knicks-rally-top-nets-100-97

The Knicks Wall’s Statistical Projections

knicks-rally-top-nets-100-97
Through the use of a weighted statistical projection, the Knicks Wall will attempt to project the rotation of the Knicks, as well as the production of players, based on per minute production. Before explaining how these per game numbers were projected, it is essential to take note of a few things. First, it will be very noticeable that Iman Shumpert, Rasheed Wallace, Marcus Camby, Chris Copeland and James White will not be in the projected rotation. For Shumpert, Wallace, and Camby, their injury situations are too fluid to make a reasonable estimate on how many minutes they will be able to play. For the purposes of this projection, the minutes are based on what the rotation may look like while those players are injured. James White and Chris Copeland are not included because there is no evidence to suggest that they will get any consistent minutes, even with said injuries. In terms of injuries, Amar’e Stoudemire is projected on the assumption that he returns on Christmas (while this is not necessarily true, the formula needs an estimate of games missed to be accurate as possible). Lastly, all stats are as of December 18, 2012.

Before projecting the production of a player on a per game basis, a minute per game projection is needed. This is done via the following formula: MPG= [((M1/G1)x10)+((M2/G2)x5)+((M3/G3)x3)+(M4/G4)+(Mc/Gc)]/20

The variables are as follows: M1 and G1 equal the minutes and games played this season. M2 and G2 are the minutes and games played for last season and so on, for the last three years. Mc and Gc are minutes and games in the player’s career.

All of these factors are taken into account to project the player’s minutes based on production for the entirety of a player’s career while keeping in mind that his production this year is the most indicative of how he will play. For Jason Kidd, for example, it is relevant to note what he has done for his entire career in projecting his numbers, but his numbers this year and last will still be closer to how he will actually play.

There are also three adjustments taken into account, first of which is the age adjustment. The age adjustment is based on the assumption that minutes for a player will increase as he approaches his prime and decrease as he exits it. As such, the minutes will be multiplied by a certain factor depending on the player’s age. For players aged 18-24, the factor is 1.1. For players 25-30, the factor is 1.0. If the player is between 31 and 34, the factor is .9 and players 35+ will have a factor of .8. The second adjustment is the injury adjustment. If a player has not played yet due to injury, it will be taken into account that his minutes may be limited, depending on the severity of the injury. If the player has not played due to injury, his minutes will be multiplied by the factor (1-.003x games missed). The last adjustment is the rotational adjustment which takes into account the fact that there needs to be 240 total minutes between players in a rotation. If the total minutes in a projected rotation are more than 240, each player’s minutes will be subtracted by a certain amount to make the minutes even. This is the most difficult aspect to project, as it is unknown how a coach will make-up for a cluttered rotation, but for the purposes of this simple projection, each player had the same rotational adjustment.

To show an example, this is how Amar’e Stoudemire’s minutes were projected:

[((1543/47)x10)+((2870/78)x5)+((2838/82)x3)+(1938/53)+(22099/641)]/20 = 34.357 mpg

Age Adjustment: 34.357 x 1 = 34.357

Injury Adjustment: 34.357 x (1- .003 x 27) = 31.57

Rotational Adjustment: 31.57 – .9 = 30.67 mpg or 30.7 mpg

After the minutes are projected, each player’s production is measured on a per minute basis, using a similar formula than the one used to project minutes. The per minute production is then multiplied by the measured minutes per game to project per game production. The major difference between the two formulas is the rotational adjustment and injury adjustment and the age adjustment is lessened to 1.05 for 18-24, 1.0 for 25-30, .95 for 31-34, and .9 for 35+. The formula for points for example is as follows:

PPM= [((P1/M1)x10)+((P2/M2)x5)+((P3/M3)x3)+(P4/M4)+(Pc/Mc)]/20

PPM x MPG = PPG

The same formula is used for each major stat, such as assists, blocks, steals, and rebounds. While knowing field goal percentages and other efficiency stats are essential for understanding the full body of work of a player, they will not be used for this projection. Lastly it would be important to recognize that parts of this projection system are loosely based on some aspects of the basketball-reference simple projection system. Without further ado, here are the projections for the Knicks’ ten man rotation.

Carmelo Anthony: 34.5 mpg, 25.3 ppg, 6.3 rpg, 2.6 apg, 1.0 spg, .5 bpg

Tyson Chandler: 29.7 mpg, 11.1 ppg, 9.3 rpg, .6 apg, .7 spg, 1.0 bpg

Raymond Felton: 32.7 mpg, 14.1 ppg, 2.8 rpg, 6.6 apg, 1.3 spg, .1 bpg

Ronnie Brewer: 21.4 mpg, 5.9 ppg, 3.0 rpg, 1.5 apg, 1.1 spg, .2 bpg

Jason Kidd: 23.3 mpg, 5.7 ppg, 2.9 rpg, 3.7 apg, 1.3 spg, .3 bpg

J.R. Smith: 28.4 mpg, 13.1 ppg, 4.1 rpg, 2.4 apg, 1.2 spg, .3 bpg

Steve Novak: 18.4 mpg, 7.1 ppg, 1.7 rpg, .4 apg, .3 spg, .2 bpg

Pablo Prigioni: 10.1 mpg, 2.2 ppg, 1.1 rpg, 1.7 apg, .7 spg, o.o bpg

Amar’e Stoudemire: 30.7 mpg, 17.9 ppg, 7.2 rpg, 1.3 apg, .7 spg, 1.1 bpg

Kurt Thomas: 10.5 mpg, 1.8 ppg, 2.4 rpg, .5 apg, .2 spg, .4 bpg

This 10 man rotation projects 104.2 points, 40.8 rebounds, 21.3 assists, 8.5 steals and 4.1 blocks per game. If these projections turn out to be true, the team would score 1.6 more points per night, grab 1.32 rebounds more per night, dish out 1.49 more assists per night and record .4 more blocks per night, than current per-game numbers.