see, you can basically fudge the stats until you get an argument one can support. one of the fallacies of stat-geek-dom.
So a stat's predictive value is determined by the general public's familiarity therewith? Hmm...I'm gonna have to go ahead and disagree with you on that one... If you don't understand the stat, how do you know if they're related or not?
No, the value of "irrelevant stats" (its the new imaginary capspace btw, don't know how to do the "tm" font here) is that they are used for specific instances. I think that many of the irrelevant stats can be disproven by other irrlevant stats based on the author's way of phrasing things.
No, the fallacy is the notion you evidenced with your statement. You look at stats as a mechanism to support a conclusion you've already reached. Stats should be used as data points to help arrive at a conclusion not yet determined, or as a tool by which theories and hypotheses can be tested, refined, confirmed, or rejected.
But there are an infinitesimal number of irrelevant stats that one can use. Therefore, the data is worthless.
Stats can be used for anything--to support an argument, to negate one, etc. Statistics are best used to provide insight not obtained by simple observation. The good thing is that numbers don't come with biases. Equations do, but numbers don't. If you build the right equation, you can find an advantage that no one else possesses. It's especially important in a league with a cap or with a team with financial restrictions. There's a tradeoff between production and what you pay for it.
So, because there's more data, it's no good. If there were less data, it would have value. Have you determined at which quantity of data points the data has maximum value? That data would be very valuable to me.
Disagree. Lots of these stats, when translated to real time NBA instances and trades..don't really apply. You can't quantify team chemistry with a statistic really or how well players play together. There are probably some weird +/- stats, but when you try to apply them in the real world...they just won't really work out like you expected. Too many variables from game to game.
Based on what, do they not really apply? You're being very vague. If this is just a longer way of saying, "I don't understand the numbers and am not willing to devote the time to do so," that's perfectly fair. Everyone's allowed to enjoy basketball the way they want to. But you're making assertions without any reasoning or examples. Many teams use advanced statistics in their evaluations, so I fail to see how they don't really apply to the practice of team-building.
I just don't believe you can use obscure statistics to plan a basketball team. I think evaluating players as people and not numbers is more valuable.
So what. People understand Newtonian physics as well - this does not make Quantum mechanics irrelevant. There is a difference between what is true and what is easy to understand. It is easy to understand that Zach scores a lot and rebounds a lot. It is also true that the deficiencies in his game are so big that these two easy to digest stats do not really translate to meaningful contributions in basketball games. ... and we have seen that this is nonsense - because the numbers show that the Blazers are not as jumpshot happy as people claim, and the numbers show that offense is not their problem - it's defense, despite the fact that numbers like "points against" show them as among the best in the league - but when you actually look at pace - you see that it is bullshit. There is nothing easier to understand than team success in basketball. If you do not understand that - you are following the wrong sport.
I think you are confusing two types of data - the on/off numbers can also be related to who you have replacing you - if Zach's replacement in Memphis is ? (I do not know who it is, honestly, but I think Memphis's bench is supposed to be one of the worst in the league) - where LMA was often replaced by respectable bench players (Outlaw, Dante, Howard) - then the on/off numbers are less relevant. I suspect that point differential makes a lot more sense when talking about teams as a whole for comparison sake than pure on/off numbers - if you do not actually consider who is the "replacement" that comes in for you when you are off... Who is replacing him and who is playing against the Blazers when he goes out is a consideration as well. I also think that Win% is a lot less volatile than on/off numbers - since a big blow-out or smack-down can distort on/off numbers - but will be much less of an issue when displayed in the Win% column. Win% rewards consistency a lot better than on/off, for the record. And yet, these numbers do not translate well to wins... It is a conundrum, for sure.
Bullshit. I have watched Zbo play plenty of times this year. He is still Zbo of old. He still doesn't run the court on defense, he still jacks up every shot he can, and he still doesn't rebound on the defensive end of the court, and he still jaws at the ref rather than playing defense when he doesn't get a call. If you want that on your team great. Go watch Memphis.
