Return of Nuggets 1: Forum #2 Comments From July 2008, Part 3
Forum commentary I did from March 2008 through July 2008, when I didn't have time to do the detailed and extensive reports that I like to do, is being posted in early October, 2008. The primary themes are how the Nuggets are blowing a great (and expensive!) opportunity to play the game of basketball in such a way that respects the sport and that takes as much advantage as possible of who they have on the roster. The 2006-09 Nuggets have turned out to be an excellent case study of how not to run a basketball team; many things you should not do if you are a basketball manager or coach can be identified from what the Nuggets actually did during these years.
In these comments, do not look for the usual huge amount of detail and proof that you see in the ordinary releases here at Nuggets 1. Some of this is more like everyday conversation than like top quality sports writing. On the other hand, some of the comments do include some detailed reasoning and proof that I pride myself on in the full reports.
______________________________________________
JULY 2008 FORUM COMMENTARY ON THE NUGGETS, ESPECIALLY ABOUT THEIR MISTAKES
Much respect is due JR Smith, who allowed the Nuggets to squeak into the playoffs and in general prevented them from disintegrating.
On second thought, damn JR Smith; if he had not been so good, the Nuggets might have been forced to make real changes in the off season, leading to a real chance of winning a playoff series.
Oh well, its not his fault that the Nuggets organization has proved that it is not ready for prime time.
Whereas, JR Smith is now ready for prime time...
_______________________________________
WHY CARMELO ANTHONY PLAYS BETTER INTERNATIONALLY
Anthony said the Nuggets "quit" in game 3 of the Lakers series, but didn't know why. The answer is that the Nuggets knew they had no chance of defeating the Lakers with the approach and setup for basketball they were running. Unlike what too many think, there is almost no such thing as pro NBA players "quitting" just because they decide to quit out of laziness or some other stupid reason. They quit if and when it is rational to quit, when it is obvious they have no chance. This is entirely normal and actually reasonable, though neither you nor I might like it.
This same reality largely explains why Carmelo Anthony is considered one of the top half dozen players in the world in basketball when playing internationally, such as at the Olympics, but can not crack the top dozen in the NBA. He doesn't really think the Nuggets can win a playoff series, or at least not more than one series, so his whole effort from season tip-off to season buzzer is ramped down a little from what it would be if he truly thought the Nuggets could win it all.
For the most part, this kind of thinking is not something C Anthony realizes is happening, or realizes is slightly affecting what he does on the court. It is mostly a subconscious thing, except in rare moments such as that game 3 of the series, when it became obvious to Carmelo that neither he nor the Nuggets were trying as hard as they could anymore.
In the NBA, the man is coasting and blaming his coasting subconsciously on the perceived shortcomings of the Nuggets franchise. He probably rationalizes that he is too young and lacking in managerial experience to be able to do much fixing of those complicated problems with the franchise.
But all players including C Anthony know that the USA can win the Olympics, so you are likely to see a 100% effort from all players there.
So Anthony plays better in international because he tries a little bit harder, because he is completely confident the USA can win. The other main reason he is better in international is that he is part of a system that makes more sense than any system he has played under in Denver.
Editorial Note:
This was another preview of an upcoming full report topic, seen here first.
I'm not sure about doing these "spoilers". On the plus side, this is a great way to make important notes on days I get some ideas, but I am not doing much writing. Also, not everyone here reads my full reports, or is it hardly anyone reads the full reports, word for word, lol? (I'll get more readers with the new editing plan.) Those who can't read a lengthy report will see some of my important stuff here instead of never.
On the other hand, anyone who reads complete reports will be getting a little repetition later on. I promise though, that there will be more evidence, more flavor, and more context in the full report than you see in any short post like this.
__________________________________________
And when C Anthony just before the 2007-08 season said that the Nuggets had a very good chance of winning 60 games, there were two main reactions. Most of the Nuggets' brain trust, the Charles Barkley types who don't like Allen Iverson, and anyone down on the Nuggets in general, simply dismissed it and claimed Anthony was exaggerating and not fully realistic. Meanwhile, the most enthusiastic fans of the Nuggets and of Anthony welcomed the prediction and more or less believed it to be realistic.
The thing is, whether subconsciously, or consciously and on purpose, Anthony revealed what he really thought about how the season would go by what he did not say, not by what he said. He didn't say anything about the playoffs! Here you had a team that had for 4 straight years won just one playoff game each year, the ultimate lack of movement and improvement, and yet one of the most important players of that team said nothing about doing better than that in the 2008 playoffs. And with one of the most expensive and talented teams no less.
Anthony was really saying, whether knowingly or not, and probably not, that no matter how well the Nuggets would do in the regular season, it was not safe to predict even winning just one playoff series. Which means he had little if any confidence that the Nuggets could actually win a series. But this revealing assessment by him blew right by everyone during the start of the season festivities.
In other words, Anthony faked everyone out with his deceptively optimistic preseason statement, maybe himself included. Until now.
_________________________________________
The following comments are related to the 2007-08 Season NBA Real Player Ratings:
How good each player is is on each player. How good the players are form the building blocks for a team, and which building blocks are available is the responsibility of the front office and, to some degree, the owner. What is built or not built with the building blocks is entirely the responsibility of the coaching staff.
So if the Nuggets are chock loaded with talented players, thanks to those players efforts, and thanks to the front office and owner getting them, it is a total scandal that they can not win a playoff series, and a humiliation that they could not win a single playoff game. Scandal and humiliation against who now? The coaches, because they had as good or better a set of building blocks than any other coaches did, yet could not do anything with them other than get their full share of obvious regular season wins against lessor and middle level teams.
I don't have the time to do it now, but I would wager I could easily prove the same with any reasonable combination player measurement: that the Nuggets are chock loaded with high performing players.
The more I have worked with this measure, the more I have realized that, even though it is no where near as complicated as the Hollinger per, it is somewhere between slightly and substantially better than it is. I am particularly impressed with the 1.4 factors installed for assists and blocks, which adjusts these critical things so that they are closer in value to a 2-point score than if you just counted them straight up.
______________________________________
A couple of good questions about the Real Player Ratings:
Where is your mathematical proof that this is the, in your words, holy grail of player ratings?
What is the statistical basis for using the multipliers that you have used?
My answers were:
The reason its the holy grail is that there is nothing better I have seen. Most statistics, whether or not weighted, are just counts without giving you the per time measure. The statistics that do give you per time measures are generally one variable, as if everyone is embarrassed to combine items when doing per time. What is there to be afraid of?
The statistical basis is that these are the ESPN factors, and I don't yet believe I have the computer wherewithal to tweak the factors and make my own underlying measure. Right now, my technological capability is such that it would take me 40-100 hours of work to produce this if I tweaked the numbers and made my own statistic, whereas I can do the whole thing, lock, stock, and barrel, in about 5 hours while starting with ESPN. Yet my tweaks would only be moving players up or down, by at the very most, 10 ranks or so.
But I could be grossly underestimating what I could do database wise, since I have never made a full scale effort as of yet, only partial efforts. If only there was a place on the internet where you could simply operate a database without being a veteran database programmer. There probably is, but I have not found it yet!
If and when I get "full database capability," which is one of a few major computer things I am still green on, I would tweak as follows.
The main remaining shortcoming with the Real Player Rating is that it underweights defense to some extent, and it underweights made them miss defending severely, because it does not count it at all. Since made them miss defending can not be exactly measured statistically, and so is left out by literally everyone, the last thing you should do is have relatively small weights on any defensive factors that you can count, so one of the most important tweaks I would do would be to upgrade steals to 1.5. I would also upgrade blocks and assists very slightly to 1.5.
I would bump up turnovers from .7 to between 1.0 and 1.2, because they so often lead to easy scores, and so they are a defensive liabilty. And I would bump up the miss shot subtractions from .8 to 1.0.
As it stand now, the total points awarded for a field goal is 3: the points plus 1.0 per made shot. The points awarded for a three is 3.5: the three points plus .5 for the made three. I see the rationale for the bonuses, but I don't see the rationale for the different bonus amounts, so I would make both of them .7 or .8.
All of these tweaks, as I said, would change player ranks by, at the most, about 10 positions. You can tweak to your heart's content, but you are not going to change your view of reality very much, as long as your initial, basic approach is comprehensive and smart.
Finally, if I was really on a perfection binge, I would attempt to do the near impossible, by installing some kind of adjustment for made them miss defending. One approach I could take would be to start with the pace adjusted defensive efficiency of each team, and normalize that so that the resulting impact on the Real Player Ratings would be up or down by, at the most 10%, relative to the median Real Player Rating, which is about .700. For example, all of the Nuggets would immediately suffer a 4-8% hit on their ratings, lol.
After I did that, I could then profit from putting players into just one of 3 categories: above average made them miss defender, average, and below average. Then the players would get both the team adjustment and the adjustment for their category, which would be about another 8% up, zero, and 8% down.
But if I could ever find any assistance at all on the internet on rating made them miss defending, I could increase the number of categories in which players could go, increasing the accuracy.
I am already using a 38% total adjustment range when I adjust the ratings for the Nuggets, up to 19% up and up to 19% down.
Another approach to tackling the near impossible, if I was attempting to do the entire NBA, while ignoring the team defense measure, would be to simply divide players into 5 categories. The highest adjustments I could confidently do, with no assistance from anywhere, would be: up 10%, up 5%, no change, down 5%, and down 10%.
So in summary, there are two long term projects planned for the Real Player Rating:
1. Conversion to use of my database and custom formula. (Target: 2009-2010)
2. Addition of carefully estimated made them miss defending adjustments. (Target: 2010-2011)
Right now, it's as good or better than anything out there. These two projects would make it close to perfect.
________________________________________
Someone asked about how Leon Powe, an underrated Celtics player, could be rated so high. My answer was:
As long as he stops his share of shots from going in, he was in fact a top 30 NBA player in 2007-08.
There ended up being a lot more forwards than guards with very high ratings, but low minutes. There were 7 forwards and 1 guard among the top 100 players, who did not play at least 20 minutes per game.
This is a strong clue that the guard positions are tougher to excel in than are the forward positions, and/or that there are currently more really good forwards in the NBA than really good guards. It's really two sides of the same coin.
Among the top 100 players, there was a grand total of one guard who did not get to play at least 20 minutes per game: JR Smith. This makes Karl's decision making even more wrong, and it was already just about as wrong as you can get! Thanks though, for getting me to check this, because when it comes to criticizing this coach, I will never be finished until he is gone, or until I switch to the Raptors or something.
_____________________________________________
If a player has coaches who don't really know what they are doing, there is no chance for him or anyone on his team to become an optimized team player, so if that player ends up, in effect, concentrating as much or more on his personal production than on optimizing his team's offense or defense, whether unconsciously or consciously, than how much can you blame him? What else is there for him to do?
Players are payed to play, and coaches are payed to coach. Coaching is more complicated than playing, but being able to play at the NBA level is more rare than being smart enough to coach a team correctly. There are many thousands of people who are smart enough to coach the Nuggets better than they are coached, but no where near as many who can play as well as Anthony, Iverson, Smith, Camby, and Martin.
Players can not be expected to coach themselves, though smart, veteran players can do it to some extent. But it is almost ridiculous to say: "Carmelo Anthony and J.R. Smith, you did not have a clue as to making sure the Nuggets optimized and played like a team, so you are not really good basketball players." No, they are still great players who, by the way, never graduated from college, so they probably could not if they wanted to figure out how to coach the Nuggets.
On the other hand, as I said, great players with a huge amount of experience, such as the Celtics big three this year, are able to instinctively coach themselves to some extent, which reduced the Laker's coaching advantage over the Celtics, which in turn allowed the Celtics to win the series in just 6 games.
____________________________________________
The Nuggets are lacking only perimeter defense, they have interior defense down. They are the worst perimeter defense team in the NBA! It's time for the Nuggets to overcompensate: they need to gamble a little more on defense, go for steals and get right in the face of the 3-point shooters, even at the risk of a few 3-point fouls. How can you lose from gambling when you are the worst perimeter defense team in the NBA? You can't lose. George Karl should be encouraging AI and JR to gamble, within reason, not cringing and biting his nails over violations of his style book.
The Nuggets need to spend damn near half of their practice time on how to defend the outside in general and the perimeter in particular, including, obviously, how to cover and rotate out to the perimeter, to cut down on open threes.
If Marcus Camby is hell bent on getting 20 rebounds, then let him do it, and don't be afraid of Kenyon Martin and/or Carmelo Anthony going way out to contest a three.
The three point shot was instituted not only to award the extra skill needed to make a longer shot, but also to penalize teams who try to play defense on the cheap, by loading up the interior defense. The Nuggets need to realize this obvious reality, and refocus their defense, or continue to be burned by outside shooting.
The Nuggets need to surrender a few more points on the inside, so that they can cut down on what they give up, by a greater number of points, what they surrender on the outside. The defense is badly unbalanced, so job one is to refocus it. Putting in more defensive effort without refocusing it isn't going to work.
________________________________________________
Stoudemire and LeBron James were essentially tied as the best players in the NBA. Why Steve Kerr thought the Suns had to have O'Neal to win it all, over Shawn Marion, is beyond me. They already had enough on that team to win it all. And they had a coach who could do it, too.
Powe was essentially tied with Pierce; the .005 difference is miniscule. Of course, Pierce played more than twice as much as Powe, so he was much more important to the Celtics than Powe was. And if you know that Powe would have fallen off if given more burn, then you know that Pierce is really better. But this tells you that Powe was virtually a star in his role as a reserve. He deserves to be where he is at.
I think my base formula tweaks would put Pierce ahead of Powe by a little, but I simply don't know who is the better made them miss defender. If Powe is a slightly better made them miss defender than Pierce is, then they could be tied again.
Now if you want to determine who is really better between closely ranked players, simply ask yourself, who is the better made them miss defender? Because the tweaks are small potatoes compared to the made them miss defending adjustment that you can make. How many shots did each player prevent from going in the hoop per 12 minutes (or whatever period of time you want)? And how good are these players at being in position to try to prevent a shot from going in in the first place?
_____________________________________________
Consider how stupid it was for Karl to have pressured, over the past couple of seasons, C Anthony to get more rebounds, calling it one of the most important things he must do to be a full success. You have on the Nuggets Marcus Camby, the ultmate rebounding machine, and Kenyon Martin, no rebounding slouch either. You also had, at the time Karl concocted his orders to Melo, Nene, an excellent rebounder, and you also had (until now) Najera, who everyone keeps emphasizing is a defensive oriented player, with defensive rebounding a part of being a good defensive player. You also had and still do have Linas Kleiza, a surprisingly good forward who is certainly good enough to get some rebounds. Why in the hell would anyone think that it would be good for the Nuggets if Carmelo Anthony hangs close to the hoop to get more defensive rebounds?
Remember the spectacle of Camby and Melo fighting over rebounds? What was that all about? It was Camby in effect saying, what the hell are you doing in here, buddy?
By telling Anthony to do this, you are indirectly excusing him in advance for relatively open and wide open midrange and longer twos, and of course threes. How can he learn how to get where he needs to be to be able to contest shots when he is honing in on rebounds? So is it any wonder that the Nuggets became the worst outside defensive team in the NBA, easily victimized by good passing, good assisting, and good outside shooting? You had guards who desperately needed defensive help and you told Carmelo Anthony to go for more rebounds instead of telling him to help out the guards!
I'm not saying he shouldn't go for more rebounds if Nene is out, or if Camby is off, or whatever, but for the Coach to make it a seasons long project for Anthony to get more rebounds was stupid from the Nuggets point of view. If Karl was coaching a bad rebounding team, it would be another story, but he was not. At the time he ordered Melo to get more rebounds or be considered a partial failure as a player, Nene and several other great rebounders were fully available. So it was a coaching error, pure and simple.
But Melo's next team, if it is not that good of a rebounding team, will be very thankful for Melo being able to rebound more, even though it actually hurt the Nuggets for him to learn how to do that.
___________________________________________
Maybe Karl was just trying to get Carmelo into the top 10 in the Real Player Ratings, which was accomplished by him getting more rebounds. Remember that made them miss defending gets you nothing in this or any other rating. Laugh out loud.
