As always thanks for durvasa for all the discussions on player metrics. They are truly informative. And to save answer to Jay's comment, yeah it pretty much sucks that you pay certain amount of cash to find out they don't guarantee certain state by giving definite answer to the problem. But all aspects of life are essentially random, and predictions based on averages (technically not correct but you get the idea) are the best you can do. Taking such advise from financial experts, for instance, is likely to make you richer than co-ordinating your moves based on sunspot. I do find financial specialists grossly over-paid, and I wouldn't ever conceive myself letting other people manage my money until I am retired but these are for different reasons. A_P30 makes a valid point regarding some of the limitations of Hollinger Player Efficiency Metrics though. I am pretty sure one the main assumption in the model is that all players faces equal opponents. Now based on my experience of playing Championship Manager (That is a soccer management game for you Americans ), I tend to assign players inferior quality for lesser opponents to manage minutes and keep key players fresh for big matches. Given that actual managers and coaches running the team will probably be more risk-averse since it is their livehood that is at risk and somewhat different scheme than mine, but I do think principle of playing inferior player against inferior players still applies. That means there might be bias between different "classes" of players thus leading to Chuck Heyes kind of stat. In all admission though, this probably doesn't quite explain the whole story, and we should treat last year's PER on Chuck as some kind of anomality and wait judgement until more data comes up. With regards to the validty of ordering player's ability based on PER metric - it is hard to provide definite answer on this issue. I do think in certain cases fans are probably in a better position judge an ability of player because by the simple fact that they are better informed. I mean it is just impossible to follow all players rigorously for an analyst like Hollinger - and with regards to certain projections such as Head - I do think the GARM is probably more accurate than the analyst. There is no statistical metric for intangible attributes such as being a winner and having good work ethic, and in extrapolating young players growth prospects, it is clear these are just as essential as some of the more visible quantifiable elements of the player. Now PER assumes that endowment of these intangible work ethics are uncorrelated with more visible elements of variables that are used to calculate the metric, which in essence is similar to postulating that all players are essentially the same with regards mental part of the game. This is clearly untrue and fans with deep knowledge on specific players and frequent observations are better placed to make a judgement than Hollinger. Nonetheless, if you balance the fact that there are general tendency to over-value their own player, as well as some behaviour which can only be considered as irrational, I find Hollinger the best basis on which compare players. I can't guarantee that I won't biased on player evaluation nor I claim to have the same detailed knowledge on an opponent's players. BTW, I found Sam's comment extremely funny. After all, I am considering the prospects of learning abacus to improve my own arithmetic ability.
An observation about 2 of the new guys the team obtained: Battier, and Snyder-- What they have in common are 1. They have very high +/- ratings from last season, including high +/- in on offense and 2. They don't get too many touches on offense, measured in terms of shot attempts and assits. 3. As a consequence, their own PER isn't very high, though their impact on team play is large. I think this shows the value of just moving the ball without wasting clock and posessions. As long as you are a decent enough shooting threat to space the floor, and you don't force anything, you can often do a favor for your team without either scoring a basket or making a pass that leads to a basket-- just move the ball on without turning it over or eating too much clock is sometimes sufficient to increase your team's chances.
I agree even if Yao plays like he did at the end of last season he's numbers will go down which is a good thing in two ways 1) we have many solid shooters this year 2) because of these shooters yao will get some rest on and off the court instead of so much pressure, which means he will be more energatic on the court and quicker than usual.
Nope. He will get all his touches and average 24-25 ppg. We have many solid shooters that will spread the floor so Yao can operate even better. Just watch. Some team is gonna try that collapsing bologna on us and Novak is gonna light them up for about 8 three pointers and 35 points and that will be the end of the collapsing D.
Two points I would like to highlight: 1. He argues that Juwon Howard SUCKS AS A STARTER! Not that he sucks period, but isn't good enough to start and should not start. There is a big difference there. I believe we can do with having Juwon there as a back up. 2. Alston did what he was brought in to do, GET YAO THE BALL. So I don't the trade was as bad as people make it out to be. In the '05 season, we didn't really need MJ that much. In '06 it looked like we really needed him b/c of the lack of offense which was a result of injuries.
I don't understand why people are so caught up with the idea of Mike James. What we need is a distributor, not necessarily a scorer. We can always find a decent scorer who comes in spurts, but you don't want an individual who limits the effectiveness of your two stars by hogging the ball so much... and that is exactly what Mike James is. If you look at the league right now, it is much easier to find a scorer than it is to find a distributor... that is what makes Jason Kidd, and Nash so valuable. Not only are they distributors, but they are exceptional at it and they make their teammates better.
