I think you're mistaking data with data analysis. In your original post you are more questioning people's ability to analyze data that cannot be broken down into simple terms like traditional box score stats and eyes. My point is that you can. Especially NBA teams and their entire analytics department. And to lesser extents people with requisite knowledge who do these for fun or make some money on the internet. Of course there are limitations to this kind of analysis, of which you can search on the internet if you want more detailed explanations. But NBA teams have certainly been making use of these kind of analysis.
Jalen could shoot 20% from the field for an entire season and rocketchamp will call him our best player
1) Who are u? 2) What are u talking about? always confused when some random poster I don’t recognize mentions me out the blue
ah, now I see…u must be one of those random Sengoofy roaches …yeah, you’re wasting your time trying to get my attention or conversate with me
My original post did not go through well then. I did not want to say the problem was that it cannot be broken into simple terms. It is a long and complicated subject, so it is normal that some things will be misunderstood. - Measuring a players impact using team stats is very noisy and also very hard to quantify a player's part in creating that team stat. That is an area where decision makers need to use their advanced knowledge of the game and their 'eye test' much more compared to stats measuring a certain skill, for example a player's performance in certain game situations. It doesn't mean that the stat is completely useless and nba teams naturally use it, but it does't mean that the stat is a very good representation of what it is trying to measure. - If something is calculated with an advanced methodology, it doesn't automatically mean that it is a meaningful stat, i.e. it measures something in an accurate way and is mostly robust. And the thing that is called 'advanced' is really the simplest statistical technique . It is just more advanced than counting stats, but not really advanced in terms of statistics. - By the eye test, I don't mean simply watching the game and deciding something. All the stats are just there to help with your decision making, there will be contradictory information and each stat comes within its context. At the end it is a person to interpret a huge pool of information, and make an informed decision combining these info with their 'eye test'. In a vacuum one player having a better on/off impact stat than the other without watching games and tracking other stats has almost zero information. - Of course nba teams make use of these kinds of analysis. But they wouldn't decide that one player has more impact than the other by looking at this number, which is often what is done over the internet.
I think you are using the reasoning of "this stuff is complicated, therefore I need to trust the people who are knowledgeable at this stuff and let their sets of eyes determine who's good and who's not." But these days, front offices who use analytics are instead saying, "This stuff is complicated. I need to clone myself 1000 times all watching the same game footage and record details of every minutia in order to fully understand how everything work together. But obviously that's impossible. So what we instead do is teach all my knowledge of basketball by putting in parameters that are important to winning to a computer, and have it learn how to utilize these factors to see who's causing the team to win or lose. And said computer can do the work at a million times faster speed without making mistakes." BTW, this isn't just a basketball thing. This is essentially why all the big tech companies are so big, profitable, and taking over the world.
Jalen is struggling hard in November so far.. might be one of his worst months so far in his 3 year career. 15ppg: 15fga on 36%fg 26%3p Jalen used to have like one good game every 10 games.. so far it's been 10 bad games into Nov so maybe tonight's game he turns the corner?
I have a phd in math and work as a data scientist in big tech, so I have a pretty good idea how things are done in the big tech. And it doesn't work like you feed the data into the system and it spits out 'do this', which is similar in basketball, only done with way more elementary methods. Coaches eventually make decisions and their knowledge of the game and what they see plays a huge role in those decisions. You said: 'So what we instead do is teach all my knowledge of basketball by putting in parameters that are important to winning to a computer, and have it learn how to utilize these factors to see who's causing the team to win or lose.' You tell the computer how to decide who is causing the team to win or lose. And your methods are quite simple and my whole point is that these methodologies are quite weak, noisy in determining who is causing the team to win or lose. I have excellent understanding of how those stats are calculated. That's why I don't say 'oh this is very complicated so it should be good'. Actually to the contrary, they are not complicated. And I can see their strong and weak points. This is the discussion I wanted to carry. I tried to stress what the weak points are.
Alright if that's the case then I'd love to hear what you think of this article. This article was an interview with the Rockets GM 15 years ago on essentially what should be a very outdated version of what I'm talking about. And as a long time Rockets fan I know that what he said was not rebuked by any statistician/mathematician since. I know this is an old article, but like I said, these information don't really get out to the public much anymore. So it's not easy to find any NBA inside guy who's willing to divulge so much information like here. I think Cleaningtheglass is probably as close as it gets since it was started by someone who was with the Bucks(?). But feel free to point out why his way of thinking is wrong, because afaik people believe in this stuff. https://www.nytimes.com/2009/02/15/magazine/15Battier-t.html
And he'd be right. We don't have a Go to/A list player yet, we have very good and good players but not a Giannis, AD, Curry type player. However, Sengun and Jalen are our best players. Jalen still has consistency issues and Sengun still has stamina issues, problems playing against players with length and he does not spread the court. Since Jalen can score at all three levels. I would say he's our best player but it really doesn't matter because neither is an A list player
LOL this may literally be the stupidest thing ever said on this forum. If jalen was shooting 20% and ACTUALLY the best player on the team we would likely be in the 2-80 range
Every player CAN score at all 3 levels if we include players like Jalen who is bad at scoring at all 3 levels.... he's really bad at the rim, he's really bad from midrange, and his entire career he's been one of the worst 3 point shooters in the NBA. Typically though we only count guys who can do those things efficiently.
Well you would think its stupid, because you can relate to being stupid. However he's not shooting 20% and we're 10 - 5, with him being the leading scorer in 6 out of those 10 wins.
Do you actually watched the games or do you just like to make comments to make yourself appear to be a basketball fan? I think its the later. If you actually watch the games, I think you will truly understand what is happening on the court. It's a great game!
