Correction - We did use the small TE along with cash considerations and a potential swap of second round picks in 2008.
It is NOT the LEngth Of juwan's contract, it is the quantum. In a funny way that comes in hand in the expiring year!
Walangjo, Van Gundier - just to clarify Roland Beech modified the Roland Rating only for the 2005-2006 season stats on 82games. It now tries to combine PER and defense into it. Prior to that all the Roland Ratings are is just the difference between the on-court and off-court +/-. As I mentioned above there are all sorts of problems with this simple view of things. EDIT: That's why I'm bothered by folks pointing to the 82games data and saying that Battier's stats are off the charts. If you look at Battier's "new" Roland Rating figures for 2005-2006 they aren't nearly that impressive. That's why he changed it for the 2005-2006 stats. However he's still experimenting with it and hasn't laid out how he's doing the calculations (at least I haven't found where he's discussed it on the apbrmetrics forum etc.). http://www.82games.com/0506/rolandratings0506.htm
You guys are getting carried away- that "fair" salary is in relation to the team that they are playing on. Just +/- and On court and OFf court numbers- are not enough- they can be a strarting point, that is what makes the analysis difficult in Basketball as it involves 4 other guys- and your stats depend on the other 4
One thing to me that does not need interpreting is Juwan's +/- numbers because they are so terribly bad across the board. When a player has negative numbers with both Yao and McGrady, there is a huge problem. I really wished 82 games would put up the player triplet numbers to go along with the player pairs numbers. Triplet number would indicate to us how everyone plays alongside Yao and McGrady together and since we are looking at how to improve our starting lineup, it would reveal just exactly who is hitting and who is missing when they are playing with Yao and TMac. And I would be willing to bet that Juwan would be at the bottom of the pile of those players. I sure hope we either move him or that at the very least he just gets minutes off the bench backing up Yao. Because he is simply too slow to play power forward in the NBA. Since we now have Novak who is a much much better shooter than Juwan, we no longer need him for anything in my humble opinion. The last thing we need on this team is Juwan playing 30 mpg alongside Yao this year. Juwan with Yao is simply TERRIBLE. Make it happen Morey.
That's just the thing... I'm not sure I like the "PER difference" that much as a measure of player worth... First, the opposing player PER is pretty unreliable. There are so many factors it does not account for. For one thing, help D and rotation figures prominently into it... If you go help and leave your man open and a teammate misses the rotation and your man drains a shot, it hurts your opponent PER even though it's hardly your fault. On the other hand, if you stick to your man at the cost of playing correct rotation, it helps your own "opponent PER" but hurts the team. Also the definition of "oppponet", I think, is simply the guy who technically plays the same position as you do on the court... when Battier, for example was asked to cover the opposing SG, he was supposely still playing SF and guarding SF. And there's also the factor of good defenders being asked to guard more difficult assignmetns and such thigns. I just don't like interjecting that into player ratings. Also, I think PER itself is biased toward "heavy use" players and does't seem to subtract enough for low fg%, low ft%, not going to the line, and turnovers (they are somewhat accounted for, but not quite enough)... guys who get few touches per game gets punished somewhat for not "making plays" that show up on the boxscore. A good case is Jason Collins-- great defender, but doesn't go for steals, doesn't go for that many blocks and does more boxing out for rebounds than actually catchin them... his PER is HORRIBLE but he makes a positive impact on the game. PER helps you compare the individual production of "go-to guys" with other "go-to guys", but doesn't really help when you try to compare role players with go to guys or even compare between roleplayers. In any case, even with the "New Roland Rating", Battier looked pretty good and his "fair salary" based on it is a whopping $10.5 mil... __________________________________________________ The most important thing, though, about the use of any stat (from 82games.com or those generated for the pros) is that they help you ask some interesting and important questions rather than give you answers. For example, when a front office guy see that Snyder has a high +/- rating and a decent PER, what they should do is not immediately grab him and pay him a big salary... the response should be to to ask why his numbers are this good and both look at more detailed stats and watch some films to see if he's doing positive things or just lucked into these good numbers. The same thing applies to, say, Deshawn Stevenson, who has a reputation of a being a good defender, but his stats show that ORL was better DEFENSIVELY with him on the bench for BOTH of the last 2 years... that helps check a front office staff's instinct to pay him for his defense until you figure out if his negative numbers are indeed his fault or a result of bad luck.
Wesley and Bogans were free agents - no big deal there - I wanted or expect them gone as well. Stromile is and asset we gave away for free - not very wise - you cannot tell me we could have not gotten someone of value + picks for him. Snyder is Snyder - good addition that could have come our way regardless. So instead [or just Snyder, Battier, Novak] we could have had; Novak, Gay, Snyder, Mike James (if we just pushed harder), Stromile (who we could trade), and $4.6 TPE. I prefer this scenario. James/Alston/JL3 Snyder/Head Mac/Gay/Novak Howard/Swift/Hayes Yao/Deke Yeah it is depressing but I am trying hard to get over it. We dropped at least two balls (the ones we know of) in the last 2 weeks. I hope we learnt some important lessons here. I am leaning towards your stance on this. Though I am still somehwta depressed by the happenings of our off season so far, all I can do now is give them the benefit of doubt and hope that these moves work out and that Rudy Gay becomes a bust in the NBA and Mike James fade into oblivion in Minny. All of that have to break right for me to feel good in hindsight about all of this.
Van gundier- I haven't seen how Roland Beech is doing the "new" Roland ratings besides showing us the results of it so I don't have any comments regarding the usage of PER along with the +/- stats. I understand conceptually what he's trying to do but don't have any idea how he's approaching it. I like what Dan Rosenbaum has been trying to do with them using some form of multi-linear regression to build correlations as well as isolation techniques to try and make better determination of an individual's influence on the court vs. what is lacking in the unadjusted +/- numbers. I think Beech calls this "DanVal". Regarding PER it's pretty well known that it favors box-score stats so it's not a good indicator of other influencers like you say that aren't measured via the boxscore.
I just don't see Swift as having much trade value. I am sure that he is currently available from Memphis. If you were the Rockets GM, what would you give up to get him back?
Uh, not really. Notice the Knicks have a total team fair salary of 38 million when they have a team salary over 100 million.
