Doctor MJ wrote:sp6r I think you're doing great work in this thread.
Thank you.
You've also made clear that you yourself are not part of the "strong RAPM" contingent, but haven't elaborated as to why presumably because you weren't asked. So, why u no strong RAPM?
If my lady and I are discussing where we should go for dinner it really isn’t a big deal. Neither party has to work hard to persuade the other because we eat out enough together that if we end up at a restaurant we don’t like it isn’t a big deal. We suffer through a meal and don’t go back. By contrast a good deal of persuasion is necessary if you’re debating whether you should get married or where you should buy a home. Those are decisions that require a lot of thought because the ramifications of getting it wrong are severe.
The decision to discard the box score from your analysis is one with heavy ramifications that require a good deal of persuasion. The supporters of the strong RAPM hypothesis have never persuaded me that their premise is correct
It would be helpful to begin by describing what the box score. The box score is simply an attempt to count events that occur on the court. The original box score was extremely small. It did not include any of the following:
Offensive Rebounds
Defensive Rebounds
Turnovers
Steals
Blocks
Gradually, over a period of years those stats were incorporated into the traditional box score. The traditional box score was also expanded to include three points made and attempted after the three-point shot was added to basketball. In short, the box score is part of the counting category of stats.
This point was heavily disputed in another recent thread but I am correct and will repeat it: the expansion of the box score is not finished. As an aside, the idea behind Elgee’s Opportunities Created is that by recording what he defines as opportunities created you can improve your individual player evaluation. In that sense he is quite similar to the creators of the original box score who also felt recording individual events could improve your player evaluation.
The video tracking stats that have been created in recent years are all attempts to count individual events that occur on the court. Here are some stats that are included in the video tracking stats category:
Shot charts
Amount of time a ball handler holds the ball
Defensive challenges, etc.
Look at that list and compare it to the list of things that were added to the box score in the 70s. It should be obvious that they are attempting to do the same things: count event that occur on the court. That is why I predict in a very short time video tracking stats will be recognized as belonging to the same category of stats as the box score.
The only important difference between the two stats is in how the statistics are recorded. The new video tracking stats are generally automated while the traditional box score stats are recorded by human beings.
To incorporate the video tracking stats in your analysis but to discard the traditional box score means either (i) the original box score categories have no value but the video tracking stats categories do or (ii) the human error is so substantial that it cannot be trusted. The former hits me as so bizarre that I’m not sure how to respond. I will attempt a response if someone holding that view comes forward. That leaves us with the human error factor.
There is human error in the box score but that in and of itself is not justification alone for discarding the box score. The view that anything that contains human error must be discarded has massive and obvious implications for rethinking everything. By everything I mean literally everything.
The human error justification for discarding the traditional box score is only justified if the human error is so severe to make the box score an essentially unreliable document. Now, the professionalism of early box score keepers is quite questionable but in recent decades the box score recorders have improved dramatically. Corrections are made to the box score after games.
Overall the individual box score as it is recorded now is a reasonable measurement of each individual’s performance in the counting categories. If someone disputes this I invite them to conduct a study of recent NBA games. Due to league pass and youtube, full NBA games are readily available.
That means the only reason to discard the data is if you feel it has no value.
What follows got heavily criticized as simply committing the Appeal to Authority fallacy. It isn’t and I think it would be helpful to describe the Appeal to Authority fallacy. The Appeal to Authority fallacy is as follows
An Appeal to Authority is a fallacy with the following form:
Person A is (claimed to be) an authority on subject S.
Person A makes claim C about subject S.
Therefore, C is true.
http://www.nizkor.org/features/fallacie ... ority.htmlIt is only a fallacy if the expert in question is not an expert on the subject. By contrast it is completely legitimate to use the opinion of experts as evidence for a position.
I really shouldn’t choose such a hot-button topic to make a point like this, but it’s such a good example that I can’t resist. Take climate change. The majority of the climate science community has concluded that the evidence supports the hypothesis that human activity has and will lead to substantial, detrimental changes to our planet’s climate.
Does that prove that climate change is real? No. Proving is something that mathematicians do. It does, however, set the standard for those who believe that climate change is not real. The scientific consensus is prima facie evidence of the truth of climate change. Jo(e) Public is justified, in the absence of the time and skills to investigate for themselves, in believing that climate change is more likely to be true than false. Those who wish to believe that climate change is probably not real have the burden of showing that the scientists are wrong.
http://letterstonature.wordpress.com/20 ... ers-guide/In the U.S. climate change debates people who are concerned about climate change frequently cite the statistic that 97% of peer reviewed articles on the subject support the view that human activity is causing changes to the climate. The users of the statistic are citing it as evidence that human activity plays a role in climate change. That is valid evidence and is not the appeal to authority fallacy
To return to the topic at hand, as best I can tell the people running the NBA still incorporate the box score into their analysis. Admittedly, the internal workings of NBA teams are often a black box but in interviews with coaches and general managers, the box score often comes up. The grizzlies recently hired a general manager who is best known for his work with the box score. Looking at players contracts and the box score it seems fairly obvious that NBA teams still use the box score as a big factor in roster development.
Does that prove the box score has value? No. Is it evidence for the value of the box score? Yes.
The leading figures in the plus/minus revolution often incorporate box score stats in their analysis. It pisses off people a good deal that Englemann removed RAPM and replaced it with XRAPM but that did occur. Mysticcb who probably has the best understanding of plus/minus stats still uses the box score in his all in one stat.
Does that prove the box score has value? No. Is it evidence for the value of the box score? Yes.
From my perspective, the fact that box score still has a lot of value to the people who control the sport is a major factor for why I place a heavy persuasion burden squarely on the strong RAPM crowd. I don’t feel that burden has ever been adequately met.
Finally, here are some stray points. I do not feel the sample size issue is ever fully resolved. NBA careers are extremely short. Very few NBA players reach 50,000 minutes played. The average office worker reaches that point in 6 months. Obviously 50,000 minutes of game-time in the NBA is a lot more valuable for evaluation than office time but the point stands we are dealing with a small sample size.
Coaching decisions play a big role. We discussed it with AI, but 76ers management repeatedly tried to pair him with a secondary scorer and neglected the 3 point line. From a plus/minus perspective it appears he failed with many different players. In reality he failed with one type of player.
I meandered a good deal but I think that describes a lot of the reasons I reject the strong RAPM premise. I do value +/- stats and hope to respond to the team portion sometime this weekend.
