically I think +/- at times can be TOO all-inclusive to accurately describe what an individual does for his team. 5 man unit +/- which Mark Cuban swears by seems to me a better way to use the stat.

    Raw +/- is something I look at in a box score, but it often just tells you that one team kicked the other's butt all night. I'm sure Judd Buechler and Bill Wennington had some terrific raw +/- nights playing for the '96 Bulls while scoring a combined 7 points with 4 rebounds and 4 fouls.

  21. ElGee Says:

    @Neil - thanks. Feeling is certainly mutual.

    I like what you've done here attempting to estimate "shots created" in this manner, but I think the same problems with relying on assists for creation creep in to play here as well (eg Rondo). That said, I'd be interested in the data for past seasons. What does it look like in the 80s? What does 87 Magic look like? Jordan? Etc. I also wonder if it could be improved by incorporating percentage of FG's assisted on (82games has it).

  22. petey Says:

    Interesting that LeBron & Kobe are the only non PGs in the top ten. Or maybe just expected. Thinking about them brings to mind something one hears frequently from commentators - the idea that their mere presence on the floor draws so much of the defense's attention that teammates get better shots/succeeed at a higher rate than when they are off the floor. I'm thinking of say, how does the threat of Kobe on the perimeter receiving a pass from Pau in the high post create more space for a diving Odom to catch a quick flip from Pau and finish at the rim? How about LeBron drawing so much attention cutting and moving that Bosh or Miller sees more daylight than normal and buries a jumper? It seems to be a really common observation (He makes the game so much easier teammates etc) and definitely part of the whole "observed performance" idea - is there a way to quantify it and factor it into this type of stat? It could be as crude as a measure of teammates' shooting % with and without said star on the floor being some sort of multiplier. (I.e. if LeBron's teammates shoot 10% better with him on the floor than without, multiply the shot creation percentage listed here by 1.10) If they shoot worse, wouldn't that also be a good indicator of whether or not a supposed "ball-hog" like Kobe causes his teammates to just stand-around watching him pick up stats (both points and assists) or if his overall impact on the floor offensively really does lead to more & better shots for his team.

  23. Alex Sonty Says:

    #13 - Great work. I added you to my Google Reader.

  24. sudoku Says:

    Kobe for MVP!

  25. Neil Paine Says:

    #22 - You're actually talking about what are called "skill curves":

    http://www.basketball-reference.com/blog/?p=8522

    http://www.basketballgeek.com/2009/02/17/basketball-on-papers-skill-curves/

    http://www.48minutesofhell.com/spurs-stats-skill-curves-matt-bonner-usage-efficienc

    It's sort of a fundamental debate in statgeekery:

    http://sonicscentral.com/apbrmetrics/viewtopic.php?t=1070

  26. P Middy Says:

    23 - DITTO!

  27. ATL Says:

    Question:

    In your explanation of how you got to your %Team, you used Kobe as an example. Where did you get the numbers 3334 and 1002 from? And how did you get to 179? Not sure if it is just me, but I am a little unclear on how you finally got to your %Team.

    Thanks.

  28. Neil Paine Says:

    3334 was the number of FGA we estimate the Lakers taking while Kobe is on the floor. They took 4747 shots, he was on the floor for 70.2% of their minutes, so .702*4747 ~ 3334. Same with free throws -- 1426 attempts total, 70% of the available minutes, an estimated 1002 FTA while he was on the court.

    179 assisted FTA is a little more complicated. I estimated that Kobe had potential assists on 595 of the Lakers' 3334 FGA when in the game. I assume this relationship also applies to potential assists on FTA. To make FTA and FGA equivalent I multiply FTA by 0.44 (so each is denominated in possessions). 1002*0.44 = 441. Using the relationship I found for FGA, I multiply (595/3334) * 441 to get roughly 79 free throw possessions being potentially assisted by Bryant. Then to get that back into FTA, I divide by 0.44, giving Kobe 179 potentially assisted FTA.

  29. Neil Paine Says:

    #20 - If I'm a coach, though, I'm encouraging my players (my role players, at the very least) to focus on that stat. Most stats become troublesome if the players are too self-aware, because they start to play to the stat at the expense of team performance. But when the stat is team performance, you want them to obsess over it as much as possible.

  30. Jon L. Says:

    Obviously Carmelo's attempts will decrease now as he is moving to the Knicks. He will actually have to share the ball with Amare.