Setting Up Excel File for RAPM

[KwameBeanJordan]

Hey guys,

I've been trying to compute some RAPM, using Jacob Frankel's guide, but I am unsure how to set up the excel file for the computations. I've downloaded the 2009 regular season matchups and player ID's from basketballvalue, extracted them, and opened them in an excel file (tab delimited). Now, the guide says

Now we have to set up the independent variables. Each of the 400 players who play in our dataset is an independent variable and there’s a column for each independent variable. Download this sheet, that has a list of player IDs. Transpose that list along the top row of the initial sheet (should be J1:RC1) then fill this formula in cells J2:RC37264

=IF(ISNUMBER(FIND(” “&J$1&” “,$B2)),1,IF(ISNUMBER(FIND(” “&J$1&” “,$C2)),-1,0))

which I don't understand. It's written as if the last column in the data is I but it goes far beyond that. Though I'm using 2009 data, rather than 2011, the issue would remain the same. As such, the formula for the cell entries won't work. I understand what entries we want in each cell, but I am not familiar enough with excel to create them on my own, and I don't how the excel should be set up to be used in R. Anyways, I'd appreciate any help.

[permaximum]

I haven't read that guide but I just looked at the page roughly and saw "glmnet". That library produces wrong results because of that supposedly efficient method it uses to calculate ridge regression. Unfortunately I haven't seen any other library that includes weighted ridge regression calculation. So, with R, you're out of luck. Unless you can set weighted ridge regression calculation up yourself.

So, I suggest throwing that guide out of window before wondering about the formulas written in it.

[DSMok1]

Hey guys,

I've been trying to compute some RAPM, using Jacob Frankel's guide, but I am unsure how to set up the excel file for the computations. I've downloaded the 2009 regular season matchups and player ID's from basketballvalue, extracted them, and opened them in an excel file (tab delimited). Now, the guide says

Now we have to set up the independent variables. Each of the 400 players who play in our dataset is an independent variable and there’s a column for each independent variable. Download this sheet, that has a list of player IDs. Transpose that list along the top row of the initial sheet (should be J1:RC1) then fill this formula in cells J2:RC37264

=IF(ISNUMBER(FIND(” “&J$1&” “,$B2)),1,IF(ISNUMBER(FIND(” “&J$1&” “,$C2)),-1,0))

which I don't understand. It's written as if the last column in the data is I but it goes far beyond that. Though I'm using 2009 data, rather than 2011, the issue would remain the same. As such, the formula for the cell entries won't work. I understand what entries we want in each cell, but I am not familiar enough with excel to create them on my own, and I don't how the excel should be set up to be used in R. Anyways, I'd appreciate any help.

What this expression is doing is checking to see if the text (ID) located in cell J1 is found in the lineup cells B2 and C2. If it IS found in cell B2, the formula returns 1. If it is found in cell C2, it returns -1. If it is found in neither, the formula returns 0.

The $ used in J$1, $B2, and $C2 lock the row or column references so this expression can be copied freely and will translate appropriately, always looking at the ID row 1 at the top and the Lineup columns B and C.

[DSMok1]

I haven't read that guide but I just looked at the page roughly and saw "glmnet". That library produces wrong results because of that supposedly efficient method it uses to calculate ridge regression. Unfortunately I haven't seen any other library that includes weighted ridge regression calculation. So, with R, you're out of luck. Unless you can set weighted ridge regression calculation up yourself.

So, I suggest throwing that guide out of window before wondering about the formulas written in it.

Where are the issues with glmnet documented? I vaguely remember some issues with this.

[permaximum]

I haven't read that guide but I just looked at the page roughly and saw "glmnet". That library produces wrong results because of that supposedly efficient method it uses to calculate ridge regression. Unfortunately I haven't seen any other library that includes weighted ridge regression calculation. So, with R, you're out of luck. Unless you can set weighted ridge regression calculation up yourself.

So, I suggest throwing that guide out of window before wondering about the formulas written in it.

Where are the issues with glmnet documented? I vaguely remember some issues with this.

I don't know if it's documented or not but I experimented a lot with glmnet while I was calculating RAPM a couple of years ago and there's no doubt in my mind that's it's problematic. At least when you use weight function. If you try to calculate RAPM, you may see players with only a few possessions at the top or bottom in results, getting very far away from the prior without really suffering the ridge penalty or leaning against the prior.

I believe in one of my debates with mystic he acknowledged this issue with glmnet as well.

[KwameBeanJordan]

I haven't read that guide but I just looked at the page roughly and saw "glmnet". That library produces wrong results because of that supposedly efficient method it uses to calculate ridge regression. Unfortunately I haven't seen any other library that includes weighted ridge regression calculation. So, with R, you're out of luck. Unless you can set weighted ridge regression calculation up yourself.

So, I suggest throwing that guide out of window before wondering about the formulas written in it.

Where are the issues with glmnet documented? I vaguely remember some issues with this.

I think you are refering to this. I guess the glmnet package uses cyclical coordinate descent. That doesn't mean much to me, at least not yet, but maybe you two understand it.

What this expression is doing is checking to see if the text (ID) located in cell J1 is found in the lineup cells B2 and C2. If it IS found in cell B2, the formula returns 1. If it is found in cell C2, it returns -1. If it is found in neither, the formula returns 0.

The $ used in J$1, $B2, and $C2 lock the row or column references so this expression can be copied freely and will translate appropriately, always looking at the ID row 1 at the top and the Lineup columns B and C.

So, the ID's for all the home players are in a single cell and likewise for the away players? Currently, my columns go,

Game ID, Start Time, End Time, Time Elapsed, Elapsed Seconds, Home Player 1 ID, ..., Home Player 5 ID, Away Player 1 ID, ..., Away Player 5 ID, etc...

and I'll need to have Game ID through Elapsed Seconds all in Column A, the home players all in Column B, and the away players all in Column C.

Do you happen to know a quick fix? I don't think a different choice of delimiter would work since I wouldn't want to concatenate the ID's.

[DSMok1]

I haven't read that guide but I just looked at the page roughly and saw "glmnet". That library produces wrong results because of that supposedly efficient method it uses to calculate ridge regression. Unfortunately I haven't seen any other library that includes weighted ridge regression calculation. So, with R, you're out of luck. Unless you can set weighted ridge regression calculation up yourself.

So, I suggest throwing that guide out of window before wondering about the formulas written in it.

Where are the issues with glmnet documented? I vaguely remember some issues with this.

I think you are refering to this. I guess the glmnet package uses cyclical coordinate descent. That doesn't mean much to me, at least not yet, but maybe you two understand it.

What this expression is doing is checking to see if the text (ID) located in cell J1 is found in the lineup cells B2 and C2. If it IS found in cell B2, the formula returns 1. If it is found in cell C2, it returns -1. If it is found in neither, the formula returns 0.

The $ used in J$1, $B2, and $C2 lock the row or column references so this expression can be copied freely and will translate appropriately, always looking at the ID row 1 at the top and the Lineup columns B and C.

So, the ID's for all the home players are in a single cell and likewise for the away players? Currently, my columns go,

Game ID, Start Time, End Time, Time Elapsed, Elapsed Seconds, Home Player 1 ID, ..., Home Player 5 ID, Away Player 1 ID, ..., Away Player 5 ID, etc...

and I'll need to have Game ID through Elapsed Seconds all in Column A, the home players all in Column B, and the away players all in Column C.

Do you happen to know a quick fix? I don't think a different choice of delimiter would work since I wouldn't want to concatenate the ID's.

Yes, your current structure would work fine. Just do IF(OR(Z$1=$F1,Z$1=$G1,Z$1=$H1,Z$1=$I1,Z$1=$J1),1,IF(OR(Z$1=$K1,Z$1=$L1,Z$1=$M1,Z$1=$N1,Z$1=$O1),-1,0)). There may be a better construction for speed of calculation, but this method is constructed similarly to the last one.

[mystic]

I believe in one of my debates with mystic he acknowledged this issue with glmnet as well.

Yes, I had similar issues, but it turned out that the documentation just isn't really good. glmnet has standardize set to TRUE by default. When using glmnet for RAPM calculations someone should use standardize=FALSE instead. Also, it is important to use alpha=0.
Using standardize=FALSE solved most of the issues, but I still get different results in comparison to the results I get when solving the equation directly (and I'm extremely sure that the latter results are the correct results), but the differences are way smaller than before.

Haven't read the link in the OP, but Eli Witus explained how to get the data into a csv file via Excel here: http://www.countthebasket.com/blog/2008 ... lus-minus/

Maybe that is more helpful ...

[permaximum]

I believe in one of my debates with mystic he acknowledged this issue with glmnet as well.

Yes, I had similar issues, but it turned out that the documentation is just really good. glmnet has standardize set to TRUE by default. When using glmnet for RAPM calculations someone should use standardize=FALSE instead. Also, it is important to use alpha=0.
Using standardize=FALSE solved most of the issues, but I still get different results in comparison to the results I get when solving the equation directly (and I'm extremely sure that the latter results are the correct results), but the differences are clearly smaller than before.

Haven't read the link in the OP, but Eli Witus explained how to get the data into a csv file via Excel here: http://www.countthebasket.com/blog/2008 ... lus-minus/

Maybe that is more helpful ...

It's been a quite some time but I recall I "probably" used standardize=FALSE anyways. But I'm sure I always used alpha=0 because that was the ridge penalty. Between 0 and 1 you would get elastic-net and with 1 you would get lasso. Ridge was better than others by a small margin.

[mystic]

It's been a quite some time but I recall I "probably" used standardize=FALSE anyways.

I can't remember it correctly, but I think you posted some numbers and there were a couple of low possessions players on top of the list. I got similar results with glmnet and couldn't really figure out what the reason for that could be (well, lambda.min was really low, and I suggested using the bigger lambda.1se instead, that changed it a bit; I explicitely recommend now using lambda.min). Later someone pointed out that standardize is set to TRUE by default, and upon setting standardize=FALSE the results became pretty good (although still slightly different than solving it via matrix algebra).

That would be the results for the 2012 matchupfile from bbv. Poss is the amount of total possessions a player was on the court (off + def), RAPM is the result based on solving it via matrix algebra directly, FALSE is with standardize=FALSE, TRUE is without (sorted by):

Code: Select all

Name                          Poss    RAPM   FALSE    TRUE
N'Diaye, Hamady                   7    0.11    0.12   28.75
Benson, Keith                    31    0.27    0.31   18.30
Diogu, Ike                       53    0.29    0.34    9.26
James, Mike                     449    1.46    1.47    6.17
Williams, Sean                  417    1.03    1.20    5.76
Reid, Ryan                       63    0.11    0.10    5.00
Whiteside, Hassan               382    0.90    1.00    4.81
Magloire, Jamaal               1155    1.70    2.06    3.74
Gibson, Taj                    4504    4.88    5.30    3.59
Foote, Jeff                     150    0.22    0.29    3.52
Lopez, Brook                    143    0.20    0.31    3.41
Hobson, Darington               147    0.20    0.30    3.39
Lucas, John                    2520    3.10    3.15    3.38
Ginobili, Manu                 2817    3.18    3.22    3.33
Kapono, Jason                   975    1.18    1.36    3.19
Bonner, Matt                   4609    4.14    4.51    3.17
Fesenko, Kyrylo                  65    0.13    0.11    3.11
Foster, Jeff                    535    0.81    0.82    3.06
Udrih, Beno                    3737    3.64    3.99    2.99
Anderson, Alan                  704    1.21    1.18    2.98
Alabi, Solomon                  164    0.39    0.37    2.85
Smith, Jerry                    122    0.15    0.19    2.71
Uzoh, Ben                       635    1.14    1.06    2.64
Dampier, Erick                  257    0.30    0.45    2.63
Sanders, Larry                 2295    2.36    2.59    2.60
Curry, Stephen                 2878    3.12    3.10    2.58
Gallinari, Danilo              5185    3.84    4.05    2.57
Butler, Jimmy                  1266    1.51    1.59    2.50
Gordon, Eric                   1174    1.29    1.31    2.46
Macklin, Vernon                 428    0.59    0.68    2.45
Miller, Mike                   2607    2.25    2.44    2.45
Anderson, James                1846    1.80    1.96    2.41
Harden, James                  7181    3.69    4.09    2.37
Randolph, Zach                 2776    2.21    2.50    2.36
Harrellson, Josh               2016    1.67    1.91    2.36
Carter, Vince                  5679    3.44    3.61    2.26
Collison, Nick                 4776    2.38    2.57    2.19
Rose, Derrick                  5036    2.47    2.90    2.18
Nowitzki, Dirk                 7682    4.24    4.68    2.17
Parker, Tony                   6875    3.92    4.23    2.13
Wade, Dwyane                   6061    2.56    2.98    2.11
Hamilton, Jordan                967    0.99    1.18    2.05
James, LeBron                  8442    3.65    3.81    2.01
Arenas, Gilbert                 794    0.87    1.05    1.98
Lin, Jeremy                    3132    2.14    2.48    1.96
Battie, Tony                    993    0.77    0.79    1.94
Allen, Tony                    5780    2.61    2.92    1.92
Udoh, Ekpe                     4571    2.68    2.83    1.91
Granger, Danny                 7605    3.47    3.73    1.90
Paul, Chris                    8082    3.23    3.54    1.89
Garnett, Kevin                 6358    3.02    3.29    1.88
Griffin, Blake                 8861    3.79    4.05    1.87
Duncan, Tim                    5871    2.92    3.07    1.84
Bledsoe, Eric                  1705    1.14    1.42    1.83
Bosh, Chris                    7288    2.94    3.02    1.82
Haddadi, Hamed                  788    0.69    0.82    1.81
Anderson, Ryan                 6106    3.01    3.14    1.80
Hudson, Lester                 1074    0.87    0.99    1.76
Young, Thaddeus                5882    2.70    2.93    1.74
Chalmers, Mario                6502    2.06    2.05    1.73
Korver, Kyle                   5019    2.33    2.42    1.72
Butler, Caron                  6920    2.46    2.55    1.71
Sefolosha, Thabo               3352    1.62    1.89    1.67
Gasol, Marc                    8979    3.44    3.73    1.66
Stackhouse, Jerry               841    0.84    0.94    1.64
Novak, Steve                   3557    1.90    2.13    1.61
Maynor, Eric                    515    0.36    0.34    1.54
Horford, Al                    1132    0.55    0.74    1.51
Aldridge, LaMarcus             6745    2.69    2.87    1.51
Billups, Chauncey              2284    0.94    0.95    1.50
Deng, Luol                     7645    2.82    2.98    1.49
Frye, Channing                 6175    2.48    2.57    1.49
Anthony, Joel                  4808    1.37    1.41    1.49
Pondexter, Quincy              3804    1.88    2.02    1.46
Felton, Raymond                6272    2.12    2.31    1.44
Westbrook, Russell             8806    2.72    2.92    1.44
Bradley, Avery                 4496    1.70    1.98    1.44
Allen, Lavoy                   2201    0.85    0.94    1.43
Smith, J.R.                    3222    1.61    1.78    1.42
Mills, Patrick                  766    0.48    0.62    1.41
Martin, Cartier                 679    0.35    0.46    1.40
Radmanovic, Vladimir           2247    1.31    1.61    1.38
Conley, Mike                   8194    2.79    2.92    1.38
Brand, Elton                   5655    2.08    2.09    1.33
Green, Danny                   5339    1.84    1.96    1.31
Williams, Jordan                949    0.35    0.45    1.30
Smith, Josh                    6793    2.42    2.59    1.23
Hamilton, Richard              2349    0.82    1.06    1.23
Rubio, Ricky                   5380    2.33    2.31    1.22
Emmett, Andre                    97    0.06    0.07    1.20
West, David                    7060    1.67    1.77    1.17
George, Paul                   7251    1.84    1.92    1.16
Splitter, Tiago                4109    1.58    1.51    1.16
Thompson, Mychel                342    0.25    0.27    1.15
Kidd, Jason                    5113    1.56    1.73    1.15
Miller, Andre                  6759    1.93    2.13    1.14
Gasol, Pau                     8804    2.45    2.67    1.14
Meeks, Jodie                   5440    1.23    1.34    1.14
Pargo, Jannero                 2099    1.15    1.40    1.14
Nelson, Jameer                 5345    1.67    1.68    1.14
Howard, Dwight                 6866    2.14    2.16    1.13
Thomas, Kurt                   2706    1.06    1.19    1.11
Rondo, Rajon                   6942    1.81    2.05    1.11
Durant, Kevin                  9567    1.99    2.01    1.09
Hibbert, Roy                   7126    1.42    1.48    1.09
Nash, Steve                    7400    1.93    2.08    1.08
Millsap, Paul                  7235    2.34    2.49    1.07
Iguodala, Andre                7573    2.14    2.15    1.04
Williams, Deron                2557    0.73    0.94    1.04
Matthews, Wes                  7310    1.63    1.74    1.03
Fields, Landry                 6848    2.21    2.36    1.01
Aminu, Al-Farouq               5140    1.51    1.88    1.00
Richardson, Jason              4916    1.38    1.34    0.99
Williams, Louis                5631    1.49    1.50    0.99
Kleiza, Linas                  2316    1.04    1.18    0.98
Koufos, Kosta                  3075    1.15    1.22    0.96
Stone, Julyan                   688    0.60    0.57    0.96
Bogut, Andrew                  1389    0.71    0.82    0.92
Wright, Brandan                2865    0.51    0.65    0.91
Forbes, Gary                   1180    0.77    0.83    0.90
Jordan, DeAndre                6650    1.09    1.00    0.90
Robinson, Nate                 4417    1.22    1.46    0.89
Varejao, Anderson              2897    1.10    1.45    0.89
Ibaka, Serge                   6686    1.24    1.27    0.88
Balkman, Renaldo                452    0.17    0.26    0.87
Harrington, Al                 6684    1.49    1.55    0.87
Johnson, Amir                  3825    1.34    1.43    0.87
Humphries, Kris                3261    1.03    1.22    0.85
Bryant, Kobe                   8124    1.50    1.59    0.85
Brewer, Ronnie                 5747    1.19    1.20    0.84
Price, A.J.                    2098    0.89    1.00    0.83
Mahinmi, Ian                   4053    0.90    1.00    0.83
Battier, Shane                 5212    0.88    1.02    0.81
Teague, Jeff                   6190    1.47    1.46    0.80
Smith, Jason                   3331    1.34    1.45    0.78
Fisher, Derek                  5665    0.73    0.74    0.77
Barnes, Matt                   5193    0.88    0.98    0.76
Mbah a Moute, Luc              3458    1.29    1.37    0.76
Hill, George                   4554    0.94    1.05    0.75
Jeffries, Jared                2473    0.78    0.72    0.73
Haywood, Brendan               4159    0.86    0.75    0.73
Dalembert, Samuel              5243    1.19    1.15    0.72
Davis, Baron                   1994    0.59    0.67    0.72
Hill, Grant                    5192    0.83    0.80    0.71
Asik, Omer                     3331    0.70    0.60    0.69
Noah, Joakim                   6936    1.16    1.19    0.69
Jefferson, Al                  7277    1.30    1.58    0.68
Beaubois, Rodrigue             4168    0.71    0.73    0.67
Love, Kevin                    8228    1.85    2.08    0.67
Turkoglu, Hedo                 5340    0.71    0.68    0.66
Andersen, Chris                1856    0.38    0.35    0.66
Camby, Marcus                  4768    0.83    0.70    0.66
Johnson, Joe                   6022    1.06    1.01    0.64
Turiaf, Ronny                   983    0.23    0.20    0.60
Vasquez, Greivis               5923    1.20    1.30    0.60
Ebanks, Devin                  1404    0.28    0.27    0.59
Bayless, Jerryd                1430    0.27    0.57    0.58
Parsons, Chandler              6637    0.99    0.97    0.53
Holiday, Jrue                  7311    0.90    0.88    0.53
Dragic, Goran                  6283    1.05    1.15    0.52
Hilario, Nene                  3791    0.68    0.61    0.52
Bynum, Andrew                  7671    0.89    1.04    0.52
Dudley, Jared                  7541    0.74    0.68    0.48
Thomas, Isaiah                 6243    1.17    1.38    0.47
Cook, Daequan                  3692    0.31    0.28    0.47
Lee, Courtney                  6391    0.95    1.03    0.47
Pierce, Paul                   7185    0.77    0.67    0.46
Wallace, Gerald                6484    0.92    0.81    0.45
Henry, Xavier                  2592    0.30    0.37    0.45
Bass, Brandon                  6276    0.47    0.58    0.44
Moore, E'Twaun                 1068    0.39    0.30    0.43
Bargnani, Andrea               2885    0.41    0.59    0.41
Stuckey, Rodney                5265    1.49    1.60    0.41
Tinsley, Jamaal                1677    0.16    0.24    0.40
Gortat, Marcin                 7883    0.49    0.58    0.40
Ford, T.J.                      740    0.47    0.35    0.39
Scola, Luis                    7634    0.79    0.92    0.36
Shumpert, Iman                 6191    1.00    1.04    0.36
Boozer, Carlos                 6798    0.52    0.42    0.35
Lawson, Ty                     8269    0.63    0.80    0.35
Mason, Roger                   2229    0.39    0.50    0.34
Favors, Derrick                4754    0.72    0.83    0.32
Seraphin, Kevin                3200    0.51    0.70    0.29
Sessions, Ramon                5982    0.92    0.88    0.28
Lowry, Kyle                    5699    0.70    0.70    0.24
Bynum, Will                    1711    0.41    0.46    0.23
Anthony, Carmelo               6568    0.64    0.75    0.23
Dunleavy, Mike                 4868    0.76    0.48    0.23
Collison, Darren               7072   -0.11   -0.16    0.23
Moore, Mikki                    298    0.07    0.09    0.22
McGrady, Tracy                 2444    0.43    0.44    0.22
Smith, Ishmael                  613   -0.13   -0.20    0.22
Ilyasova, Ersan                5842    1.22    1.12    0.20
Fernandez, Rudy                2694    0.32    0.02    0.18
Lopez, Robin                   3204    0.34    0.43    0.17
Wilkins, Damien                3098    0.50    0.60    0.16
Brewer, Corey                  4996    0.31    0.47    0.16
Pietrus, Mickael               3121   -0.13    0.03    0.16
Budinger, Chase                4832    0.11    0.26    0.14
Johnson, James                 3679    0.57    0.58    0.13
Ayon, Gustavo                  3569    0.52    0.65    0.12
Thompson, Jason                6222    0.91    0.97    0.12
Greene, Donte                  2974    0.20    0.57    0.11
Harris, Tobias                 1775    0.43    0.45    0.11
Watson, Earl                   3744    0.23    0.27    0.10
Neal, Gary                     4347    0.13    0.05    0.09
Chandler, Wilson                750   -0.02    0.07    0.08
Hayward, Gordon                7020    0.09    0.09    0.08
Allen, Ray                     5734   -0.24   -0.21    0.08
Harris, Manny                  1312   -0.02    0.05    0.07
Duhon, Chris                   3867   -0.03    0.07    0.05
Gay, Rudy                      9185   -0.49   -0.58    0.04
Haslem, Udonis                 5572   -0.14   -0.22    0.02
World Peace, Metta             6212   -0.28   -0.30    0.02
Booker, Trevor                 4602    0.85    1.08    0.02
Jordan, Jerome                  410    0.06    0.07    0.01
Vucevic, Nikola                2818   -0.30   -0.21    0.00
Blake, Steve                   4332    0.04   -0.02   -0.03
Dyson, Jerome                   658   -0.22   -0.15   -0.03
Maxiell, Jason                 4779    0.62    0.70   -0.05
Brown, Shannon                 5118    0.01   -0.01   -0.05
Tolliver, Anthony              3427   -0.01    0.07   -0.05
Hollins, Ryan                  1925    0.18    0.05   -0.05
Price, Ronnie                  1864   -0.06    0.06   -0.06
Mohammed, Nazr                 2573   -0.20   -0.37   -0.07
Ellis, Monta                   7777    0.36    0.34   -0.08
Pachulia, Zaza                 4872   -0.17   -0.17   -0.09
Harris, Devin                  5939   -0.07   -0.29   -0.09
Turner, Evan                   5684   -0.62   -0.64   -0.10
Batum, Nicolas                 6042   -0.24   -0.41   -0.13
Hawes, Spencer                 2955   -0.61   -0.46   -0.14
Rush, Brandon                  6291    0.30    0.23   -0.15
Honeycutt, Tyler                313    0.10    0.10   -0.15
Jones, James                   2156   -0.40   -0.55   -0.15
Chandler, Tyson                7312   -0.13   -0.26   -0.17
Faried, Kenneth                3968    0.20    0.05   -0.19
Leonard, Kawhi                 5264   -0.32   -0.43   -0.19
Derozan, DeMar                 5538    0.28    0.28   -0.19
Afflalo, Arron                 7948   -0.33   -0.41   -0.20
Ellington, Wayne               3761   -0.45   -0.27   -0.22
Irving, Kyrie                  5488    0.39    0.49   -0.22
West, Delonte                  3718   -0.43   -0.37   -0.25
O'Neal, Jermaine               2036   -0.10   -0.29   -0.26
Jackson, Reggie                1831   -0.49   -0.63   -0.30
Martin, Kenyon                 3464   -0.29   -0.12   -0.32
Cousins, DeMarcus              7361   -0.10    0.18   -0.32
Green, Gerald                  1357    0.07   -0.04   -0.33
Hill, Jordan                   1976   -0.14   -0.16   -0.34
Kaman, Chris                   4904    0.23    0.06   -0.34
Williams, Marvin               4197   -0.36   -0.38   -0.35
Martin, Kevin                  4772   -0.27   -0.35   -0.35
Murphy, Troy                   3300   -0.32   -0.37   -0.36
Webster, Martell               4432   -0.32   -0.11   -0.37
Diaw, Boris                    4881   -0.07    0.19   -0.37
Leuer, Jon                     2071    0.01    0.05   -0.37
Jackson, Stephen               4190   -0.35   -0.21   -0.39
Perkins, Kendrick              6551   -1.28   -1.49   -0.39
Pendergraph, Jeff               401    0.07    0.05   -0.39
Mayo, O.J.                     6789   -0.71   -0.71   -0.40
Marion, Shawn                  7001   -1.02   -1.10   -0.42
Gibson, Daniel                 3340   -0.11   -0.18   -0.44
Beasley, Michael               4193   -0.23   -0.51   -0.44
Ridnour, Luke                  6643   -0.57   -0.55   -0.45
Pekovic, Nikola                4775   -0.12   -0.21   -0.46
Mack, Shelvin                  2240   -0.21   -0.14   -0.46
Johnson, Armon                  399   -0.19   -0.20   -0.46
Morris, Markieff               4417   -0.36   -0.33   -0.47
Lee, David                     7904   -0.31   -0.34   -0.49
Foye, Randy                    6288   -1.48   -1.76   -0.52
Terry, Jason                   7268   -0.89   -1.05   -0.52
Orton, Daniel                   456   -0.32   -0.24   -0.53
James, Damion                   623   -0.38   -0.30   -0.54
Barbosa, Leandro               4377   -0.32   -0.39   -0.54
Thompson, Klay                 5696   -0.66   -0.66   -0.54
Brockman, Jon                   871   -0.01    0.01   -0.54
Thornton, Marcus               6950   -0.30   -0.42   -0.57
Redick, J.J.                   5407   -1.03   -1.14   -0.58
Singleton, Chris               4645    0.01    0.02   -0.58
Walker, Bill                   2397   -0.41   -0.36   -0.59
Pavlovic, Sasha                1538   -0.25   -0.29   -0.60
Landry, Carl                   3504   -0.30   -0.33   -0.61
Howard, Josh                   3592   -0.63   -0.69   -0.63
Gee, Alonzo                    6114   -0.69   -0.68   -0.63
Randolph, Anthony              2051   -0.17   -0.19   -0.66
Garcia, Francisco              3041   -0.39   -0.28   -0.68
Wallace, Ben                   3144   -0.30   -0.28   -0.68
Williams, Shelden              1804   -0.29   -0.36   -0.68
Miles, C.J.                    3883   -0.64   -0.73   -0.71
Patterson, Patrick             5381   -1.09   -1.23   -0.72
Williams, Derrick              5535   -1.05   -0.96   -0.72
Jefferson, Richard             6294   -1.16   -1.28   -0.74
Ariza, Trevor                  4661   -0.60   -0.43   -0.75
Williams, Mo                   5475   -1.63   -1.66   -0.77
Jamison, Antawn                7533   -0.81   -0.77   -0.77
Samuels, Samardo               2669   -0.63   -0.55   -0.80
Johnson, Ivan                  2308   -0.18   -0.54   -0.80
Johnson, Carldell               422   -0.13   -0.12   -0.82
Hinrich, Kirk                  3114   -0.84   -0.96   -0.83
Farmar, Jordan                  801   -0.47   -0.51   -0.85
Crawford, Jamal                5489   -1.59   -1.69   -0.88
Douglas, Toney                 2434   -0.56   -0.60   -0.91
Amundson, Louis                2753   -0.58   -0.64   -0.94
Wright, Chris                   548   -0.16   -0.20   -0.95
Wafer, Von                     1572   -1.10   -1.10   -0.96
Wright, Dorell                 6169   -1.25   -1.31   -1.00
Jennings, Brandon              8249   -1.32   -1.38   -1.01
Burks, Alec                    3098   -0.92   -1.06   -1.02
McGuire, Dominic               3945   -1.13   -0.97   -1.02
Wall, John                     7948   -1.12   -1.03   -1.03
Carroll, DeMarre               1042   -0.34   -0.42   -1.03
Calderon, Jose                 4818   -1.07   -1.30   -1.06
Jerebko, Jonas                 4716   -0.98   -0.96   -1.08
Bell, Raja                     2906   -1.08   -1.23   -1.08
Hansbrough, Tyler              5242   -1.45   -1.50   -1.08
Milicic, Darko                 1777   -0.51   -0.61   -1.09
Livingston, Shaun              3921   -0.85   -0.84   -1.09
Harangody, Luke                 662   -0.59   -0.62   -1.13
Simmons, Bobby                 1565   -0.42   -0.65   -1.16
Barea, Jose                    4024   -0.98   -1.14   -1.20
Hayes, Chuck                   3791   -0.78   -0.97   -1.20
Gooden, Drew                   5122   -1.26   -1.19   -1.20
Blair, DeJuan                  4617   -1.75   -2.01   -1.24
Cole, Norris                   4310   -1.94   -2.16   -1.28
Evans, Tyreke                  8154   -1.99   -2.01   -1.28
Daniels, Marquis               1705   -0.79   -0.83   -1.32
Davis, Glen                    4201   -1.68   -1.80   -1.33
Monroe, Greg                   6763   -1.42   -1.41   -1.35
Stevenson, DeShawn             1648   -0.93   -0.88   -1.35
Cunningham, Dante              4302   -1.57   -1.90   -1.36
Brown, Kwame                    732   -0.26   -0.32   -1.38
Young, Nick                    6525   -1.60   -1.78   -1.38
Outlaw, Travis                 1762   -0.68   -0.65   -1.40
Parker, Anthony                4364   -1.42   -1.21   -1.40
Miller, Brad                    562   -0.38   -0.43   -1.40
Evans, Maurice                  896   -0.36   -0.43   -1.47
Hickson, J.J.                  4112   -1.75   -1.62   -1.47
Okafor, Emeka                  2778   -0.68   -0.92   -1.52
Belinelli, Marco               6840   -2.30   -2.32   -1.54
Smith, Nolan                   1426   -1.08   -1.10   -1.54
Johnson, Chris                  554   -0.45   -0.40   -1.55
Johnson, Trey                   227   -0.15   -0.18   -1.56
Stoudemire, Amare              5410   -2.41   -2.52   -1.58
Johnson, Wesley                5630   -2.33   -2.46   -1.59
Telfair, Sebastian             3142   -1.43   -1.55   -1.63
Pittman, Dexter                1018   -0.71   -0.80   -1.63
Brooks, Marshon                3066   -1.62   -1.80   -1.63
Davis, Ed                      3515   -1.27   -1.53   -1.67
Biedrins, Andris               2744   -1.38   -1.33   -1.68
Watson, C.J.                   3850   -2.13   -2.38   -1.71
Lewis, Rashard                 2771   -1.07   -0.94   -1.71
Crawford, Jordan               5789   -2.23   -2.33   -1.74
Boykins, Earl                   414   -0.34   -0.36   -1.75
Ivey, Royal                    1376   -0.95   -1.35   -1.75
Smith, Greg                     252   -0.12   -0.23   -1.76
Salmons, John                  4895   -1.76   -2.02   -1.77
Przybilla, Joel                1309   -0.94   -1.07   -1.79
Mullens, Byron                 4752   -1.61   -1.66   -1.80
Gordon, Ben                    4555   -1.64   -1.81   -1.80
Odom, Lamar                    3766   -1.77   -1.98   -1.82
Carroll, Matt                  1976   -1.12   -1.20   -1.84
Higgins, Cory                  1360   -0.79   -0.91   -1.85
Cook, Brian                     942   -0.73   -0.85   -1.86
Prince, Tayshaun               6825   -2.61   -2.59   -1.86
Brown, Derrick                 4545   -1.78   -1.66   -1.87
Richardson, Quentin            2347   -1.58   -1.71   -1.88
Childress, Josh                1703   -1.16   -1.20   -1.88
Knight, Brandon                6903   -2.60   -2.72   -1.92
Collins, Jason                  777   -0.45   -0.62   -1.94
Vesely, Jan                    3146   -1.61   -1.68   -1.95
Williams, Terrence             1836   -1.31   -1.37   -1.96
Najera, Eduardo                 740   -0.57   -0.55   -1.97
Speights, Marreese             5032   -2.82   -3.26   -1.99
White, D.J.                    3704   -1.90   -2.04   -1.99
Blatche, Andray                2427   -1.08   -1.29   -1.99
Petro, Johan                   1543   -1.37   -1.40   -2.06
Delfino, Carlos                5443   -2.44   -2.54   -2.06
Mozgov, Timofey                2603   -1.78   -1.95   -2.09
Goudelock, Andrew              1387   -1.05   -1.18   -2.12
Jack, Jarrett                  5251   -2.61   -2.82   -2.13
Gray, Aaron                    1808   -1.10   -1.36   -2.15
Henderson, Gerald              5695   -2.65   -2.68   -2.18
Jones, Dahntay                 3849   -2.17   -2.26   -2.20
Selby, Josh                     931   -0.79   -0.88   -2.22
Gomes, Ryan                    1648   -1.67   -1.81   -2.24
Daye, Austin                   1963   -1.35   -1.37   -2.26
Wilcox, Chris                  1663   -1.32   -1.64   -2.26
Singleton, James                208   -0.27   -0.21   -2.26
Thomas, Lance                  2138   -1.85   -1.80   -2.27
Smith, Craig                   1504   -1.38   -1.47   -2.37
Johnson, JaJuan                1007   -0.63   -0.80   -2.39
Augustin, D.J.                 4462   -2.78   -2.74   -2.42
McGee, JaVale                  5817   -3.20   -3.43   -2.42
Thomas, Tyrus                  3212   -2.21   -2.18   -2.44
Walton, Luke                   1119   -1.01   -1.14   -2.45
Walker, Kemba                  5892   -3.23   -3.41   -2.45
Butler, Rasual                 1511   -1.31   -1.32   -2.46
Casspi, Omri                   4665   -3.04   -3.34   -2.51
McRoberts, Josh                2306   -2.00   -2.27   -2.51
Diop, DeSagana                 1126   -1.03   -1.07   -2.53
Morrow, Anthony                2443   -2.26   -2.34   -2.56
Tyler, Jeremy                  1798   -1.54   -1.64   -2.57
Fredette, Jimmer               4284   -2.72   -2.89   -2.58
Stephenson, Lance              1623   -1.25   -1.41   -2.58
Kanter, Enes                   2978   -2.27   -2.51   -2.61
Scalabrine, Brian               370   -0.28   -0.31   -2.66
Villanueva, Charlie             466   -0.37   -0.52   -2.73
Summers, DaJuan                 736   -0.63   -0.74   -2.74
Thompson, Tristan              4587   -3.35   -3.58   -2.75
Green, Willie                  2632   -2.10   -2.43   -2.79
Stiemsma, Greg                 2413   -2.50   -2.59   -2.83
Morris, Marcus                  467   -0.63   -0.73   -2.89
Williams, Reggie               2500   -2.24   -2.13   -2.94
Redd, Michael                  2832   -2.64   -2.84   -2.96
Young, Sam                     1262   -1.90   -1.92   -2.96
Jenkins, Charles               3019   -2.98   -2.93   -3.00
Howard, Juwan                   547   -0.88   -0.93   -3.03
Bibby, Mike                    1981   -2.29   -2.33   -3.03
Cardinal, Brian                1004   -1.29   -1.48   -3.07
Erden, Semih                   1238   -1.28   -1.53   -3.10
Silas, Xavier                   146   -0.36   -0.33   -3.10
Okur, Mehmet                   1071   -1.56   -1.55   -3.11
Carter, Anthony                 749   -0.89   -1.01   -3.12
Lee, Malcolm                    937   -1.16   -1.34   -3.15
Biyombo, Bismack               4476   -3.50   -3.62   -3.17
Gladness, Mickell               751   -0.95   -1.09   -3.33
Babbitt, Luke                  1417   -2.07   -2.21   -3.34
Williams, Elliot                519   -0.72   -0.85   -3.34
Russell, Walker                1054   -1.25   -1.41   -3.39
Maggette, Corey                2633   -2.46   -2.67   -3.41
Joseph, Cory                    716   -1.07   -1.18   -3.53
Evans, Reggie                  2784   -3.16   -3.44   -3.53
Harris, Terrel                 1000   -1.53   -1.78   -3.59
Adrien, Jeff                    254   -0.51   -0.52   -3.60
Evans, Jeremy                   716   -0.83   -1.01   -3.66
Aldrich, Cole                   661   -1.03   -1.29   -3.66
Gaines, Sundiata               1624   -2.30   -2.44   -3.69
Flynn, Jonny                   1343   -2.13   -2.29   -3.71
Jianlian, Yi                    792   -1.39   -1.63   -4.38
Dawson, Eric                     30   -0.05   -0.06   -4.47
Dooling, Keyon                 2157   -3.57   -3.86   -4.56
Jones, Dominique                986   -2.05   -2.24   -4.68
Clark, Earl                    1658   -3.13   -3.47   -4.85
Barron, Earl                     34   -0.08   -0.09   -4.87
Davis, Josh                     508   -1.24   -1.35   -5.16
Pargo, Jeremy                  1641   -3.02   -3.42   -5.24
Morris, Darius                  531   -1.15   -1.29   -5.39
Warrick, Hakim                 1899   -3.52   -3.96   -5.45
Sloan, Donald                  1973   -3.92   -4.18   -5.47
Nocioni, Andres                 215   -0.65   -0.80   -5.51
Williams, Shawne                443   -1.21   -1.31   -5.53
Leslie, Travis                  178   -0.67   -0.73   -5.76
Jones, Solomon                  997   -2.43   -2.61   -5.91
Gadzuric, Dan                    51   -0.12   -0.15   -6.32
Fortson, Courtney               371   -1.11   -1.31   -6.37
Harper, Justin                  275   -1.07   -1.13   -6.48
Watkins, Darryl                 351   -1.14   -1.16   -6.94
Thabeet, Hasheem                397   -1.36   -1.48   -7.00
Ahearn, Blake                    96   -0.30   -0.37   -7.07
Thompkins, Trey                 471   -1.61   -1.81   -7.16
Thomas, Malcolm                  61   -0.21   -0.24   -7.36
Liggins, DeAndre                404   -1.54   -1.68   -7.42
Hayward, Lazar                  519   -1.63   -1.98   -7.54
Ubiles, Edwin                   121   -0.43   -0.49   -7.77
Dentmon, Justin                 107   -0.34   -0.44   -8.02
Eyenga, Christian               298   -1.08   -1.29   -8.78
Hughes, Larry                   407   -1.58   -1.81   -8.86
Moon, Jamario                   344   -1.29   -1.60   -9.60
Kennedy, D.J.                   102   -0.65   -0.68  -10.52
Brackins, Craig                 329   -1.75   -1.93  -10.76
Horner, Dennis                   72   -0.41   -0.49  -11.51
Byars, Derrick                   18   -0.12   -0.13  -13.52
Elson, Francisco                 63   -0.49   -0.57  -14.87
Azubuike, Kelenna                15   -0.12   -0.13  -15.48
Curry, Eddy                     187   -1.51   -1.73  -17.38
Skinner, Brian                   18   -0.20   -0.22  -22.91

I think the difference is clearly to see. All those low poss players at the top and the bottom are way closer to 0 as they should be.

But I'm sure I always used alpha=0 because that was the ridge penalty. Between 0 and 1 you would get elastic-net and with 1 you would get lasso. Ridge was better than others by a small margin.

Yeah, I think I remember you using alpha=0 in that one script you posted anyway, but I wanted to make sure that the OP is aware of it.

[permaximum]

@mystic

Thanks for all those reminders, the table itself and some crucial information. That table looks too familiar and I can say without standardization things look accurate. Then I probably didn't bother to read the description for standardize in the glmnet documentation although I thought I used it for some basketball-related calculation.

As for small differences the glmnet provides in the results, when I first started using glmnet I read about it's very efficient and quick solving method which provides "almost" perfect results. I tested different libraries which includes ridge regression calculations such as ridge.cv from parcor and lm.ridge from MASS. lm.ridge and ridge.cv were slower and more RAM taxing but they both gave the same results which should be 100% accurate compared to glmnet's almost accurate results. Unfortunately as you know only glmnet provides weighted ridge regression in R.

Still, it's better than nothing and results look similar enough compared with the actual. Funny, this problem was one of the reasons I stopped doing any basketball-related analytic stuff.

[KwameBeanJordan]

Yes, your current structure would work fine. Just do IF(OR(Z$1=$F1,Z$1=$G1,Z$1=$H1,Z$1=$I1,Z$1=$J1),1,IF(OR(Z$1=$K1,Z$1=$L1,Z$1=$M1,Z$1=$N1,Z$1=$O1),-1,0)). There may be a better construction for speed of calculation, but this method is constructed similarly to the last one.

Thanks, that worked. I should have known there was an "OR" command in Excel. I had been trying to combine the "IF" function with the "FIND" function over a range of cells and it wasn't working.

Haven't read the link in the OP, but Eli Witus explained how to get the data into a csv file via Excel here: http://www.countthebasket.com/blog/2008 ... lus-minus/

Maybe that is more helpful ...

Yeah, thanks a ton, that guide is very useful. Last night it took several hours for me to set up a single excel file due to computing times. Using the formulas in that guide it took just a couple of minutes. Not to mention it's very thorough and could be used by some with absolutely no knowledge of Excel.

[permaximum]

Yeah, thanks a ton, that guide is very useful. Last night it took several hours for me to set up a single excel file due to computing times. Using the formulas in that guide it took just a couple of minutes. Not to mention it's very thorough and could be used by some with absolutely no knowledge of Excel.

There are some serious wrong information on that guide. I can't remember atm but it will lead to wrong results if you try to calculate RAPM.

[jbrocato23]

I think this passage from the help page might explain why you guys are having trouble with glmnet:

Note also that for "gaussian", glmnet standardizes y to have unit variance before computing its lambda sequence (and then unstandardizes the resulting coefficients); if you wish to reproduce/compare results with other software, best to supply a standardized y.

Try running your rapm regression in glmnet with lambda=2 and standardize = F.

[KwameBeanJordan]

I ran the ridge regression for the 2009 regular season with standardize=FALSE and alpha=0. Lambda from 10-fold cross validation was about 0.39. Here are the results. I can't tell if the "weird" results are just because it's single year, NPI, or because of issues with glmnet. Sorry for the formatting.

Code: Select all

Last Name	 First Name	RAPM
Odom	 Lamar	12.31604942
Wade	 Dwayne	11.8765314
Kidd	 Jason	11.27945562
James	 LeBron	11.03529885
Paul	 Chris	9.198197105
Allen	 Ray	8.758585813
Bonner	 Matt	8.453281324
Garnett	 Kevin	8.389645919
Murray	 Ronald (Flip)	8.111565661
Ming	 Yao	7.550693524
Iguodala	 Andre	7.280219036
Telfair	 Sebastian	6.955628379
Hilario	 Nene	6.952163792
Collins	 Mardy	6.698537662
Hayes	 Jarvis	6.609101395
Parker	 Anthony	6.423443631
Hill	 Grant	6.186705823
Foster	 Jeff	6.088112656
Douglas-Roberts	 Chris	5.992923648
Granger	 Danny	5.956554307
Wallace	 Rasheed	5.845939892
Hinrich	 Kirk	5.707718038
Udoka	 Ime	5.623271265
Kirilenko	 Andrei	5.544449309
Gasol	 Pau	5.454487359
Hollins	 Ryan	5.334297199
McDyess	 Antonio	5.212261217
Foye	 Randy	5.158671278
Singleton	 James	5.12141204
Bosh	 Chris	5.062289582
Young	 Thaddeus	5.007335729
Walton	 Luke	4.927350102
Roy	 Brandon	4.908400221
Nelson	 Jameer	4.902344019
Wallace	 Gerald	4.870096386
Anthony	 Carmelo	4.825343673
Battier	 Shane	4.823911994
Millsap	 Paul	4.781569197
Pierce	 Paul	4.601448312
Weaver	 Kyle	4.398207722
Batum	 Nicolas	4.38092317
Redd	 Michael	4.33398484
Milicic	 Darko	4.325750567
Moon	 Jamario	4.285770007
Cardinal	 Brian	4.142546525
Smith	 J.R.	4.11946775
Bryant	 Kobe	4.043974622
Jaric	 Marko	3.889485428
Brown	 Devin	3.789652904
Turiaf	 Ronny	3.757028389
Thompson	 Jason	3.725364336
Stoudemire	 Amare	3.665040902
Bynum	 Will	3.621943647
Smith	 Josh	3.601936654
Rondo	 Rajon	3.559299883
Brown	 Kwame	3.553469219
Magloire	 Jamaal	3.489628942
West	 David	3.332831241
Wafer	 Von	3.32347593
Okafor	 Emeka	3.21520411
Udrih	 Beno	3.16616916
Johnson	 Amir	3.151854329
Afflalo	 Arron	3.143567016
Williams	 Deron	3.1342971
Bogut	 Andrew	3.095992632
Aldridge	 LaMarcus	3.084918119
Evans	 Maurice	3.066128404
Jefferson	 Al	3.02261335
Przybilla	 Joel	2.969752596
Martin	 Kenyon	2.950992753
Andersen	 Chris	2.905510421
Johnson	 Joe	2.817861328
Blatche	 Andray	2.795720579
Wallace	 Ben	2.793657254
West	 Delonte	2.752831675
Mbah a Moute	 Luc	2.727471616
Randolph	 Zach	2.691843429
Jones	 James	2.670968779
Nocioni	 Andres	2.577677604
Mason	 Roger	2.485041796
Davis	 Glen	2.237945106
Deng	 Luol	2.211217632
Battie	 Tony	2.20831788
Carter	 Vince	2.186501337
Radmanovic	 Vladimir	2.171790002
Young	 Nick	2.168966305
Westbrook	 Russell	2.165081043
Blount	 Mark	2.160453789
Chandler	 Tyson	2.118713733
Gordon	 Ben	2.094284898
Herrmann	 Walter	2.087852006
Bass	 Brandon	2.046412804
Parker	 Tony	2.014860161
Mason	 Desmond	2.010104992
Noah	 Joakim	1.889202827
Okur	 Mehmet	1.87524535
Oden	 Greg	1.85437902
Richardson	 Quentin	1.851153815
Barnes	 Matt	1.837738365
Allen	 Tony	1.83438973
Maggette	 Corey	1.824347056
Robinson	 Nate	1.811929267
Ollie	 Kevin	1.77872978
Jack	 Jarrett	1.65469783
Amundson	 Louis	1.630131703
Songaila	 Darius	1.585679815
Varejao	 Anderson	1.524559077
Alexander	 Joe	1.516891731
Stojakovic	 Peja	1.512613935
Dudley	 Jared	1.51150629
Balkman	 Renaldo	1.495941038
Korver	 Kyle	1.46334628
Jamison	 Antawn	1.425706882
Hibbert	 Roy	1.414401248
Ilgauskas	 Zydrunas	1.37515858
Graham	 Joey	1.36760235
Wright	 Brandan	1.349153488
Lopez	 Brook	1.341471727
Harrington	 Al	1.338683372
O'Neal	 Shaquille	1.323041244
Gibson	 Daniel	1.284552897
Lee	 David	1.265992022
Blake	 Steve	1.186409223
Graham	 Stephen	1.181859652
Anderson	 Ryan	1.151676132
Artest	 Ron	1.140214782
Turkoglu	 Hedo	1.072903692
Duncan	 Tim	1.053549273
Stuckey	 Rodney	1.041356337
Morrow	 Anthony	0.972778617
Haywood	 Brendan	0.944737633
Bell	 Charlie	0.922812506
Randolph	 Anthony	0.909657534
Billups	 Chauncey	0.887742392
Szczerbiak	 Wally	0.871453708
Felton	 Raymond	0.860739438
Posey	 James	0.828030822
Gray	 Aaron	0.819350519
Lewis	 Rashard	0.81605232
Jackson	 Stephen	0.815000674
Williams	 Sean	0.747756234
Rush	 Kareem	0.736976326
Brewer	 Corey	0.683642845
House	 Eddie	0.683019756
Simmons	 Bobby	0.681824538
Ford	 T.J.	0.641563489
Howard	 Dwight	0.599767864
Lue	 Tyronn	0.595529228
Harpring	 Matt	0.593756695
Harris	 Devin	0.573913187
Sessions	 Ramon	0.570096924
Butler	 Caron	0.516710288
Jackson	 Bobby	0.50713767
Conley	 Mike	0.477797763
Ellis	 Monta	0.465755532
Wright	 Julian	0.438722439
Carter	 Anthony	0.37395917
Nowitzki	 Dirk	0.365048468
Rose	 Derrick	0.354499797
Vujacic	 Sasha	0.336594818
Fernandez	 Rudy	0.290368224
Ratliff	 Theo	0.189821727
McGrady	 Tracy	0.174351338
Dampier	 Erick	0.159164432
Moore	 Mikki	0.138730652
Powe	 Leon	0.130031037
Bell	 Raja	0.116650438
Humphries	 Kris	0.055504608
Lee	 Courtney	0.054641712
Scola	 Luis	0.022540554
Finley	 Michael	0.001485102
Jeffries	 Jared	-0.01135624
Ely	 Melvin	-0.01843289
Dragic	 Goran	-0.056003407
Hassell	 Trenton	-0.076810476
Augustin	 D.J.	-0.096431429
Brewer	 Ronnie	-0.11328789
Ginobili	 Manu	-0.137237296
Williams	 Mo	-0.160712444
Nash	 Steve	-0.216331254
Dalembert	 Samuel	-0.22343989
Diaw	 Boris	-0.232506505
Maxiell	 Jason	-0.238786909
Marion	 Shawn	-0.244554643
Redick	 J.J.	-0.251507814
Williams	 Shelden	-0.262639122
Knight	 Brevin	-0.304775618
Cook	 Daequan	-0.318765159
Miller	 Brad	-0.322691185
Quinn	 Chris	-0.327207235
Thomas	 Tim	-0.353290209
Biedrins	 Andris	-0.354760461
Thomas	 Kurt	-0.394519797
Brooks	 Aaron	-0.396495696
Nesterovic	 Rasho	-0.401690722
Barbosa	 Leandro	-0.424026617
Duhon	 Chris	-0.447553693
Camby	 Marcus	-0.457264986
Allen	 Malik	-0.458914497
Gay	 Rudy	-0.460485532
Davis	 Baron	-0.524668835
Green	 Jeff	-0.529540355
Rose	 Malik	-0.529671876
Terry	 Jason	-0.564288939
Speights	 Marreese	-0.604700282
Gordon	 Eric	-0.66732338
Kaman	 Chris	-0.686873032
Villanueva	 Charlie	-0.698214386
Johnson	 Anthony	-0.710294639
Najera	 Eduardo	-0.715751318
Prince	 Tayshaun	-0.760767678
Watson	 Earl	-0.807704924
Arthur	 Darrell	-0.820967776
Kapono	 Jason	-0.885937099
Collison	 Nick	-0.890345925
Buckner	 Greg	-0.901312724
Miles	 C.J.	-0.9093213
Douby	 Quincy	-0.962301537
Bynum	 Andrew	-0.987122121
Calderon	 Jose	-0.990023715
Scalabrine	 Brian	-1.044961567
Carney	 Rodney	-1.045015137
Azubuike	 Kelenna	-1.045619006
Ross	 Quinton	-1.060205466
Love	 Kevin	-1.087401563
Daniels	 Marquis	-1.093314097
Miller	 Andre	-1.112266339
Hill	 George	-1.11316351
Boone	 Josh	-1.200956452
Warrick	 Hakim	-1.210525987
Horford	 Al	-1.223514516
Mayo	 O.J.	-1.223778012
Bibby	 Mike	-1.232663375
Howard	 Josh	-1.233533442
Williams	 Marvin	-1.274626149
Pachulia	 Zaza	-1.288160489
Krstic	 Nenad	-1.32094039
Jones	 Fred	-1.337772501
Pavlovic	 Sasha	-1.38023789
Webster	 Martell	-1.398459821
Mobley	 Cuttino	-1.420827501
Watson	 C.J.	-1.45198423
Peterson	 Morris	-1.481996354
Sefolosha	 Thabo	-1.493863774
Jianlian	 Yi	-1.514221216
Gasol	 Marc	-1.549669886
Gomes	 Ryan	-1.593247406
Brown	 Shannon	-1.60693356
Bogans	 Keith	-1.648775551
Elson	 Francisco	-1.69193221
Diop	 DeSagana	-1.70240628
Green	 Willie	-1.741140071
Murphy	 Troy	-1.82131743
Atkins	 Chucky	-1.836393796
Miller	 Mike	-1.874320767
Oberto	 Fabricio	-1.887304122
Price	 Ronnie	-1.911348712
McGuire	 Dominic	-1.942476184
Wilkins	 Damien	-1.995378364
Belinelli	 Marco	-2.030233007
Law	 Acie	-2.03547271
Wright	 Antoine	-2.041160504
Taylor	 Mike	-2.069664363
Thomas	 Tyrus	-2.088922052
Fisher	 Derek	-2.144742078
Beasley	 Michael	-2.174921669
Martin	 Kevin	-2.181025699
Outlaw	 Travis	-2.207936061
Boozer	 Carlos	-2.307622554
Brown	 Bobby	-2.346158213
Wilcox	 Chris	-2.394121044
Ukic	 Roko	-2.40197247
Lopez	 Robin	-2.492740414
Bowen	 Bruce	-2.515387459
Farmar	 Jordan	-2.55278094
Landry	 Carl	-2.587775482
Barea	 Jose	-2.653936557
Stevenson	 DeShawn	-2.781176939
Evans	 Reggie	-2.782106111
Hayes	 Chuck	-2.815237565
Koufos	 Kosta	-2.834095339
Alston	 Rafer	-2.853736677
Brand	 Elton	-2.874219439
Chalmers	 Mario	-2.886807318
Perkins	 Kendrick	-2.893579315
Smith	 Joe	-2.893755276
Gadzuric	 Dan	-2.937119768
Iverson	 Allen	-3.079090981
Kleiza	 Linas	-3.11454609
Williams	 Louis	-3.166315696
Pietrus	 Mickael	-3.187397154
Barry	 Brent	-3.190249777
O'Neal	 Jermaine	-3.247717322
Diawara	 Yakhouba	-3.35214309
Ridnour	 Luke	-3.495698653
Thornton	 Al	-3.575891825
Hughes	 Larry	-3.636935224
Haslem	 Udonis	-3.665554948
Jefferson	 Richard	-3.711116327
Marks	 Sean	-3.737765555
Solomon	 Willie	-3.757652076
McCants	 Rashad	-3.808491417
Crawford	 Jamal	-3.836575689
Smith	 Craig	-3.865006131
Skinner	 Brian	-3.877807917
Mohammed	 Nazr	-3.891136877
Gortat	 Marcin	-3.89343883
Novak	 Steve	-3.915413799
Marbury	 Stephon	-3.985748547
Greene	 Donte	-4.015011767
Gooden	 Drew	-4.044457818
Diener	 Travis	-4.302566726
Crittenton	 Javaris	-4.330614955
Howard	 Juwan	-4.456837068
Richardson	 Jason	-4.489089542
Armstrong	 Hilton	-4.490297644
Bargnani	 Andrea	-4.507606855
Dixon	 Juan	-4.535635696
Rodriguez	 Sergio	-4.553094553
McGee	 JaVale	-4.562850392
Stackhouse	 Jerry	-4.775890987
Hawes	 Spencer	-4.8627335
Chandler	 Wilson	-4.893609433
Petro	 Johan	-4.991658586
Jones	 Solomon	-5.08162245
Garcia	 Francisco	-5.128463183
Durant	 Kevin	-5.252985678
Davis	 Ricky	-5.315195953
Ariza	 Trevor	-5.331560836
Morrison	 Adam	-5.347620495
Lowry	 Kyle	-5.40081842
Salmons	 John	-5.418413517
Dooling	 Keyon	-5.640375878
Butler	 Rasual	-5.740804426
Anthony	 Joel	-5.743847664
Jordan	 DeAndre	-6.122952289
George	 Devean	-6.18068668
Hamilton	 Richard	-6.230281298
Hart	 Jason	-6.255887757
Hickson	 J.J.	-6.280278717
James	 Mike	-6.434333349
Vaughn	 Jacque	-6.557000931
Jones	 Dahntay	-6.764371097
Ivey	 Royal	-6.801467081
Dunleavy	 Mike	-6.824878577
Bayless	 Jerryd	-6.849230606
Carroll	 Matt	-6.928231035
Powell	 Josh	-6.983124086
Green	 Gerald	-7.297763639
Head	 Luther	-7.773401035
Frye	 Channing	-8.520804181
Collins	 Jason	-8.791717235
Daniels	 Antonio	-11.15968912
Rush	 Brandon	-11.19596718

Note: If you have any questions about what I did I won't have internet access for about 30 hours.