The ignorance or shall I say laziness of some posters on this board is really coming to life in this thread. The rank isn't about who's the better player! It's a peice of a bigger puzzle though. Thats like a lamborghini owner going crazy because some Car magazine rated a Honda Civic higher than the Diablo based on Economic efficiency.
Hey Sam, I did pointed out during "Barrymania" that his presence brought winning records to the Sonics. So I was actually smarter than you doubters.
Perfect analogy! Thank you. Oh please, you know that during Barrymania I would have been playing up the standard error rates in Rosenbaums calculations!
That explains that call from Sac Town where they asked us to trade Ostertag for Yao straight up. They said "yeah, haven't you seen the latest Rosenbaum Player Ratings?"......
Seems like these stats are all based on relativity, but have no common starting point by which to make the individual players relative to one another.
Brian Cardinal and Sabonis are in the top 20 NBA players plus/minus ratings for the last two years at about the same rating of Jason Kidd? Is that why Jerry West gave Cardinal 39 millions? Ak47 above Shaq and Kobe? Are you kidding me? This method gives too much lenience to the role players than the stars - just because it offers an angle through stats playing doesn't mean it offers a good angle on NBA players overall. This method might be more suited for evaluating role players than the star players. It's like saying a rador system is good because it cranks out nice stats when it often mistake birds as planes, then when people bash this system because it treats little birds as the bombers the defenders say oh it's because you are lazy and don't bother to understand it. Yep, this Rosenbaum guy is treating little birds in the NBA like the bombers who is a prime example of how stats manipulation can screw one's senses.
An obviously wrong/flawed rating system is worse than no rating system at all. Rosenbaum's system is supposed to measure how a player contribute to team's wins. I don't care what complex formulas he used, but the results are laughable. AK47 helped Utah win more than all the following players with their teams: Duncan, TMac, Shaq, Kobe? No way. Greg Ostertag helped Utah more than Yao Ming helped Houston? Are you insane? In fact, let's compare Rosenbaum's rating against the simple (assumed not too accurate) Efficiency ranking. Obvious misranked players in Rosenbaum's system: AK47 at 2, Baron Davis at 8, Rasheed Wallace at 11 (he is like 3rd option at POR and DET) , James Posey at 18 (Memphis 10-man rotation will work fine without Posey), Rafer Alston in front of Lamar Odom, Big Z in front of Boozer, Greg Ostertag in front of Yao Ming On the other hand, I can't really identify any obvious mis-ranked players. All the top 50 players play an essential role in the team's winning. Rosenbaum's rating is a total joke. All the GMs that paid those totally bloated contracts this FA season must have used his system. Rosenbaum Player Rating OBS FNAME LNAME OVRATE1 SEOVR1 WV5_R IWV_R IEFF40_R IPOSSPCT IEFF MIN4 GP4 MPG4 1 Kevin Garnett 16.1 2.4 1 1 1 28 108 3231 82 39.4 2 Andrei Kirilenko 12.2 2.3 4 2 12 22 106 2895 78 37.1 3 Tim Duncan 11.4 2.4 6 4 2 28 106 2525 69 36.6 4 Tracy McGrady 11.0 2.4 10 3 6 32 111 2676 67 39.9 5 Shaquille O'Neal 10.2 2.5 7 6 3 27 110 2464 67 36.8 6 Kobe Bryant 9.0 2.3 27 5 9 30 108 2447 65 37.6 7 Dirk Nowitzki 8.7 2.3 5 10 4 24 115 2915 77 37.9 8 Baron Davis 8.3 2.3 11 8 48 29 100 2686 67 40.1 9 Ray Allen 8.1 2.1 9 12 25 28 110 2149 56 38.4 10 Elton Brand 8.0 2.3 56 7 7 23 108 2676 69 38.8 11 Rasheed Wallace 7.3 2.0 8 34 35 22 105 2389 68 35.1 12 Vince Carter 7.1 2.2 3 22 34 30 101 2784 73 38.1 13 Predrag Stojakovic 6.0 2.3 49 15 27 21 118 3263 81 40.3 14 Shawn Marion 5.9 2.2 19 16 24 22 106 3218 79 40.7 15 Ron Artest 5.8 2.1 17 17 53 25 100 2714 73 37.2 16 Jason Kidd 5.6 2.4 13 19 17 28 97 2449 67 36.6 17 Paul Pierce 5.6 2.2 15 21 29 30 100 3099 80 38.7 18 James Posey 5.5 2.0 40 11 51 18 112 2452 82 29.9 19 Brad Miller 5.4 2.0 44 9 11 19 109 2622 72 36.4 20 Pau Gasol 5.2 2.2 29 20 14 25 105 2459 78 31.5 21 Marcus Camby 5.1 2.4 70 13 15 16 100 2162 72 30.0 22 Emanuel Ginobili 5.0 2.3 45 14 55 22 103 2260 77 29.4 23 Ben Wallace 4.8 2.2 22 26 23 15 95 3048 81 37.6 24 Andre Miller 4.7 2.2 31 25 57 23 102 2838 82 34.6 25 Theo Ratliff 4.6 2.1 38 24 50 14 98 2664 85 31.3 26 Mike Bibby 4.5 2.3 57 23 52 23 109 2982 82 36.4 27 Nene Hilario 4.4 2.2 2 49 70 18 101 2504 77 32.5 28 Jermaine O'Neal 4.4 2.2 35 27 8 27 100 2788 78 35.7 29 Eddie Jones 4.4 2.3 23 35 79 22 109 2998 81 37.0 30 Sam Cassell 4.3 2.2 101 18 13 28 110 2838 81 35.0 31 Reggie MillerR 4.3 2.2 24 32 95 16 119 2254 80 28.2 32 Doug Christie 4.1 2.1 18 36 91 15 102 2778 82 33.9 33 Richard Jefferson 4.1 2.2 39 31 46 21 106 3136 82 38.2 34 Chauncey Billups 4.1 2.3 36 33 44 24 110 2760 78 35.4 35 Darrell Armstrong 3.7 2.1 32 37 94 20 100 2248 79 28.5 36 Morris Peterson 3.7 1.9 14 60 121 17 104 2144 82 26.1 37 Gary Payton 3.5 2.1 25 47 45 23 105 2825 82 34.5 38 Vlade Divac 3.5 2.2 34 39 38 20 97 2317 81 28.6 39 Michael Finley 3.5 2.1 30 40 77 21 111 2778 72 38.6 40 Kenyon Martin 3.1 2.3 61 42 22 24 98 2251 65 34.6 41 Allen Iverson 3.0 2.1 47 41 62 33 98 2040 48 42.5 42 Zydrunas Ilgauskas 3.0 2.2 102 38 21 24 103 2538 81 31.3 43 Rashard Lewis 2.9 2.1 58 43 61 22 106 2930 80 36.6 44 Steve Francis 2.7 2.2 20 51 56 26 99 3193 79 40.4 45 Donyell Marshall 2.7 2.1 99 30 20 19 105 2989 82 36.5 46 Rafer Alston 2.7 2.2 50 45 108 19 101 2581 82 31.5 47 Lamar Odom 2.7 2.2 80 29 30 24 99 3003 80 37.5 48 Greg Ostertag 2.6 2.2 68 46 63 14 98 2153 78 27.6 49 LeBron James 2.2 2.3 71 50 54 29 97 3122 79 39.5 50 Yao Ming 2.1 2.2 21 59 5 24 108 2693 82 32.8 2003-04 Regular Season Efficiency Rating Efficiency Formula: ((PTS + REB + AST + STL + BLK) - ((FGA - FGM) + (FTA - FTM) + TO)) / G Player G EFF 1. Kevin Garnett (Minnesota Timberwolves) 82 33.13 2. Tim Duncan (San Antonio Spurs) 69 26.80 3. Elton Brand (Los Angeles Clippers) 69 25.01 4. Shaquille O'Neal (Los Angeles Lakers) 67 24.93 5. Dirk Nowitzki (Dallas Mavericks) 77 24.17 6. Tracy McGrady (Orlando Magic) 67 23.75 7. Predrag Stojakovic (Sacramento Kings) 81 22.99 8. Brad Miller (Sacramento Kings) 72 22.89 9. Kobe Bryant (Los Angeles Lakers) 65 22.68 10. Shawn Marion (Phoenix Suns) 79 22.44 11. Carlos Boozer (Cleveland Cavaliers) 75 22.23 12. Andrei Kirilenko (Utah Jazz) 78 21.59 13. Jermaine O'Neal (Indiana Pacers) 78 21.55 14. Zach Randolph (Portland Trail Blazers) 81 21.02 15. Ray Allen (Seattle SuperSonics) 56 20.91 16. Sam Cassell (Minnesota Timberwolves) 81 20.67 17. Yao Ming (Houston Rockets) 82 20.57 18. Paul Pierce (Boston Celtics) 80 20.49 19. Jason Kidd (New Jersey Nets) 67 20.45 20. Lamar Odom (Miami Heat) 80 20.43 21. Kenyon Martin (New Jersey Nets) 65 20.28 22. Stephon Marbury (New York Knicks) 81 20.27 23. Amare Stoudemire (Phoenix Suns) 55 20.07 24. Erick Dampier (Golden State Warriors) 74 20.03 25. Donyell Marshall (Toronto Raptors) 82 19.96 26. Baron Davis (New Orleans Hornets) 67 19.81 27. Ben Wallace (Detroit Pistons) 81 19.78 28. Pau Gasol (Memphis Grizzlies) 78 18.96 29. Corey Maggette (Los Angeles Clippers) 73 18.93 30. Richard Jefferson (New Jersey Nets) 82 18.84 31. LeBron James (Cleveland Cavaliers) 79 18.77 32. Steve Nash (Dallas Mavericks) 78 18.51 33. Marcus Camby (Denver Nuggets) 72 18.49 34. Vince Carter (Toronto Raptors) 73 18.38 35. Zydrunas Ilgauskas (Cleveland Cavaliers) 81 18.20 36. Allen Iverson (Philadelphia 76ers) 48 18.19 37. Michael Redd (Milwaukee Bucks) 82 17.99 38. Mike Bibby (Sacramento Kings) 82 17.93 39. Shareef Abdur-Rahim (Portland Trail Blazers) 85 17.75 40. Andre Miller (Denver Nuggets) 82 17.72 41. Jamaal Magloire (New Orleans Hornets) 82 17.71 42. Kenny Thomas (Philadelphia 76ers) 74 17.54 43. Michael Finley (Dallas Mavericks) 72 17.29 44. Rashard Lewis (Seattle SuperSonics) 80 17.18 45. Antoine Walker (Dallas Mavericks) 82 17.11 46. Steve Francis (Houston Rockets) 79 17.10 47. Rasheed Wallace (Detroit Pistons) 68 17.09 48. Carmelo Anthony (Denver Nuggets) 82 16.87 49. Ron Artest (Indiana Pacers) 73 16.84 50. Jason Richardson (Golden State Warriors) 78 16.69
aint no better way to numb the senses(especially the common senses) than to read responses like what we have in this thread... admirable job by Sam and Jeff to try to install some sense....but in the end, it's like banging your head into a titanium wall. a real shame...cause this is the kind of discussion that could have been fun..
Stats are lies, lies and lies. Obviously some stats gurus think they have more sense than Mark Twain.
Can someone tell briefly what the methodologies are ??? What kinda angle and usefullness of this rating for people who don't bother or unable to read those beta zero beta one stuff ???
one suspects if Rosenbaum's system was used to evaluate the GARM, choujie would be ranked higher than SamFisher
Well just taking a look at some of the players in his top 10, Shaq ranks 33 in minutes played, Kobe 25, Duncan 27, AK47 31. Garnett however is 11 and McGrady 9. So at best I think the results are mixed. Remembering that the best players in the league play the most minutes, some of those rankings (minutes played) aren't very high. At the very least the results are mixed. I also assumed that based on his explination because the more time you're on the court the more your results would even out. With Baron Davis and McGrady, (Davis is also ranked high in minutes played) it is understandable that they have more affect on their team's outcome because they're teams are so inept without them. I know he tries to normalize the numbers but I'm sure the weakness of your teamates helps your numbers.
My beef about his rating is interpretation. I got my Master degree in a top statistics program and now am working as statistical consultant. A rule of thumb in our profession is to use simple and easily interpretable method. Let the data speak for itself. Statistically, Rosenbaum's analysis is valid. However, interpretation of his results looks dubious. The objective of his study is also anything but straightforward. That's why a lot of us are confused about his results. I would say he's a bad statistical consultant and don't think his rating would be popular among basketball fans.
The rank may not be about who's the better player overall, but any statistic ranking is about who's the better player in one or some aspects, which begs the following question: What is this mysterious economic efficiency you are talkling about that puts Brian Cardinals in the top 20 player ratings?
Actually, there is a thread "Macgrady won't help the rockets" and now we have an article titled "Measuring How NBA Players Help Their Teams Win" with Yao behind Tag etc. interesting. Here is a quote from this "interesting" article. "In other words, in 2003-04 replacing one of the average players on the Orlando Magic with one of these five players likely would have made them a bit better than the New Jersey Nets and Memphis Grizzlies." Yeah not only he made the "interesting" list he actually claims it will have immediate effect. So replace Yao with Tag would made Rockets better. Interesting... More interesting stuf, here is the fomula he uses: MARGIN = b0 + b1X1 + b2X2 + . . . + bKXK + e interesting... I'll definitely read on, must have more interesting stuff to come...
My guess is that the more you play, the more accurate the rating is. This is true with almost all statistical analyses--sample size. But his observation is that highly rated players in this system are usually players who play a lot, which, according to his conclusion, confirms that the coaches usually knows which players have better impact on the court. BTW, for those who say something like "statistics" lie, this is not a stat churning system, at least not in the usual idea of basketball stats. It doesn't count any of the numbers such as ppg, rpg, etc. It only counts the result when a player is on the floor relative to the impacts of his teammates and his opponents.
