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.

Losing Your (Inefficient) Leading Scorer Hurts Your Team

Posted by Neil Paine on May 23, 2011

Last week, I ran a post (prompted by this post at the Wages of Wins) wherein I tried to determine the offensive impact when a team loses its leading scorer. I found that, since 1986 at least, a team loses about 2 points of offensive rating relative to the league average when its top scorer by PPG doesn't play.

I got a lot of great feedback from that initial post, so I decided to try my hand at a sequel after making a number of improvements to the study:

  • One complaint was that I was lumping efficient scorers in with inefficient ones in the original study. No one is really debating whether losing LeBron James will hurt an offense, but one of the core questions is whether losing Carmelo Anthony or Rudy Gay has a negative impact as well. To that end, I'm now isolating only teams with inefficient leading scorers. This means a team's PPG leader, minimum 1/2 of team games played, with either a Dean Oliver Offensive Rating or True Shooting % that was equal to or below the league's average that season.
  • Another complaint was that I looked at offense alone, rather than the total impact of the player's loss. So now I'm looking at the change in team efficiency differential (offensive efficiency minus defensive efficiency) when a player is in and out of the lineup.
  • While I accounted for strength of opponent in the last study, I didn't account for home-court advantage. Now I have added an HCA term to what we would predict an average team to put up vs. a given opponent (+4 pts/100 of efficiency differential to the home team), in addition to an SOS term (the opponent's efficiency differential in all of its other games).

What follows is a massive table that shows the results of this new study. The outcome (the bottom-right cell) is the average change in efficiency differential when an inefficient leading scorer plays vs. when he does not play, weighted by possessions without the leading scorer. If it is positive, it is evidence that even inefficient scoring is an attribute that teams find difficult to replace in a salary-capped economic system; if it is negative, it is evidence that scoring is overrated if it's not done efficiently, and that inefficient #1 options can be replaced with relative ease.

To the data dump (mouse over column headers for descriptions):

w/ Scorer w/o Scorer
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
1986 ATL Wilkins 7806.8 109.0 106.2 2.8 2.6 409.5 97.7 103.3 -5.6 0.5 2.1
1986 CLE Free 7630.6 106.3 108.6 -2.4 -1.8 715.1 101.7 109.2 -7.6 -7.0 5.2
1986 DAL Aguirre 7567.6 112.5 111.2 1.4 0.8 845.4 110.9 112.5 -1.5 0.7 0.1
1986 IND Williams 7859.6 103.2 106.3 -3.1 -2.5 412.9 99.1 106.3 -7.3 -7.3 4.8
1986 NYK Ewing 4925.7 100.5 103.8 -3.2 -2.6 3158.1 99.5 109.0 -9.5 -9.2 6.7
1986 PHO Davis 7376.1 104.7 108.4 -3.7 -3.6 1232.8 105.2 103.0 2.2 1.8 -5.4
1986 SAC Johnson 8378.4 106.5 109.5 -3.0 -3.0 0.0 0.0 0.0 0.0 0.0 0.0
1986 SAS Mitchell 8532.8 106.9 108.7 -1.8 -1.8 0.0 0.0 0.0 0.0 0.0 0.0
1986 SEA Chambers 6525.2 107.0 106.7 0.3 0.1 1533.9 103.3 105.0 -1.8 -2.2 2.2
1986 WSB Malone 7967.9 103.8 105.2 -1.4 -1.0 192.1 89.0 109.9 -20.8 -21.7 20.7
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
1987 CLE Harper 8310.2 103.0 106.7 -3.8 -3.3 0.0 0.0 0.0 0.0 0.0 0.0
1987 GSW Carroll 8310.0 109.0 111.6 -2.6 -2.4 109.0 118.4 99.1 19.3 4.3 -6.7
1987 IND Person 8174.9 106.4 107.0 -0.6 -0.3 0.0 0.0 0.0 0.0 0.0 0.0
1987 LAC Woodson 7599.4 102.1 112.4 -10.3 -10.0 821.6 98.6 117.1 -18.5 -16.1 6.1
1987 MIL Cummings 8241.1 109.8 105.9 3.9 3.9 0.0 0.0 0.0 0.0 0.0 0.0
1987 NYK Ewing 6202.7 105.2 109.9 -4.7 -4.3 1891.9 104.8 116.6 -11.8 -10.2 5.9
1987 SAS Robertson 8267.7 106.0 110.8 -4.8 -4.7 101.6 114.2 133.8 -19.7 -18.4 13.7
1987 UTA Malone 8469.7 104.4 104.0 0.4 0.1 0.0 0.0 0.0 0.0 0.0 0.0
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
1988 ATL Wilkins 7535.8 111.9 108.5 3.4 3.6 375.4 110.3 99.6 10.7 11.4 -7.8
1988 CLE Daugherty 7703.5 107.1 106.4 0.8 1.1 282.5 110.5 109.4 1.1 0.4 0.7
1988 DEN English 8476.7 110.4 106.8 3.6 3.2 192.9 111.5 98.0 13.5 7.8 -4.6
1988 IND Person 7728.1 106.9 107.7 -0.9 -0.6 287.3 112.1 111.7 0.3 5.3 -6.0
1988 LAC Woodson 8077.9 97.7 107.9 -10.3 -9.9 208.9 101.0 109.6 -8.6 -11.5 1.6
1988 MIL Cummings 7469.4 108.3 107.7 0.7 1.1 551.4 109.5 110.6 -1.1 -0.4 1.6
1988 PHO Davis 6826.0 107.6 112.1 -4.5 -4.8 1470.2 106.0 110.1 -4.1 -3.4 -1.4
1988 SAC Theus 7382.9 107.0 112.8 -5.9 -5.8 903.5 106.1 110.2 -4.1 -3.6 -2.2
1988 SAS Robertson 8586.4 108.5 113.1 -4.7 -4.5 0.0 0.0 0.0 0.0 0.0 0.0
1988 UTA Malone 8304.2 107.2 103.5 3.6 3.1 0.0 0.0 0.0 0.0 0.0 0.0
1988 WSB Malone 7899.3 106.8 107.6 -0.8 -0.3 202.9 105.0 106.0 -1.0 -5.5 5.2
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
1989 ATL Wilkins 7879.2 112.7 107.5 5.2 5.2 197.6 111.9 115.4 -3.5 4.9 0.3
1989 CHH Tripucka 7165.0 103.5 112.1 -8.6 -8.1 1078.3 106.7 114.6 -8.0 -6.6 -1.5
1989 DAL Aguirre 4279.7 110.1 108.5 1.6 1.0 3594.5 104.9 109.6 -4.7 -4.9 5.9
1989 DEN English 8876.7 109.0 107.4 1.6 1.0 0.0 0.0 0.0 0.0 0.0 0.0
1989 IND Person 7975.2 107.2 111.5 -4.3 -3.0 211.2 104.7 103.7 0.9 -7.8 4.8
1989 LAC Norman 8425.3 101.0 110.5 -9.6 -9.0 193.4 106.0 110.1 -4.1 -3.8 -5.2
1989 MIA Edwards 7854.6 98.0 109.6 -11.6 -11.3 300.6 106.1 109.8 -3.7 -6.3 -4.9
1989 MIL Cummings 7897.6 110.6 106.7 3.8 3.9 183.3 109.7 112.9 -3.3 4.3 -0.4
1989 NJN Hinson 8226.5 103.4 109.7 -6.3 -5.9 0.0 0.0 0.0 0.0 0.0 0.0
1989 SAC Smith 8154.3 104.8 110.2 -5.4 -5.1 99.0 105.0 118.2 -13.1 -5.6 0.5
1989 SAS Anderson 8447.2 101.3 108.2 -6.9 -7.0 98.3 94.6 108.9 -14.2 -5.3 -1.8
1989 WSB Malone 7734.7 106.5 108.9 -2.4 -2.2 588.3 109.1 107.8 1.4 2.5 -4.6
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
1990 CHH Gilliam 5890.3 103.7 111.0 -7.2 -6.2 2149.5 98.7 108.7 -10.0 -10.0 3.9
1990 DEN Lever 8352.2 108.1 106.9 1.2 1.3 321.3 113.6 109.9 3.7 4.6 -3.3
1990 DET Thomas 7648.5 110.7 104.0 6.7 6.2 95.8 96.0 108.5 -12.5 -11.7 17.8
1990 HOU Olajuwon 8310.5 105.3 103.9 1.4 1.6 0.0 0.0 0.0 0.0 0.0 0.0
1990 MIA Seikaly 7478.6 99.8 109.0 -9.2 -9.3 769.7 101.3 115.6 -14.3 -11.5 2.3
1990 MIN Campbell 7470.9 104.4 109.1 -4.6 -4.1 0.0 0.0 0.0 0.0 0.0 0.0
1990 NJN Hopson 7824.2 101.2 109.6 -8.4 -8.2 280.9 103.9 98.6 5.3 2.4 -10.7
1990 ORL Catledge 7744.3 106.4 115.3 -8.9 -9.1 823.3 103.0 108.3 -5.3 0.3 -9.4
1990 WSB Malone 7476.3 108.0 110.0 -2.0 -2.1 707.3 106.9 110.6 -3.7 -4.6 2.5
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
1991 BOS Bird 5977.0 114.2 106.2 8.0 7.5 2121.5 109.3 109.3 0.0 -0.7 8.2
1991 CHH Newman 7890.6 105.4 110.7 -5.3 -5.0 99.8 114.3 127.3 -13.0 -8.4 3.4
1991 DEN Adams 7458.6 106.1 114.1 -8.0 -8.0 1860.6 103.0 118.8 -15.9 -13.3 5.3
1991 LAL Worthy 7355.3 112.5 106.0 6.4 6.5 380.6 116.9 96.2 20.8 17.4 -10.9
1991 MIN Campbell 7134.8 107.6 111.5 -3.8 -3.7 474.7 103.4 113.5 -10.1 -9.7 6.0
1991 NJN Theus 8074.7 103.1 107.9 -4.8 -4.9 97.9 114.4 97.1 17.4 16.2 -21.1
1991 NYK Ewing 7761.8 107.7 108.2 -0.4 -0.5 87.5 105.1 89.1 16.0 4.9 -5.5
1991 WSB King 6340.2 103.6 108.5 -4.9 -4.5 1688.8 103.2 108.9 -5.7 -6.6 2.1
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
1992 CHH Gill 8044.0 107.3 111.8 -4.5 -4.0 305.8 114.1 101.4 12.8 10.8 -14.8
1992 DEN Williams 7972.6 101.3 109.2 -7.9 -7.5 103.0 99.0 113.6 -14.6 -23.3 15.9
1992 LAL Worthy 4995.6 108.0 109.6 -1.6 -1.7 2600.1 108.9 109.3 -0.4 0.3 -2.1
1992 MIN Campbell 7387.7 106.1 113.1 -7.0 -6.6 404.7 97.6 112.7 -15.1 -17.7 11.1
1992 ORL Anderson 5851.9 104.7 111.0 -6.2 -6.1 2163.5 101.7 111.1 -9.4 -8.5 2.4
1992 SAC Richmond 7975.0 104.9 110.4 -5.5 -5.0 220.0 84.1 110.5 -26.4 -24.0 18.9
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
1993 BOS Lewis 7586.5 108.9 108.6 0.4 0.4 190.0 124.7 101.0 23.7 20.8 -20.3
1993 DAL Harper 6193.6 100.7 115.2 -14.5 -14.6 1940.2 98.2 116.2 -18.0 -15.7 1.1
1993 DEN Abdul-Rauf 8111.0 105.2 106.8 -1.6 -2.1 92.2 103.0 116.0 -13.0 -9.7 7.6
1993 MIA Rice 7837.7 108.4 109.6 -1.2 -1.0 0.0 0.0 0.0 0.0 0.0 0.0
1993 MIL Brickowski 6348.0 106.2 110.0 -3.8 -3.3 1515.9 109.0 113.2 -4.2 -4.2 0.9
1993 NYK Ewing 7712.1 106.6 100.1 6.5 6.2 97.1 108.1 104.0 4.1 -0.8 7.0
1993 POR Drexler 4894.8 108.9 106.2 2.7 2.7 3281.5 108.7 105.0 3.7 3.4 -0.7
1993 WSB Grant 7040.1 105.1 111.6 -6.5 -6.0 976.8 98.0 109.8 -11.9 -11.5 5.5
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
1994 ATL Wilkins 4612.9 109.9 103.5 6.4 6.1 3100.0 104.8 100.4 4.5 3.9 2.2
1994 DAL Jackson 7719.3 101.1 110.3 -9.2 -8.7 0.0 0.0 0.0 0.0 0.0 0.0
1994 DEN Abdul-Rauf 7671.8 104.5 103.2 1.3 1.4 192.2 105.6 93.1 12.5 8.5 -7.1
1994 LAC Manning 4193.0 101.1 106.4 -5.3 -5.7 3967.3 106.1 112.2 -6.2 -4.8 -0.9
1994 MIN Laettner 6645.8 102.1 110.1 -8.0 -7.1 1088.1 105.1 108.4 -3.3 -6.1 -1.0
1994 PHI Weatherspoon 7852.6 102.3 110.3 -8.0 -7.7 0.0 0.0 0.0 0.0 0.0 0.0
1994 POR Robinson 8085.7 108.8 106.1 2.7 2.7 0.0 0.0 0.0 0.0 0.0 0.0
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
1995 ATL Blaylock 7206.1 107.1 105.9 1.3 1.3 186.9 107.0 100.6 6.4 0.7 0.5
1995 BOS Wilkins 7226.1 109.2 111.3 -2.1 -2.3 488.4 110.2 110.4 -0.2 2.7 -5.0
1995 DEN Abdul-Rauf 6737.3 109.2 109.3 -0.1 0.3 834.2 114.5 105.0 9.5 5.6 -5.3
1995 DET Hill 6513.5 105.9 113.4 -7.5 -7.3 1110.0 103.8 113.7 -9.9 -9.3 2.0
1995 GSW Sprewell 6819.1 107.6 112.7 -5.2 -4.7 1269.1 105.0 112.2 -7.2 -7.6 2.9
1995 LAC Vaught 7575.7 102.0 111.6 -9.6 -8.9 189.0 104.2 117.4 -13.2 -14.5 5.6
1995 MIL Robinson 7440.5 106.7 111.6 -4.9 -4.7 185.6 110.4 106.1 4.3 2.5 -7.2
1995 MIN Rider 6834.4 103.1 113.2 -10.1 -9.4 639.5 106.5 114.0 -7.5 -6.7 -2.7
1995 NJN Coleman 5229.4 107.5 110.0 -2.5 -2.1 2368.3 102.3 107.6 -5.3 -6.6 4.4
1995 NYK Ewing 7194.8 107.7 104.6 3.1 2.8 273.3 111.2 98.8 12.4 13.6 -10.8
1995 POR Robinson 7047.9 110.5 105.7 4.8 4.5 617.7 107.5 111.9 -4.4 -1.1 5.6
1995 WSB Webber 5096.9 105.9 112.7 -6.8 -6.7 2681.4 106.1 110.4 -4.2 -4.0 -2.7
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
1996 BOS Radja 5211.8 105.9 110.4 -4.4 -4.3 2730.0 108.9 110.7 -1.8 -2.4 -1.9
1996 DAL Jackson 7889.0 106.6 111.7 -5.1 -4.9 0.0 0.0 0.0 0.0 0.0 0.0
1996 DEN Abdul-Rauf 5300.6 106.0 110.0 -4.0 -3.1 2266.5 105.6 106.0 -0.4 -2.3 -0.8
1996 DET Hill 7017.2 109.1 106.0 3.1 2.9 176.9 93.8 101.7 -7.9 1.4 1.5
1996 GSW Sprewell 7268.6 109.2 111.0 -1.8 -1.7 372.7 107.1 103.8 3.2 2.0 -3.7
1996 MIL Baker 7367.8 106.4 112.3 -5.9 -5.5 0.0 0.0 0.0 0.0 0.0 0.0
1996 MIN Rider 6999.3 104.3 111.1 -6.8 -6.0 662.7 109.2 103.5 5.7 -1.1 -4.9
1996 NJN Gilliam 7117.6 103.0 107.2 -4.3 -4.2 362.8 97.9 109.7 -11.9 -11.5 7.3
1996 NYK Ewing 6897.9 106.6 103.3 3.3 3.3 568.3 108.2 115.4 -7.2 -6.3 9.6
1996 PHI Stackhouse 6650.5 103.1 113.5 -10.5 -10.2 884.2 100.8 114.8 -14.0 -11.8 1.6
1996 POR Robinson 7295.5 106.7 104.0 2.7 2.5 342.9 105.3 106.7 -1.5 0.7 1.8
1996 TOR Stoudamire 6486.1 105.0 111.8 -6.8 -6.1 1154.4 102.3 117.7 -15.4 -17.9 11.8
1996 VAN Anthony 6290.7 99.2 110.0 -10.9 -10.7 1196.0 94.0 105.3 -11.3 -10.1 -0.6
1996 WSB Howard 7602.3 109.4 108.3 1.1 1.1 88.8 104.7 97.9 6.8 -1.8 2.9
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
1997 BOS Walker 7895.4 104.5 112.1 -7.6 -7.0 0.0 0.0 0.0 0.0 0.0 0.0
1997 CLE Brandon 6480.7 105.9 103.2 2.7 3.1 312.2 98.7 106.7 -8.0 -6.6 9.7
1997 DEN Ellis 5072.4 106.8 113.5 -6.7 -7.0 2544.1 102.3 109.1 -6.8 -6.4 -0.7
1997 HOU Olajuwon 7165.2 109.3 104.3 5.0 4.3 369.9 111.9 109.5 2.4 3.4 1.0
1997 LAC Vaught 7526.5 105.9 108.4 -2.6 -2.8 0.0 0.0 0.0 0.0 0.0 0.0
1997 MIA Hardaway 7134.7 107.6 101.4 6.1 6.1 88.6 115.1 99.3 15.8 19.6 -13.4
1997 MIL Robinson 7158.0 106.8 108.7 -1.9 -1.6 159.2 109.9 122.5 -12.6 -6.2 4.7
1997 MIN Gugliotta 7332.9 106.2 107.8 -1.6 -1.9 90.3 104.2 111.9 -7.8 0.7 -2.7
1997 NJN Gill 7560.6 105.5 110.4 -4.9 -4.3 0.0 0.0 0.0 0.0 0.0 0.0
1997 NYK Ewing 7084.4 105.2 101.9 3.3 3.7 342.6 106.0 100.1 5.8 3.9 -0.2
1997 PHI Iverson 7277.9 105.9 112.5 -6.7 -6.5 526.2 97.1 106.8 -9.7 -3.2 -3.4
1997 SAS Wilkins 5454.6 105.5 114.3 -8.8 -9.7 1667.6 99.7 109.6 -9.9 -6.8 -2.9
1997 TOR Stoudamire 7334.1 105.4 108.8 -3.4 -2.9 98.4 99.6 107.7 -8.1 -3.6 0.7
1997 VAN Abdur-Rahim 7051.1 101.3 112.8 -11.6 -11.6 179.2 95.4 108.8 -13.4 -8.4 -3.1
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
1998 BOS Walker 7604.8 103.4 106.2 -2.8 -2.2 0.0 0.0 0.0 0.0 0.0 0.0
1998 CLE Kemp 7184.8 103.1 99.7 3.4 3.5 183.2 97.7 108.6 -10.9 -7.3 10.9
1998 DAL Finley 7396.5 101.3 108.1 -6.8 -6.9 0.0 0.0 0.0 0.0 0.0 0.0
1998 DET Hill 7177.2 106.2 104.6 1.7 2.0 90.7 105.8 97.0 8.8 6.7 -4.7
1998 GSW Smith 4524.5 95.0 106.7 -11.7 -11.2 2988.2 98.3 105.6 -7.3 -8.0 -3.2
1998 MIL Robinson 5058.1 104.9 104.9 0.0 0.2 2373.2 102.9 109.4 -6.5 -5.6 5.8
1998 NJN Van Horn 5655.2 110.9 109.0 1.9 1.7 1833.1 103.6 102.5 1.1 2.6 -0.9
1998 NYK Houston 7237.5 103.8 101.0 2.8 3.1 0.0 0.0 0.0 0.0 0.0 0.0
1998 PHO Chapman 6258.6 107.7 102.8 4.9 4.8 1288.1 110.8 101.7 9.1 6.3 -1.4
1998 POR Rider 6654.9 104.7 103.3 1.4 0.7 739.2 103.4 100.4 3.0 5.4 -4.8
1998 TOR Stoudamire 4477.9 101.6 109.7 -8.1 -7.3 3169.9 102.0 114.5 -12.6 -11.5 4.1
1998 WAS Webber 6518.9 106.3 105.1 1.2 1.9 1008.7 103.3 106.0 -2.7 -3.9 5.7
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
1999 BOS Walker 3850.7 101.0 104.0 -3.0 -3.5 738.1 103.0 99.9 3.1 3.2 -6.7
1999 CHI Kukoc 3854.7 94.1 103.5 -9.3 -9.0 542.7 86.1 106.9 -20.8 -18.9 9.9
1999 GSW Starks 4452.0 99.2 101.9 -2.7 -2.6 0.0 0.0 0.0 0.0 0.0 0.0
1999 LAC Taylor 4142.2 101.4 110.8 -9.4 -9.3 345.7 92.6 107.6 -15.0 -14.5 5.2
1999 MIN Garnett 4273.1 102.6 102.0 0.6 0.5 261.3 100.7 103.7 -3.1 -2.7 3.2
1999 NYK Ewing 3309.9 98.5 97.9 0.5 0.6 1034.2 102.6 99.4 3.2 3.7 -3.2
1999 ORL Hardaway 4430.6 101.0 98.0 2.9 3.2 0.0 0.0 0.0 0.0 0.0 0.0
1999 PHI Iverson 4273.5 101.0 98.1 2.9 3.3 170.5 97.3 109.7 -12.3 -14.5 17.8
1999 POR Rider 4236.4 105.5 98.5 7.0 6.7 260.2 104.5 96.8 7.7 8.5 -1.8
1999 SAC Webber 4071.0 103.6 103.8 -0.1 0.0 779.8 101.3 102.8 -1.5 -3.1 3.1
1999 WAS Richmond 4470.2 102.0 104.5 -2.5 -2.3 0.0 0.0 0.0 0.0 0.0 0.0
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
2000 ATL Rider 5496.7 104.3 107.7 -3.4 -3.6 2039.2 98.2 110.5 -12.4 -11.8 8.2
2000 BOS Walker 7729.5 105.4 106.2 -0.8 -1.0 0.0 0.0 0.0 0.0 0.0 0.0
2000 CLE Kemp 7865.0 101.1 104.7 -3.6 -3.8 0.0 0.0 0.0 0.0 0.0 0.0
2000 DAL Finley 7758.3 107.2 107.8 -0.6 -0.4 0.0 0.0 0.0 0.0 0.0 0.0
2000 DEN McDyess 7692.9 104.0 106.4 -2.4 -2.1 98.2 118.1 108.0 10.2 9.3 -11.4
2000 GSW Jamison 4078.3 100.8 108.9 -8.1 -8.0 3730.3 99.8 109.1 -9.3 -8.2 0.2
2000 LAC Taylor 5804.6 97.0 110.2 -13.1 -12.6 1859.8 102.9 112.8 -9.8 -8.7 -3.9
2000 PHI Iverson 6523.2 103.4 101.7 1.8 1.6 1082.4 94.7 95.2 -0.5 -1.9 3.5
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
2001 CHH Mashburn 6804.8 103.0 100.7 2.3 2.0 543.2 97.2 94.6 2.6 -1.1 3.1
2001 GSW Jamison 7707.0 98.4 108.0 -9.6 -8.8 0.0 0.0 0.0 0.0 0.0 0.0
2001 LAC Odom 6888.7 102.3 104.9 -2.6 -2.3 527.0 101.3 111.8 -10.4 -7.2 4.9
2001 SAC Webber 6720.9 106.0 99.9 6.2 6.5 1124.7 108.2 102.7 5.5 4.7 1.8
2001 WAS Howard 4958.4 99.8 107.0 -7.2 -6.8 2586.1 104.3 111.7 -7.3 -8.3 1.5
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
2002 CHH Davis 7324.6 105.1 104.0 1.1 0.7 0.0 0.0 0.0 0.0 0.0 0.0
2002 DEN Van Exel 4028.4 103.6 109.7 -6.1 -5.3 3366.5 100.5 107.4 -6.9 -6.7 1.4
2002 DET Stackhouse 6779.3 105.5 103.6 1.9 1.9 538.6 108.8 100.3 8.5 2.3 -0.4
2002 GSW Jamison 7741.5 103.5 109.2 -5.7 -5.2 0.0 0.0 0.0 0.0 0.0 0.0
2002 IND O'Neal 6598.7 104.9 104.3 0.6 0.3 968.9 104.6 106.6 -2.1 -1.4 1.7
2002 NJN Martin 6729.1 104.9 100.2 4.7 4.2 806.8 103.1 100.1 3.0 3.6 0.6
2002 NYK Houston 6927.3 102.3 106.8 -4.5 -4.6 435.7 98.7 102.4 -3.7 -4.0 -0.6
2002 PHI Iverson 5373.9 104.2 100.2 4.0 3.8 1860.7 100.1 104.6 -4.5 -4.9 8.6
2002 PHO Marbury 7499.1 104.0 104.8 -0.7 -0.4 0.0 0.0 0.0 0.0 0.0 0.0
2002 TOR Carter 5351.2 103.8 104.6 -0.8 -0.5 1894.3 102.3 102.0 0.3 -1.5 1.0
2002 WAS Jordan 5276.9 105.3 106.0 -0.7 -1.1 1934.4 106.0 110.2 -4.1 -3.4 2.4
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
2003 ATL Robinson 6322.1 103.4 108.4 -5.0 -5.1 1164.9 100.9 99.1 1.8 1.1 -6.2
2003 CHI Rose 7716.1 100.9 106.4 -5.5 -5.5 0.0 0.0 0.0 0.0 0.0 0.0
2003 CLE Davis 7419.5 97.4 107.7 -10.3 -10.1 282.0 95.0 105.0 -9.9 -13.4 3.3
2003 DEN Howard 6969.7 93.4 102.7 -9.3 -8.8 435.9 90.4 97.3 -6.9 -2.9 -5.9
2003 NOH Mashburn 7360.7 104.6 102.2 2.4 1.8 0.0 0.0 0.0 0.0 0.0 0.0
2003 PHI Iverson 7511.6 105.7 103.2 2.5 2.0 0.0 0.0 0.0 0.0 0.0 0.0
2003 SAC Webber 6427.1 106.7 100.2 6.5 6.8 1409.7 105.3 97.1 8.2 8.0 -1.2
2003 SEA Payton 4614.2 102.9 105.1 -2.2 -1.6 2614.8 107.4 103.9 3.6 3.5 -5.1
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
2004 BOS Pierce 7398.2 102.7 104.3 -1.6 -1.9 194.1 110.3 111.3 -1.0 -7.5 5.6
2004 CHI Crawford 7362.2 97.6 104.4 -6.8 -7.0 184.4 93.3 104.6 -11.4 -13.9 6.9
2004 CLE James 7203.1 101.8 104.9 -3.1 -3.3 260.6 108.6 105.9 2.7 -0.5 -2.8
2004 DEN Anthony 7605.9 104.8 103.7 1.2 1.6 0.0 0.0 0.0 0.0 0.0 0.0
2004 GSW Richardson 7009.9 104.1 104.9 -0.8 -0.3 368.8 94.9 96.8 -1.9 -0.5 0.2
2004 IND O'Neal 6799.0 104.5 98.3 6.3 5.8 353.7 109.1 96.4 12.7 10.0 -4.3
2004 NOH Davis 5988.1 103.8 102.4 1.4 0.5 1297.9 101.2 108.3 -7.1 -6.1 6.7
2004 PHI Iverson 4310.3 100.3 103.4 -3.2 -3.4 2891.0 100.1 102.4 -2.4 -2.8 -0.6
2004 PHO Stoudemire 5061.9 103.2 106.7 -3.5 -3.5 2492.8 100.3 105.4 -5.1 -3.2 -0.2
2004 POR Randolph 7027.5 104.9 106.1 -1.2 -0.6 82.9 85.6 109.8 -24.1 -25.5 24.9
2004 TOR Carter 6379.2 98.8 101.1 -2.4 -2.6 776.5 90.7 103.2 -12.5 -13.0 10.3
2004 WAS Arenas 5077.2 101.1 107.1 -6.0 -6.7 2484.3 96.4 102.8 -6.3 -6.3 -0.4
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
2005 ATL Walker 4833.3 99.5 109.6 -10.2 -9.6 2676.6 104.5 115.9 -11.4 -12.0 2.4
2005 CHA Okafor 6771.1 101.7 108.2 -6.5 -6.8 817.3 103.2 109.3 -6.1 -5.1 -1.7
2005 CHI Curry 5793.8 102.1 101.1 0.9 0.9 1793.7 102.1 100.2 1.9 0.5 0.4
2005 DEN Anthony 7023.1 107.1 104.2 2.9 3.0 622.6 102.8 108.7 -5.9 -5.6 8.5
2005 DET Hamilton 6679.8 107.0 101.7 5.3 4.8 506.9 99.6 107.1 -7.5 -7.3 12.1
2005 GSW Richardson 6783.6 106.0 107.5 -1.5 -1.3 903.3 99.6 108.2 -8.5 -6.9 5.6
2005 HOU McGrady 6932.0 107.2 102.8 4.4 4.7 348.7 104.7 96.4 8.3 4.0 0.8
2005 IND O'Neal 3846.1 105.6 105.3 0.3 0.9 3324.4 107.2 105.7 1.5 0.5 0.4
2005 NOH Nailon 5969.0 100.2 108.7 -8.5 -7.7 1231.7 103.0 109.0 -5.9 -6.1 -1.6
2005 ORL Francis 7345.7 105.8 107.9 -2.2 -2.3 384.1 101.8 108.0 -6.2 -9.6 7.3
2005 PHI Iverson 7154.1 104.2 104.7 -0.4 -1.2 660.7 101.6 105.9 -4.4 0.0 -1.2
2005 POR Randolph 4069.4 103.8 106.9 -3.1 -2.8 3243.2 104.7 111.0 -6.3 -5.5 2.6
2005 SAC Webber 4285.6 109.6 108.4 1.2 1.9 3371.7 112.9 109.2 3.7 3.6 -1.7
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
2006 CHI Gordon 7472.9 104.7 103.9 0.8 0.6 191.8 103.2 106.3 -3.1 -0.4 1.1
2006 HOU McGrady 4167.8 104.9 103.3 1.7 1.5 3037.0 99.2 105.8 -6.6 -5.7 7.2
2006 IND O'Neal 4554.2 104.5 102.1 2.3 2.4 2764.3 106.4 104.6 1.8 1.2 1.2
2006 NYK Marbury 5544.0 105.6 111.9 -6.3 -6.6 1941.0 102.3 111.4 -9.2 -8.0 1.4
2006 ORL Francis 4077.9 105.2 107.8 -2.6 -3.0 3196.8 109.3 108.7 0.5 0.6 -3.5
2006 POR Randolph 6434.4 101.9 113.0 -11.1 -10.6 707.0 102.8 111.6 -8.8 -8.0 -2.6
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
2007 GSW Davis 6190.1 109.2 107.9 1.3 1.5 1933.6 102.2 107.7 -5.5 -5.1 6.7
2007 IND O'Neal 6366.0 104.2 105.5 -1.3 -1.7 1204.3 100.0 109.8 -9.8 -9.1 7.5
2007 NOK West 4727.3 109.3 108.7 0.6 0.8 2689.6 99.1 104.9 -5.8 -5.3 6.1
2007 NYK Curry 7418.2 106.2 109.6 -3.4 -3.5 94.1 123.3 102.0 21.3 16.5 -20.1
2007 POR Randolph 6044.9 106.1 110.5 -4.4 -4.4 1246.4 104.7 111.4 -6.7 -4.8 0.5
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
2008 ATL Johnson 7483.1 107.6 109.6 -2.0 -2.4 0.0 0.0 0.0 0.0 0.0 0.0
2008 GSW Davis 8103.8 112.4 110.1 2.2 2.3 0.0 0.0 0.0 0.0 0.0 0.0
2008 MEM Gay 7724.6 105.3 112.2 -6.9 -6.4 100.8 123.0 99.2 23.8 10.9 -17.3
2008 MIA Wade 4702.2 103.2 111.3 -8.0 -7.9 2697.6 97.7 110.0 -12.3 -12.1 4.2
2008 MIN Jefferson 7509.9 104.4 111.8 -7.4 -7.0 0.0 0.0 0.0 0.0 0.0 0.0
2008 NYK Crawford 7363.2 105.3 112.3 -7.0 -6.7 182.7 107.9 120.4 -12.6 -18.9 12.1
2008 POR Roy 6546.9 108.8 108.9 -0.1 -0.1 693.0 100.6 111.3 -10.7 -6.7 6.7
2008 SEA Durant 7715.2 101.1 110.4 -9.3 -8.5 189.4 104.5 105.1 -0.5 -7.8 -0.6
2008 WAS Jamison 7119.5 109.7 110.2 -0.5 -0.7 248.9 116.9 114.5 2.4 1.6 -2.2
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
2009 ATL Johnson 7032.8 110.3 108.6 1.8 1.9 267.2 107.0 105.2 1.9 1.5 0.4
2009 DEN Anthony 6193.5 111.8 108.0 3.9 3.4 1509.9 107.9 105.2 2.7 2.7 0.7
2009 DET Hamilton 5865.1 108.4 108.9 -0.6 -0.7 1279.1 107.3 107.7 -0.5 0.5 -1.2
2009 GSW Jackson 5811.6 110.4 113.5 -3.1 -2.9 2281.8 109.1 114.7 -5.6 -6.5 3.6
2009 LAC Thornton 6549.9 103.7 112.9 -9.2 -8.6 1025.8 97.9 109.2 -11.3 -12.5 3.9
2009 MEM Gay 7106.1 104.8 111.2 -6.4 -6.1 259.9 96.2 92.7 3.5 1.7 -7.8
2009 MIN Jefferson 4614.5 106.9 110.9 -4.0 -3.4 2897.2 106.5 114.1 -7.6 -8.0 4.6
2009 NYK Harrington 6553.6 109.3 111.9 -2.6 -1.8 1384.1 105.6 108.9 -3.3 -5.8 4.0
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
2010 ATL Johnson 6870.2 112.8 107.7 5.1 4.9 529.7 111.6 105.2 6.4 5.5 -0.5
2010 CHA Jackson 6535.5 106.8 104.0 2.8 2.5 883.1 94.2 101.2 -7.0 -6.7 9.2
2010 CHI Rose 7294.0 104.3 105.7 -1.4 -1.6 375.2 102.1 111.1 -9.1 -4.4 2.8
2010 DET Hamilton 4106.9 106.7 113.1 -6.5 -7.3 3130.0 106.3 111.2 -4.9 -3.5 -3.8
2010 GSW Ellis 6432.4 108.1 112.5 -4.4 -4.3 1780.2 110.6 111.2 -0.6 0.1 -4.4
2010 LAC Kaman 7003.5 103.9 110.0 -6.1 -6.0 543.6 104.7 122.3 -17.7 -14.4 8.4
2010 MIL Bogut 6412.9 105.7 103.2 2.5 1.9 1170.2 105.2 107.0 -1.8 -0.4 2.4
2010 MIN Jefferson 7293.1 102.6 112.6 -10.0 -9.4 572.9 98.8 108.7 -9.9 -10.6 1.1
2010 PHI Iguodala 7489.4 107.0 111.3 -4.3 -4.3 0.0 0.0 0.0 0.0 0.0 0.0
2010 SAC Evans 6789.4 106.4 110.9 -4.6 -4.6 944.0 103.5 108.7 -5.2 -3.4 -1.2
2010 WAS Butler 4343.4 104.9 109.8 -4.8 -4.3 3163.8 105.4 111.2 -5.8 -6.4 2.1
Yr Tm Scorer Poss ORtg DRtg ED EDvAv Poss ORtg DRtg ED EDvAv Diff
2011 ATL Johnson 6382.1 107.4 108.4 -1.0 -1.5 895.7 104.5 104.9 -0.4 1.1 -2.6
2011 CHA Jackson 6011.3 104.5 108.1 -3.6 -3.8 1323.8 103.2 111.6 -8.5 -7.7 3.9
2011 CLE Jamison 5240.8 101.7 113.2 -11.5 -10.6 2377.4 104.9 110.8 -5.8 -7.1 -3.5
2011 DEN Anthony 4773.4 113.3 110.3 3.0 2.6 3032.7 112.2 104.1 8.1 8.7 -6.2
2011 GSW Ellis 7583.4 108.9 111.4 -2.5 -2.3 194.6 113.6 113.0 0.5 4.2 -6.5
2011 MIL Jennings 5661.2 102.2 103.1 -0.8 -1.6 1693.8 103.1 104.3 -1.2 1.0 -2.6
2011 SAC Evans 5366.9 103.4 107.4 -4.1 -4.2 2459.3 105.9 114.8 -8.9 -7.2 3.0
2011 TOR Bargnani 6159.2 107.6 114.2 -6.6 -6.9 1444.0 103.7 111.4 -7.7 -6.0 -0.9
2011 WAS Young 6029.0 103.0 111.6 -8.6 -8.3 1705.9 103.8 108.9 -5.2 -5.8 -2.5
Total 105.3 107.7 -2.5 -2.4 103.9 108.2 -4.4 -4.1 1.2

As you can see, teams were worse by 1.2 points of efficiency differential when their inefficient leading scorer didn't play, suggesting that scoring, even at below-average rates of efficiency, is difficult to replace. There are a number of alternative hypotheses, of course, so this should not be taken as ironclad, but it is a strong piece of evidence in favor of those who say it is hard to compensate for the loss of any high-usage player.

ShareThis

33 Responses to “Losing Your (Inefficient) Leading Scorer Hurts Your Team”

  1. Nathan Walker Says:

    I wonder if David Berri thinks that the Bulls were better with Kerr than Jordan. EFFICIENCY Lulz.

  2. David Says:

    Hi Neil, excellent to see these tweaks added. 3 things caught my eye.

    The 1.2 value is statistically distinct from zero. 95% CI is 1.2 +/- 2 * 0.3852 using bootstrap.

    Not sure if this is meant to debunk the Berri stuff, which replaced the scorer with an average player. I realize that is a philosophical issue too, no need to revist. My question here is if there is anyway to do something similar (replace chucker with avg player) here using, say, WinShares? Another way to put this: is there some link between where a team is in your Diff column and the player that got the minutes when the scorer was out? You have HCA, SOS (I think) in this mix so this seems like a logical extension. Although I'll admit to not being sure if it would yield anything. But I'm curious if we aren't seeing some strong vs. weak bench thing.

    You defined lead scores relative to the league average. Would not position average be more useful? I'm just not sure a guard should be compared to a center in this regard. I guess I was surprised by some of the names there, like James in 2004 or Wade in 2008.

  3. mystic Says:

    Great job, Neil. You just made my unfinished work about those "inefficient" scorers useless. I identified 57 cases since 2002-03 which I wanted to look into not only by using normal boxscore stats, but also by the +/- numbers provided by 82games.com. Adding the HCA term is really important here, I also added a term for SOS by using the average SRS values for the opponents to make the adjustment for different strong opponents during the stretch with and without the scorer. Overall I come to a similar conclusion as you (granted, I only collected the data for 36 cases so far). The interesting thing is that while the correlation coefficient to the expected point differential (using the Net+/- numbers for the players in games they played to calculate that) is at around 0.7, the teams are playing like 2pts better than expected, but they have still close to 1.5 pts worse point differential than the expected value with the scorer. When I add the more efficient scorers to the mix, the value goes up to 2.5 so far.

    So, the overall conclusion is the same. Even replacing the inefficient scorer isn't that easily achievable.

    David, we have to keep in mind that the scoring load is not replaced by the 7th or 8th man on that roster who gets more playing time due to the loss of that main scorer. The 2nd option will become the 1st option, the 3rd option the 2nd option and so on. In average the role of the main scorer will not be replaced by a well below average guy, but most times by a guy who should be an above average scorer/player anyway.

  4. AHL Says:

    At the very least, this shows why Denver was so much better off without Melo, and Memphis without Rudy Gay.

  5. ElGee Says:

    My comment turned into a supplementary post: http://www.backpicks.com/2011/05/23/more-on-low-efficiency-scorers-on-bad-offensive-teams/

    The short of it: I look at that data and see that
    (1) a lot of "inefficient" scorers played on bad offensive teams outside of the player in question
    (2) Those bad teams were helped by their "inefficient" scoring

  6. mystic Says:

    Elgee, going by your data would you agree that inefficient scorers tend to shoot more when the team surrounding them is worse offensively? While the higher usage in the end would also drag down their scoring efficiency further more?

  7. Jason J Says:

    It's interesting to see some of these names. I would not have expected Patrick Ewing to come up as an inefficient scorer so often. Also interesting to see which teams dropped off the face of the earth without his defensive presence and which ones survived.

    Two concerns come to mind on the results:

    1) Some of these guys were also their teams' best all-around players - like Ewing or McGrady or Wizards Jordan for example - so you lose the inefficient offense, but you also lose significant playmaking, rebounding, defense, and leadership (on court coordination of the team). I'm not sure if that overall 1.2 is really telling us about the importance of having inefficient high usage players or if it isn't just picking up on the fact that many of these guys bring more to the table than chucking (they'd almost have to).

    2) The usefulness of the findings might benefit some if the possessions played without the high usage player had a 200+ possession minimum or something in place to account for randomness. If a player misses one or two games worth of possessions, is the team's production over that span really telling us anything of note? Look at 2008 Rudy Gay. He missed one game that the Griz happened to play against the hapless Sonics before Durant figured out how to be Durant in this league. As a team they shot 13-16 from 3, and Bobby Jackson had a phenomenal game. Probably not a repeatable success story.

    http://www.basketball-reference.com/boxscores/200801180MEM.html

  8. DSMok1 Says:

    Wonderful research. Keep it up!

  9. Neil Paine Says:

    #7 - 2) is addressed by weighting the differences by possessions without the scorer (so a case like Gay's makes basically no impact on the overall average). 1) is an interesting point. I wonder how we could measure all-around play to eliminate situations like you mentioned, since efficiency is so intertwined with a metric like win shares or even PER. Perhaps look at non-scoring SPM?

  10. Jason J Says:

    #9 - Gotcha.

    How about factoring in versatility index?

  11. DSMok1 Says:

    I just put up a post with a chart visualizing this data. The slopes of the lines are interesting:

    http://godismyjudgeok.com/DStats/2011/nba-stats/chart-with-or-without-inefficient-scorers/

  12. ElGee Says:

    #6 - Mystic, I would absolutely agree.

    If the team can't apply on pressure on a defense, the "best" option is left with harder attempts, and those attempts are sill better than whatever the other guys could muster up without him. The usage is going up, driving efficiency is going down. Even though they're "inefficient" a lot of times they are helping the offense a good deal. (And they might not be inefficient if they were on a decent offensive club to being with.)

  13. DSMok1 Says:

    Okay, I revised and added to the last chart:

    http://godismyjudgeok.com/DStats/2011/nba-stats/chart-with-or-without-inefficient-scorers/

    Basically, If you're on a good offense, losing you hurts, period. If you're on a bad offense, it may not hurt.

    I think that's contrary to what Neil said...

  14. Greyberger Says:

    Boo versatility index. An idea who's time has come, and went.

  15. Michael E Sullivan Says:

    I'm not sure I buy your interpretation DSMok.

    Looking at your graph it seems the main thing pulling your offensive line down is that there are a few data points where the offense without the player was terrible, but with the player actually fairly good. This suggests it hurts a lot more. Basically to get at what you are asking, the appropriate question is what kind of spread is there between the best and worst with player data points for a give without player rating. The spread is much bigger on the low end than on the high end.

    The teams that were very good without the high scorer, generally did better with, but not a lot better with and they rarely did a lot worse with -- suggesting that NBA coaches and GMs are not stupid, and do not often let players take a ton of possessions when it would make their team significantly worse. This would be expected.

    So generally inefficient scorers are either on poor teams that don't have a lot of better options, or their high usage lets other players be more picky about when to take shots, and thus they are more efficient with than without the scorer around, meaning that these guys could be somewhat more valuable than raw efficiency suggests. Though, only somewhat, I notice that most of the guys with positive differentials brought some other stuff to the table.

  16. Matt, Colombia Says:

    I think this shows that WoW simply does not understand usage curves.

  17. Bob M. Says:

    This confirms that Jermaine O'neal was an offensive black hole for the Pacers in the mid 2000s. Good research.

  18. Anon Says:

    "Boo versatility index. An idea who's time has come, and went."

    Not a fan?

  19. Greyberger Says:

    I might not have a grasp on what the VI is measuring and what it means. The only context I've seen it in is coefficient tables for SPM, where it seems arbitrary and questionable to include it. To me

  20. marparker Says:

    Isn't it important to know who took over that player's minutes and what kind of usage/efficiency they were able to produce. Can we really lump in trades with injuries? All sorts of bases were left uncovered. This is a reach. It almost seems disingenuous.

  21. Rob P. Says:

    Seeing Worthy on the list for '91 and '92 made me wonder about him. A "high-ankle injury in the '91 playoffs" and a "season-ending knee injury in '92" (wiki) really seemed to alter his game. Interesting how much he trended toward taking shots from deep in his later years:

    In his first 6 seasons (471 games): 59 3PTA and 3 3PTM (.017 3PT%).
    In his final 6 seasons (455 games): 427 3PTA and 114 3PTM (.266 3PT%).

    Side note: I wondered who led L.A. in scoring (PPG) in Worthy's final two seasons:
    Sedale Threatt with 15.1 PPG (1992-1993) and Vlade Divac with 14.2 PPG (93-94)

    Threatt edged out Worthy by 14 points; meaning Worthy came really close to having another lousy, inefficient season appear on this list.

  22. huevonkiller Says:

    Wow this is some great research.

  23. Jason J Says:

    I actually think VI is a minor component of PER as well... that may be wrong... I do believe it is a Hollinger stat in any case.

    I suggested it because we're trying to isolate the inefficient scorers who also do a lot of other things, and VI measures other things without any concern for efficiency.

  24. Vjl110 Says:

    The fact that losing a starting player (who happens to be an inefficient scorer) lowers efficiency differential 1.2 points really isn’t interesting by itself. Absolutely no negative change is not an acceptable null hypothesis for this analysis. That said, there are ways you could get at answering the proposed question using this data-set.

    1)
    Try randomly selecting 100 starting players and see how losing those players affects efficiency differential. Now you have a control group which gives you a null value to compare against. If losing a random player is no different in effect than losing an inefficient scorer, that inefficient scoring isn’t necessarily important. If losing inefficient scorers hurts the team more than losing a random starter, you now have much better evidence for the import of inefficient scoring.

    2)
    Alternatively, build a regression model using some composite metric like WP48/WS/PER for each player, and another using shots/usage or some combination thereof. Then try using these two models to account for the team efficiency differentials that you found. The results of this study wouldn't be as clean as the one above, but they would still be interesting. If usage or shot attempts does not explain the strength of effect on efficiency differential within the above group of inefficient volume scorers, then I am not sure the study supports the value of inefficient scoring.

    This is a great study, but I don't feel like I have learned much until you at least find a better null value than "no effect of losing starting player".

  25. yariv Says:

    As #24 noted, and as discussed in the original post, the problem is comparing losing the leading scorer to losing any high MPG player (I would prefer "high MPG" over starter). Any intention to work on this?

  26. Owen Says:

    "I think this shows that WoW simply does not understand usage curves."

    Not that this will get answered, but I am a WOW type, and I have to admit, I don't get usage curves. The premise is obvious. The more you shoot the less efficient you should be. The less you shoot, the more efficient you should be.

    But to me it rarely works out as predicted. For instance, shouldn't Bosh have seen a huge bump in efficiency this year playing next to two stars? He was more efficient from the perspective of turnovers but his scoring efficiency actually went down to it's level of two years ago.

    You look at Kobe's numbers and the pattern is counterintuitive as well. He basically has been in a ts% band of 54.4% to 56.4% for most of his career (all but two years) despite changes in usage, shot attempts, teammates etc. His highest usage year (in his age 27 season granted) wasn't basically right where it should have been at 59.9%. If anything though his efficiency seemed to increase with higher usage (although adjusting for age probably would flatten things out).

    Clearly there is some relationship between usage and efficiency, but it strikes me as being very weak.

  27. Owen Says:

    "wasn't basically right where it should have been at 59.9%."

    should read

    "was basically right where it should have been at 55.9%."

  28. David Says:

    #26 - Neil did a recent post on usage curves. Well, not exactly, but he looked at how per minute stats change when minutes change. This, to me, is the heart of the matter. There really was no conclusive take home message. You can find players who get better or worse and also those that do not meaningful change. I've seen this elsewhere too. I also wonder where this conventional wisdom stuff come from. I know Hollinger likes this idea. But there is simply too much scatter for this idea to, say, be a parameter for player acquisition by the FO. Maybe I've missed _the_ study on this too.

    #24 - Yes, this is huge. A meaningful null should not be "no change". And I would still love to see the following: Replace every team's leading scorer by an average player using WinShares and simulate the current season. What does eff diff look like then? The conditional analysis Neil did certainly points in the right direction but there is too much variability in terms of how a team makes up for the missing chucker.

  29. Vjl110 Says:

    #28 - I think using Winshares or anything else to predict the affect of an "average" player would simply replicate Berri's method.
    The virtue of using a group of randomly selected high-mpg/starting players as a control. You don't need to make any assumptions about what features contribute to wins. This would be a really easy addition the Neil's study, and would give us a really good look at the value of "inefficient" scoring.

  30. Anon Says:

    "Not that this will get answered, but I am a WOW type, and I have to admit, I don't get usage curves. The premise is obvious. The more you shoot the less efficient you should be. The less you shoot, the more efficient you should be."

    This is based on some key assumptions, however (like running/playing an offense that maximizes your play, etc.). Not to mention that the James/Wade/Bosh trio is something that is unique in NBA history.

  31. Bill Says:

    How James/Wade/Bosh came together is unique, but three star players on one team is certainly not unique in NBA history. Did Larry Bird ever play on a team with fewer than three HOFers?

  32. David Says:

    #29 Agree. And I'd love to see this for above and below average players for PPG, REB, and AST. Or maybe by terciles. I think a larger context would help.

  33. Nathan Walker Says:

    Usage curves explain that a player's PERSONAL efficiency decreases when usage increases. Whether or not his TEAM's efficiency increases depends on the player, but we can predict it with their usage. Eli Witus found that lineups containing players with high-usage will fare better than lineups containing players with low-usage.

    The general trend I have found is that (Dean Oliver) efficiency * usage is the most significant descriptor (rather than separating the two) and produces the best predictions.