Also, in football you only play 16 games, and there is far more variance in schedules faced across all the teams (all teams don't play eachother, for instance).
Durvasa, I knew, it is too hard for you to admit/accept it. I'm sorry, but I want to tell the truth. Haven't I proven that you are totally wrong in my post#158 and post#161? All your stats are based on a false assumption "A team can play the same/similar level of defense against other teams". It is a fatal error, it is a fundmental problem, none can save you. The foundation of your stats is broken. It really doesn't matter if you admit it or not. Sorry. .
True, but after about 10 games or so your rankings and average stats are pretty close to what kind of defense/offense you have. Which just gives more credence to what you CAN do with basketball since you have to play every team more than once.
I thought we had established a long time ago that we would all love to have an upgrade to Rafer at PG, but when the players the rockets would most desire to trade are PG's worse than Rafer an upgrade is unlikely. The funny thing to me is some of the same people both insist we are contenders for a title and say we can't win with Rafer. Personally I think the offensive boost of Scola in the starting line up is adequate when combined with Battier finally shooting a decent % after slumping horribly at the start. The main problem right now is that Rafer has been averaging more points than Tmac over the past 5 games. If Tmac isn't going to average more than 14 a game then we are going to be in trouble regardless. Both Yao and Tmac have taken exactly the same amount of shots, so it isn't Tmac holding back on his shooting. A comparison for the last 5 games: Yao: 43-73 Tmac: 27-73 Rafer: 29-63 So tell me which player has been dragging the offense down because of their shooting?
Blake, very good question, thank you. You always asked good questions. No one showed up here to agree with me, which doesn't mean no one agree with me. No one agree with me, which doesn't mean I'm wrong. There are many good examples in the history, I want to use the old example again here, "Remember, people used to think The Earth is square". In my post#158, I asked durvasa to show me his/her assumptions, which is the foundation of his stats. In his post#160, he gave me one of his assumptions "If Team B plays the same level of defense against both teams ...", this false assumption should be read as "A team can/will/always play the same/similar level of defense against other teams", a false assumption too. In my post#161, I proved that his assumption is false, which means the foundation of his stats is broken. Actually, in my post#158, I already proved that he is wrong. The funniest part is that (if you review this thread and my old thread, you can see), my original intention is trying to prove offense is the dominant factor. This is just a byproduct , durvasa attacked me first by his stats, before that, I've never thought there is a big hole in his stats. I'm not The God (please let me use BrooksBall's words), I can't say 100%, I can't say I'm absolutely right, I may be totally wrong. If I'm right, I must thank him first and thank all of you here. You can call me arrogance, foolishness, ignorance, or whatever. I'm right or wrong, time can tell, history can tell. .
I probably shouldn't respond, but your mangling of my argument is just too offensive for me to ignore. That is not my assumption. You completely misunderstood my point, which was just to establish that if two teams played against similar competition (offense/defense), then you can compare how good they are (defensively/offensively) based on their performance against said competition. Your counter that no team can play exactly the same way on multiple occasions, while obviously true, is irrelevant to that point. We know that it is in fact possible for two teams to face similar level of competition, if over time they play a random schedule in the same league. That is my "assumption". Also, I'd like an answer to these questions: Do you think you know more about basketball then Dean Smith? Do you think you know more about thinking analytically than Daryl Morey?
I told you many times, I did completely understand your point, but you didn't completely understand my point. It really doesn't matter similar, same, or exactly same. It is a fundmental problem. My point is, it is very difficult to separate offense and defense. You told me there is a easy way to separate them. The thing is, you think you did separate them, but actually, you didn't. I mean you couldn't separate them by your current stats method. I may be totally wrong. Also, maybe there is a better way to separate them, but at least, I don't know and I can't see it so far. No, I don't. Also, I don't know their theories, I can't comment. I must say, it really doesn't matter who he/she is. If he is right, he is right. If he is wrong, he is wrong. Who cares. .
Uh, you are totally wrong. Maybe another analogy will make this clearer: If Bob bats .400 against the league, that doesn't tell you how good a batter Bob is because it depends on how good the pitching was. However, if Bob bats .400 against the league while Steve bats .250 against the league, then you can reasonably infer that Bob is a better batter than Steve. Similarly, if Barry has a 2.5 ERA against the Astros, that alone doesn't tell you how well he pitched. Maybe the Astros sucked at batting. But if Barry pitches a 2.5 ERA against the league, while Mitch pitches a 4.5 ERA against the league, then you can reasonably infer that Barry is a better pitcher than Mitch. This concept is trivial to understand, but you've demonstrated that you still don't get it. And this is directly analogous to offense/defense in basketball. Yes, if the Rockets score only 95 points against the T-Wolves (in 100 poss), that doesn't tell you how good their offense was (depends on the Wolves defense). But, if the Rockets score 95 points against the league on average, while the Suns score 110 points against the league on average, then one can infer that the Suns are a better offensive team. And if the Rockets allow 100 points against the Wolves, that tells you nothing about their defense by itself (depends on how the Wolves played offense). But, if they average 100 points allowed against the league, while the Suns average 105 points allowed, then you can reasonably infer that that the Rockets are a better defensive team. To expound a bit more on the "theory": The purpose of any offense is to maximize points scored per possession. The purpose of any defense is to minimize points allowed per possessions. You rate a team's offense and defense based on their ability to do these things. The way you do that is you compare how they perform against similar competition. The teams that can score more points per possessions against similar competition are the better offensive teams, and the teams that allow less points per possessions against similar competition are the better defensive teams. Further, there is no relationship between the two -- i.e. if you know a team's offensive efficiency, that tells you nothing about their defensive efficiency; this can be experimentally verified. In other words, they are separable. Don't take my word for it. I've provided links to resources. Daryl Morey uses the points per possession concept when rating team offense and defense, if you've paid any attention to his interviews. Go to the library and check out "Basketball on Paper" by Dean Oliver, if you want to understand the theory behind it. Check out Dean Smith's "Multiple Offense, Multiple Defense" and read how he uses statistical methods to evaluate teams . There are many things in basketball that are difficult to analyze statistically. How a player contributes to winning, for instance. But figuring out who are the most efficient offense and defensive teams over the course of a season is not one of them.
Let me say this way: (1) Rockets-Wolves' one game result is 95-87 (per 100 poss or whatever) (2) Rockets' seasonal average result is 96.4-88.3 (per 100 poss or whatever) In case(1), you can't say 95 is Rockets' offensive merit, can you? You can't say 87 is Rockets' defensive merit, can you? Basically, 95 is a co-production of Rockets' offense and Wolves' defense, 87 a co-production of Rockets' defense and Wolves' offense. You are right, there is no relationship between 95 and 87. Similarly, In case(2), how can you say 96.4 is Rockets' offensive merit? How can you say 88.3 is Rockets' defensive merit? 96.4 is a co-production of Rockets' offense and all other teams' defense, 88.3 a co-production of Rockets' defense and all other teams' offense. You are right, there is no relationship between 96.4 and 88.3. Also beware that, in the stats data such as 95, 96.4, 87, 88.3, the offense and opponents' defense are fully correlated altogether. They are much more complex than you thought. You can't separate offense and defense from 95-87, similarly, you can't separate offense and defense from 96.4-88.3. You think you did separate them, but actually, you didn't. Actually, you separated 95 from 87, and separated 96.4 from 88.3. You thought it is so easy, but it is not that easy. I think I say it very clearly this time, and hope you can understand it. .
At least our guys are playing with a lot more energy especially Yao now that he doesn't have to stand there in the low post every possession and bang with every other big man from every single team. The balanced offense is much better than watching the two man offense of Yao/T-Mac.
Yes, of course. You can't evaluate any offense or defense without a baseline figure for comparison, for reasons you've (sort of) given. 96.4-88.3, alone, doesn't tell us anything about your offense of defense. Just that you're a good team because you score more than you give up. However, if on average teams are scoring 90 (so, necessarily, also giving up 90) throughout the league, then now we have enough information to work with. Now, if you're scoring 96.4 against the league, that means you're an above average offensive team (everyone else is only scoring close to 90 against the league). And if you're only giving up 88.3 to the league, then you're an above average defensive team (everyone else is giving up around 90 to the league). You can't evaluate a team's offense/defense without comparing to what other teams are doing against similar competition. This is the crucial step that you skipped over in your post.
ok, let's see: average teams: 90-90 (offense-defense) Rockets: 96.4-88.3 Wolves: 98.5-99.8 By you logic, Wolves is a better offensive team than Rockets, it is questionable. "comparing to what other teams are doing against similar competition", this is an assumption, not truth. As I said many times earlier, if you want to use this assumption, you must make sure it is perfect. I proved in my post#161, there is no such a thing called "similar competition". The foundation of your stats is still very very questionable. .
"it is questionable" is not an argument. If those were the actual numbers (they are not) over the course of an entire season, then yes you could definitely argue that the Wolves were a better offensive team. They would also be a much worse defensive team. But, of course, those are numbers you pulled out your backside. These are the actual numbers this season: Rockets: 105.3-101.9 Wolves: 100.9-109.4 And I certainly can conclude from these numbers that the Rockets are both a better offensive team and better defensive team the the Wolves. Nice try. It IS truth, and I've explained why over and over. And contrary to your insistence that you've understood what I'm saying, you clearly have not been able to grasp it yet. If two teams play a random schedule in the same league, then over time the level of competition faced will become more and more similar on average. This is a fact. And you're the only person I've ever come across who is incapable of understanding it. LOL ... you're calling that a proof? No such thing as similar competition? That means you can't rely on W/L to gauge how good teams are. Maybe the Wolves have played a much tougher schedule than us through 52 games. Maybe they are actually better than us. Hell, maybe they're the best team in the league! After all, we can never say that they've played an approximately similar schedule in difficulty to anyone else according to your "proof".
Clearly, my numbers are just examples. I pointed out there is a flaw in your logic. What are you talking about? In my post#158, I already told you that you can't look the offense/defense in detail, you can only look offense/defense in the big picture. Everything you said here are all belong to the big picture. Nothing belong to detail. League average is belong to the big picture, you can't use it in detail. W/L is belong to the big picture too. I want to separate offense and defense in detail, and you told me you can separate them easily. After all these posts, you still didn't/couldn't give me a good answer. Now, do you want mislead me from "detail" to "the big picture"? Does this mean you can't find the solution in "detail"? .
You talked about "similar competition" not existing. Backtracking from that? Good idea. "Nothing belong to detail. League average is belong to the big picture" ... huh? I don't understand this "big picture" / "detail" designation you've come up with. How can I give you a good answer, when you've yet to frame a meaningful question? Let me get this straight. You think you can't separate offense from defense, because how much you score is dependent on the ability of the other team to stop you. Uh, so is rebounding. So is limiting turnovers. So is FG%. All stats are worthless now? Is that your great insight? What you've failed to grasp is that in evaluating an offense, it is not enough to just look at what you score. You have to also look at what other teams in the league are doing. That's true for anything you want to evaluate "in detail". Again, go and educate yourself. Seek out and read the books I mentioned if you're serious about understanding how to analyze offense/defense in detail. Learn from people who know what they're talking about (e.g. Daryl Morey) when it comes to objective basketball analysis. And if you're in school, I would definitely suggest a basic statistics course.
blackbird, I just realized we got really side-tracked. An attempt to steer the discussion back to the original issue. The initial issue was how to determine whether offense or defense is more important to winning. Obviously, we have a disagreement on how to define and evaluate team offense and defense, separately. You keep maintaining that your ability to score points is directly related to the strength of your opponent's defense. This is true. I also believe that it is irrelevant (every team faces about the same level of defense over the course of a season). You are talking about one type of "separability": a team's offense, and their opponent's defense. I'm talking about another: a team's offense, and the same team's defense. If I want to analyze some team's offense and defense separately, all I want is two measures that are unrelated to eachother that fully captures that team's abilities on both ends of the court. And if I want to be anal, I can also standardize the measures (relative to league average points scored per possession) ... it's not really necessary. It is provably true that offensive efficiency and defensive efficiency are unrelated and captures everything thing a team would want to achieve in order to win the game. And it is also provably true that they both correlate to winning equally. End of story. Offense and defense, at the team level, are equally important.
Agree it also gives other players like Rafer/Bonzi more opportunities to get to the basket - even if they miss-it creates offensive rebounding opporunities... I think being involved in the offense helps create energy for the other players, as well. hopefully, this will still be the case in the playoffs
There are at least 2 levels, "similar game by game level competition" is belong to detail. "similar league average level competition" is belong to the big picture. "big picture" means a large scale view or on average. "detail" means play by play or game by game or other detail information. You really need to educate yourself first. Now, everything go back to ground zero again. At the beginning, I want to find out offense/defense which is the dominant factor, therefore I need to separate them first, and I said it is very difficult to separate them. You told me you can do it easily by stats. In my example, team A's single game result is 95-87, seasonal average result is 96.4-88.3. I said "you can't separate offense/defense from 95-87, therefore you can't separate them from 96.4-88.3 as well. You think you did separate them, but actually, you didn't". Post after post, day after day, I said "No, you can't", you replied "Yes, I can. I can't separate 95 from 87, but I can separate 96.4 from 88.3 (is it so-called the separation by you?), you must consider other teams' performance here, because so and so ...". Eventually, you figure out it is a mission impossible. Then you say "wait a minute, I'm talking about a different type of separability." Questions to you, how come there is no way to separate 95 from 87, but there is a way to separate 96.4 from 88.3, which you keep saying. Is there any contradiction? Now you are talking about a different type of separability? Do you really think that, it is necessary for you to tell me that there is a beautiful creative way to separate 96.4 from 88.3? And by the separation you keep saying, how can I know which is the dominant factor or they are equivalent? Why? Another miracle? .
