[Historical and Non-Current] User Guide for NBA Real Player Ratings, July 2008

IMPORTANT NOTICE: THIS USER GUIDE IS EXPIRED. FOR THE CURRENT USER GUIDE TO REAL PLAYER RATINGS LOOK FOR THE LATEST VERSION IN THE REPORT INDEX. THIS GUIDE DOES APPLY, HOWEVER, TO REAL PLAYER RATING REPORTS PRIOR TO NOVEMBER 2008.

USER GUIDE FOR NBA REAL PLAYER RATINGS
As of July 2008

We start by taking the top 390 players, out of about 450 players who played in the NBA in 2007-08, ranked according to gross or simple player rating, which is as follows:

ADD THE FOLLOWING
Points
Rebounds
1.4 X Assists
Steals
1.4 X Blocks
Field Goals Made
0.5 X # of 3-Pointers Made
0.25 X # of Free Throws Made

SUBTRACT THE FOLLOWING
0.7 X Turnovers
0.8 X # of Missed Field Goals
0.8 X # of Missed Free Throws

Real Player Rating, the holy grail of player ratings, is then gross or simple player rating divided by minutes per game. This gives you the actual production per minute of each player, so you can now directly compare players with very different playing times. By discovering players who have high ratings, but low minutes, you can spot players who were underrated by their coaches. Among younger players, you can spot the promising ones and the ones who need more time to develop into full NBA players, time that in some cases may not be available, meaning that the player will end up playing in Europe or something.

Then we eliminate any player who did not play in at least 16 games.

Then we eliminate any player who did not play at least 7 minutes per game in the games he played in.

Finally, we lop off the players with the lowest real player ratings from the bottom (36 players in this case) and take the top 330 NBA players to form the official Real Player Ratings list. The average team in the NBA will have 11 players from this list.

If a player does not appear on the list, then one or more of the following is true:

1. The player was not among the top 390 out of 450 players for gross or simple real player rating.
2. His real player rating is very low, less than .536.
3. He played in fewer than 16 games.
4. His minutes per game were less than 7. In many cases, this would be a player who played mostly or only in garbage time.
5. The player is one of the best made you miss type of defenders in the NBA, but has very little offensive game. Since scores that a defender prevents is an unknown quantity, it is not accounted for by the Real Player Rating, and so it is possible for a player not to make the top 330 list despite being a valuable asset, albeit mostly on defense only. Bruce Bowen, the San Antonio small forward, is the most obvious example. His real player rating is only .370, yet he played over 30 minutes a game due to all the scores of the Spurs’ opponents he stops.

ADJUSTMENTS FOR MADE THEM MISS DEFENDING
Since shots that a defender stops from going in the basket, with no actual block, can not be known and are not kept track of, the Real Player Rating is not exactly perfect. But if you are very knowledgeable about the skills and efforts of players with respect to preventing scores, you can make your own adjustments based on your knowledge.

In order to improve my coverage of the Denver Nuggets, I introduced in 2008 “adjustments for defending” to the Nuggets real player ratings. Although neither I nor anyone knows anywhere near exactly how many scores were prevented by the various Nuggets, since I was very familiar with the players and what they can and do accomplish on the court, I was able to rank the Nuggets with respect to made you miss defending.

I decided that I would then assign an adjustment for made them miss defending to the real player ratings of each Nugget, in equal increments. Furthermore, I decided that the top half of the list of Nuggets according to made them miss defending would get positive adjustments, while the bottom half would get negative adjustments.

The next step was to estimate how much the adjustments should be. After giving it my best thought, I decided that a +.130 adjustment would be the best estimate I can come up with for what the best made them miss defender the Nuggets have should get. So Kenyon Martins’s Real Player rating is .777, but his Real Player Rating adjusted for made them miss defending is .907.

The equal increment adjustments were symmetric as to zero, so the best made them miss defender, as just discussed, received a +.130 adjustment, while the worst made them miss defender received a -.130 adjustment. Notice that this means that the best made them miss defender is a full .260 better than the worst made them miss defender, than the basic Real Player Ratings indicate. Since the Real Player Ratings of the entire list of 330 players range from .536 to 1.268, a range of .732, the .260 range for made them miss defending is, it seems clear to me, an adequate but not excessive correction of the fact that no one actually knows how many shots various players prevented from going in.

So what you do, if you think you know about how good a player is in made them miss defending compared with other players, is to adjust that player’s Real Player Rating up or down, by as much as .130 up if the player is among the very best made them miss defenders among the 330 players rated, and by as much as .130 down if the player is among the very worst.

Specifically, you estimate how the player would rank among the 330 players who are in the Real Player Ratings rankings, and then adjust that player’s rating according to the following guideline:.

Top 12: +.130
13-24: +.120
25-36: +.110
37-48: +.100
49-60: +.090
61-72: +.080
73-84: +.070
85-96: +.060
97-108: +.050
109-120: +.040
121-132: +.030
133-144: +.020
145-156: +.010
157-174: 0
175-186: -.010
187-198: -.020
199-210: -.030
211-222: -.040
223-234: -.050
235-246: -.060
247-258: -.070
259-270: -.080
271-282: -.090
283-294: -.100
295-306: -.110
307-318: -.120
319-330: -.130

As another Nuggets example, and maybe I got carried away slightly with the hysteria regarding Carmelo Anthony’s defending, but I estimated Anthony would be ranked in the 259-270 range among the 330 players if all these players were ranked according to made them miss defending. So Anthony’s Real Player Rating, which is 1.091, becomes 1.011, when it is adjusted for made them miss defending.

Remember that you can adjust only players you know very well as to their defending; it is most likely beyond anyone's capabilities to even approximately rank all 330 players, so there can be no full "Real Player Ratings Adjusted for Made Them Miss Defending". No one that I have found has attempted to do this anywhere on the internet! However, as I did with the Nuggets, I believe that you can calculate the adjusted ratings for a team, if you know the players on that team very well.

SCALE FOR REAL PLAYER RATINGS
All Time Historic Superstar Player 1.175 to Up
Superstar Player 1.050 to 1.174
Star Player 0.925 to 1.049
Outstanding Player 0.825 to 0.924
Major Role Player 0.750 to 0.824
Role Player 0.675 to 0.749
Minor Role Player 0.600 to 0.674
Reserve Only Player 0.525 to 0.599
Marginal or Struggling Player 0.450 to 0.524
Bust Players (or Defense Only!) Lower to 0.449