"My rule was I wouldn't recruit a kid if he had grass in front of his house.
That's not my world. My world was a cracked sidewalk." —Al McGuire

Monday, November 21, 2016

Value Add 4.0 Process and Rankings; High Ranking MU Trio

One year ago Value Add 3.0 provided the most precise value of all 4000 players in D1 basketball, and now Value Add 4.0 finally provides a measurement of how valuable each player would be if they were a Major Contributor for their team. This answers the lingering criticism of how we measure the value of a superstar stuck down the bench at Kentucky.

First we will reveal the top few players, the players from Major Conferences in the Top 1% of all players (top 40), the Big East players in the top 5% of all players (top 200) and the Marquette trio in the top 15% of all players, then we will cover the new math. While we are just updating the tables, you can find the ranking for all 3,898 players by clicking on www.valueaddbasketball.com.

No. 13 on the early list is a painful one for MU, as Brewtown Andy wrote two years ago.

RnkPlayerTeamv4.0 if MajorConfv3.0 Pts/G%usagePosGoTo Adj
1Jacob Van #15Air Force16.48MWC13.7316%PG*1.21.00
2Jeremy Morgan #20Northern Iowa16.19MVC16.1921%PF*11.00
3David DiLeo #14Central Michigan15.84MAC14.1414%PF*11.12
4Marquis Wright #1Siena12.30MAAC10.2526%PG*1.21.00
5Brandon Parrish #11TCU12.17B127.028%PF*11.73
8Markelle Fultz #20Washington12.03P1210.0331%PG*1.21.00
12Charles Cooke #4Dayton11.02A1011.0223%PF*11.00
13Marial Shayok #4Virginia10.93ACC6.7210%SG*1.11.48
17Devin Davis #15Houston10.73Amer10.7316%C*11.00
19Josh Hart #3Villanova10.51BE9.5520%SG*1.11.00
20DJ Johnson #4Kansas St.10.42B127.3611%C*11.42
22Devin Robinson #1Florida10.41SEC8.1412%SF*11.28
24Malcolm Hill #21Illinois10.36B1010.3625%SF*11.00
29ShawnDre' Jones #3Richmond9.97A108.3129%PG*1.21.00
31Tyus Battle #25Syracuse9.86ACC5.158%SG*1.11.74
32Mikal Bridges #25Villanova9.76BE7.5612%SF*11.29
33Joel Berry #2North Carolina9.76ACC7.9615%PG*1.21.02
35Anton Beard #31Arkansas9.70SEC5.139%PG*1.21.58
36Manu Lecomte #20Baylor9.61B128.0119%PG*1.21.00
63Khyri Thomas #2Creighton8.81BE6.0112%SG*1.11.33
68Kamar Baldwin #3Butler8.68BE5.1610%SG*1.11.53
107Darryl Reynolds #45Villanova7.55BE4.026%C*11.88
120Shamorie Ponds #2St. John's7.30BE6.4115%SG*1.11.03
171JP Macura #55Xavier6.63BE5.5321%PG*1.21.00
182Justin Patton #23Creighton6.49BE4.2410%C*11.53
190Jalen Brunson #1Villanova6.41BE5.3515%PG*1.21.00
222Luke Fischer #40Marquette6.09BE4.6812%C*11.30
371Haanif Cheatham #25Marquette5.00BE4.2314%SG*1.11.08
544JaJuan Johnson #23Marquette3.95BE3.9516%SF*11.00

Based on Value Add 3.0 we know that Luke Fischer improves Marquette's points per 100 trips by 4.68 (almost all on offense right now). He is worth an outstanding 130 points per 100 trips down the court. However, Fischer is only getting 12% of Marquette's overall possessions (on the court for 68% of minutes multiplied by 18% of possessions when on the court equals 12%). Based on Dean Oliver's charts in "Basketball on Paper," we can estimate that Fischer would could impact Marquette's scores by 1.3 times as many points (far right factor) if he could get the ball a solid 16% of all possessions in a game, which calculates to a potential Value of 6.09 points - the 222nd best rating in all of basketball.

We do also monitor the average performance by possession each season. While we list the five traditional positions (Point Guard, Shooting Guard, Small Forward, Power Forward and Center) this position is calculated for each player based on four factors. First, a players assist rate is divided by his defensive rebound percentage, which gives an indication on if he plays more on the perimeter or underneath. The players height is also factored, and finally these factors are compared to all of his teammates so that each team has a pretty even distribution of minutes played at each of the five positions. If a team plays a four guard set, the player who spends the most time inside and/or is tallest is going to be considered a power forward for purposes of comparison to the other "No. 4" players on other teams.

So far this season the teams' three "forward/centers" are averaging about the same value over the 351 teams. Shooting guards rank almost 10% lower in the calculations than the front line players, and point guards rank almost 20% lower, so these players have their rankings further multiplied by 1.1 and 1.2 respectively to show how much they improve their team over a replacement player at the same possession.

The initial Value Add 3.0 Calculation is thus multiplied by the position factor and then the usage factor. If a player already has at least 16% of his team's possessions then he is considered to already be performing at his maximum level.
Therefore JuJuan Johnson's 3.95 ranking is not adjusted upward because he is already possession the ball 16% of the time Marquette has it (partly by coming up with his own steals). That is good enough for 544th in the country, placing him in the top 15% of all players.

Haanif Cheatham's 4.23 is increased slightly (*1.08) based on how much more he could contribute if he had the ball 16% of the time instead of 14%, but he also has that factor multiplied by 1.1 to account for his numbers being even more impressive as a shooting guard.

No comments: