The Defense Thread

[KqWIN]

There we go... some actual grist. Or, almost.

You seem to really struggle with a couple of things. First, I've signposted all over the place that this was an initial data exploration, but you seem to think that I'm either dug into a position here or maybe that I'm unaware that I'm lacking information. It's weird. Second, you seem pretty butthurt by the idea that you actually need to explain yourself (this Pearson reference, without any description of the method, is a bare minimum). It's very little-brother of you.
In a discussion about the Jazz's defensive performance, it seems significant that they surrendered shots in the 3-10ft zone at a higher rate than anyone in the league, and that their opponents converted on those attempts at a rate above league average. And since the Jazz forced shots in the 10-16ft + 16ft to 3pt line zones at a rate that was lower than league average, it seems like there might be something worth poking further into here. To brush this off based on generalities and a Pearson reference while acting like your position is obvious to everybody is ****ing hilarious.

Forgive me, but I think some things are trivial. You decided to "well actually" me when I said that forcing floater range shots is good, but to me that's trivial if you've noticed any trends in basketball lately. But even if you haven't, I also assume that you're looking at the exact same numbers because you're posting them. Like It should be obvious to see that NBA players make 70% shots at the rim and that this is way more than the 45% they make from floater range. I thought everyone knows that rim shots are way more efficient than shots from floater range, but even if they did not, I'd assume that someone with the numbers in front of them could also come to that conclusion.

But lets talk about some of these things.

it seems significant that they surrendered shots in the 3-10ft zone at a higher rate than anyone in the league

I'd agree it's significant, and in a good way. Let me repeat my reasoning. Allowing more shots from 3-10 means less shots from the rim. There is a strong, negative correlation in the numbers, and it's also easy to see why from the games. For the most part, players shoot shots from that range because they wanted to get to the rim, but they couldn't. There is no relationship between allowing shots from 3-10 and allowing longer 2 point shots. It's not in the numbers, and I don't see a basketball reason as to why that would be the case. So while shooting further out from 2 is less efficient, it doesn't seem to be something you can control with your defense. But it's important to remember that even floater range shots are way below league average eFG of .545. So yeah, I do not see forcing floaters as a bad thing. I see it as a good sign because 1) there is a relationship between floater range and rim shots and 2) that difference is massive in terms of efficiency for that tradeoff.

and that their opponents converted on those attempts at a rate above league average

How massive is that difference in efficiency between a shot at the rim and a floater range shot? As noted above, it's about a 25% difference. It's a huge impact. The difference between what opponents shot against the Jazz in floater range compared to average? 0.7% difference. Much smaller in comparison. But as I've said before, there are gains to be made here. You want that percentage as low as possible, but it doesn't pack the same punch as turning shots at the rim into floaters. 25% compared to 0.7%. It's up to you what you find significant, but I choose to stress the large amount of floaters that were taken more than the percentage opponents shot on them.

And since the Jazz forced shots in the 10-16ft + 16ft to 3pt line zones at a rate that was lower than league average

More of the same here for me. There are gains to be made here, I don't want that to get lost. But how significant is it? For 10-16 it's once again less than 1% difference. This does make an impact, but to me is very miniscule. This is especially true when you consider the total amount of shots that come from this range. It's not much to begin with. This zone made up for 9% of shots against the Jazz last year. The difference for 16-3p was much greater at 5%. But, it only made up 6% of the shots Jazz faced. On top of that, I would hypothesize that 16-3p has more random variance for two reasons. 1) The sample size is smaller to begin with and 2) we know that 3FG% allowed is highly variable and random....this may apply to long 2's to a lesser extent. Feel free to dive into that one if you like. But even if opponents make these shots at a higher rate, they are still in the red. As I said, allowing mid range shots is generally a good thing.

All of this shot location and shot performance can be wrapped up in opponent eFG%. I mentioned this before, but the Jazz were fairly average in opponent eFG%. They ranked 13th. Room for improvement, of course, but I don't think the room for improvement is to allow less floaters. If they didn't and allowed more shots at the rim (which is the other side of the tradeoff) they would have been much worse. I think the Jazz need to do a better job a lowering the percentages of shots they do allow, but also defensive rebounding and turnovers were very key factors in their below average total defense.

At this point, I don't even know what this conversation is about. On one hand, you keep retreating and saying things along the lines of "I haven't taken a position I'm just investigating". That's cool and is why I'm offering my opinion. On the other hand, you seem to be really opposed to view and also my reasonings for those views. I'm not sure what you have provided in opposition. Bolding things and changing the font doesn't actually make them more important. If providing the Pearson coefficient to show a negative correlation between shots at the rim and from floater range is the bare minimum, what is the bare minimum for the argument against that? It's one thing to say that the Jazz allow this percentage from this zone, but why exactly is that important? I've never denied that the number you posted is true, but I'm just discussing the importance and impact of that number. You said it seems like these things are important, but why and to what extent? If you're just gonna say....."I'm in explorer mode" or whatever cool. But why so critical if you're not even going to argue against it haha.

 

[NAOS_pt2]

" I thought everyone knows that rim shots are way more efficient than shots from floater range, but even if they did not, I'd assume that someone with the numbers in front of them could also come to that conclusion."
Who the **** has argued ^this^ point with him??

"Allowing more shots from 3-10 means less shots from the rim."
Who the **** has argued ^this^ point with him??

"There is no relationship between allowing shots from 3-10 and allowing longer 2 point shots. It's not in the numbers, and I don't see a basketball reason as to why that would be the case. So while shooting further out from 2 is less efficient, it doesn't seem to be something you can control with your defense."
This is an interesting point. I wish there was an adequate explanation of it, along with some supporting evidence. I guess a quick Pearson Correlation Coefficient reference is all we're going to get (and several posts into the discussion at that).

"...But as I've said before, there are gains to be made here. You want that percentage as low as possible, but it doesn't pack the same punch as turning shots at the rim into floaters."
Again, nobody has argued that with you.

You split the following compound/and statement into two, then responded to each half independently. That's not necessarily the best way to do things. ("In a discussion about the Jazz's defensive performance, it seems significant that they surrendered shots in the 3-10ft zone at a higher rate than anyone in the league, and that their opponents converted on those attempts at a rate above league average.")

At this point, I don't even know what this conversation is about. On one hand, you keep retreating and saying things along the lines of "I haven't taken a position I'm just investigating". That's cool and is why I'm offering my opinion. On the other hand, you seem to be really opposed to view and also my reasonings for those views. I'm not sure what you have provided in opposition. Bolding things and changing the font doesn't actually make them more important. If providing the Pearson coefficient to show a negative correlation between shots at the rim and from floater range is the bare minimum, what is the bare minimum for the argument against that? It's one thing to say that the Jazz allow this percentage from this zone, but why exactly is that important? I've never denied that the number you posted is true, but I'm just discussing the importance and impact of that number. You said it seems like these things are important, but why and to what extent? If you're just gonna say....."I'm in explorer mode" or whatever cool. But why so critical if you're not even going to argue against it haha.

lol you are a total lunatic.

Let's follow up on this Pearson argument in detail later on.

In order to mop up this mess, let's just point out that your whole argument appears to hinge on the quote above that I've emboldened. It could very well be an excellent point, but there are at least two problems with it in this discussion so far. First, it is based on a data transformation that is done after an initial data exploration. I'm going to guess that you know that. Seeing that all I did was an initial data exploration (and I let everybody know that), your first reply could have been something along the lines of 'Those are interesting-but-potentially-misleading results; in order to see the real relationship here, you need to apply the following data transformation after you've completed your exploration....' But of course you didn't do that. Second, you finally referenced Pearson (with no explanation of method or rationale) and gave some numbers 4.5 hours after you initially weighed in on my first post. All the hours in between you spent with your hands on your head exclaiming that you'd explained yourself in detail... when you clearly hadn't. You're exhausting. Maybe you can point out another post in this thread so far that lived up to your standards? I'll go back and have a look, but I'm starting to think that engaging with you is mostly just a drag.

 

[KqWIN]

" I thought everyone knows that rim shots are way more efficient than shots from floater range, but even if they did not, I'd assume that someone with the numbers in front of them could also come to that conclusion."
Who the **** has argued ^this^ point with him??

"Allowing more shots from 3-10 means less shots from the rim."
Who the **** has argued ^this^ point with him??

"There is no relationship between allowing shots from 3-10 and allowing longer 2 point shots. It's not in the numbers, and I don't see a basketball reason as to why that would be the case. So while shooting further out from 2 is less efficient, it doesn't seem to be something you can control with your defense."
This is an interesting point. I wish there was an adequate explanation of it, along with some supporting evidence. I guess a quick Pearson Correlation Coefficient reference is all we're going to get (and several posts into the discussion at that).

"...But as I've said before, there are gains to be made here. You want that percentage as low as possible, but it doesn't pack the same punch as turning shots at the rim into floaters."
Again, nobody has argued that with you.

You split the following compound/and statement into two, then responded to each half independently. That's not necessarily the best way to do things. ("In a discussion about the Jazz's defensive performance, it seems significant that they surrendered shots in the 3-10ft zone at a higher rate than anyone in the league, and that their opponents converted on those attempts at a rate above league average.")


lol you are a total lunatic.

Let's follow up on this Pearson argument in detail later on.

In order to mop up this mess, let's just point out that your whole argument appears to hinge on the quote above that I've emboldened. It could very well be an excellent point, but there are at least two problems with it in this discussion so far. First, your argument is based on a data transformation that is done after an initial data exploration. I'm going to guess that you know that. Seeing that all I did was an initial data exploration (and I let everybody know that), your first reply could have been something along the lines of 'Those are interesting-but-potentially-misleading results; in order to see the real relationship here, you need to apply the following data transformation after you've completed your exploration....' But of course you didn't do that. Second, you finally referenced Pearson (with no explanation of method or rationale) and gave some numbers 4.5 hours after you initially weighed in on my first post. All the hours in between you spent with your hands on your head exclaiming that you'd explained yourself in detail... when you clearly hadn't. You're exhausting.

You have still not explained a single reason why your POV (whatever it is because you just retreat to no opinion ). There is no argument presented that a high amount of floaters is a bad thing and something to concerned with. You seem so bothered by what I’m personally doing, I’m just waiting to see a counterpoint. Your “argument” relies on bolding numbers and swearing that they are more important than I believe.

Have a good weekend, maybe through all your exploring you can come up with an actual argument instead of whining about the Pearson coefficient. Or, you can try to cook up some more personal insults. Those are always funny.

 

[Saint Cy of JFC]

 

[NAOS_pt2]

You have still not explained a single reason why your POV (whatever it is because you just retreat to no opinion ). There is no argument presented that a high amount of floaters is a bad thing and something to concerned with. You seem so bothered by what I’m personally doing, I’m just waiting to see a counterpoint. Your “argument” relies on bolding numbers and swearing that they are more important than I believe.

Have a good weekend, maybe through all your exploring you can come up with an actual argument instead of whining about the Pearson coefficient. Or, you can try to cook up some more personal insults. Those are always funny.

Should I go back and count all the times you said something to the effect of "Anyone who watches can see...." You have absolutely no foot to stand on when it comes to personal insults.

My points have been pretty clear. I did a quick data exploration, and then named a few things I'd like to investigate further. It was you who got yourself twisted into a knot, bro.

 

[NAOS_pt2]

"whining about the Pearson coefficient." LMFAO. That has to be the worst bad-faith description of how I've interacted with that point. Bravo.

 

[LogGrad98]

For a few minutes there I thought one of these guys was @One Brow . Weird.

 

[LogGrad98]

This @NAOS_pt2 guy reminds me a little bit of a guy that used to post here. NOAS or something. Can't really remember.

 

[NAOS_pt2]

This @NAOS_pt2 guy reminds me a little bit of a guy that used to post here. NOAS or something. Can't really remember.

I have, indeed, resembled my former self in this series of posts. Pretty ashamed, actually, since battling on the internet was supposed to be in my past. —Lemme just say that this back-n-forth with KqWIN is definitely one of the more exhausting and ridiculous ones I've ever had on this site. I shouldn't be surprised, either, because I knew this side of him.

For a few minutes there I thought one of these guys was @One Brow . Weird.

I feel like defending the honor of @One Brow here, since he shouldn't be compared to either me or KqKWIN. Dude stays super cool, on-point, and never whines like KqWIN.

 

[HermanG]

I have, indeed, resembled my former self in this series of posts. Pretty ashamed, actually, since battling on the internet was supposed to be in my past. —Lemme just say that this back-n-forth with KqWIN is definitely one of the more exhausting and ridiculous ones I've ever had on this site. I shouldn't be surprised, either, because I knew this side of him.

I feel like defending the honor of @One Brow here, since he shouldn't be compared to either me or KqKWIN. Dude stays super cool, on-point, and never whines like KqWIN.

 

[KqWIN]

This doesn’t even feel like a back and forth tbh. Usually when it’s a back and forth, both sides will present an opinion with rationale. In this “back and forth” one side apparently doesn’t have to provide rationale because they are “exploring” while at the same time outright ignoring the opposing rationale.

I’ve been known to write too many words, and that’s totally fair. I made the comment that allowing a lot of floater range shots is a good thing and tried to keep it short and simple. The modern NBA game is all about certain shots. Floater range shots are not one of them. I think it follows that if you concede a lot of shots that are not the shots NBA teams want it’s a good thing. Shot distribution is zero sum. But apparently that was too vague and patronizing. To me this conversation is not too different from a casual fan asking why teams are shooting so many three’s.

So then another point is brought up, that there are even less efficient shots than floaters. Ok…I talked about that. If that’s a big factor, I think two things need to be shown. 1) it needs to be shown that it’s actually playing out on the court that way. 2) It needs to be shown how it’s actually very important. I didn’t deny there were gains to be made there, but I did downplay them. No argument was made for either point, but I guess that’s ok bc if you’re investigating.

I was ok with that, so said my peace on that. Since the rim is the most important area, it makes sense that it is the point of contention. Generally, when a player shoots a floater range shot he would have preferred the shot at the rim. Offenses want to shoot at the rim, but if defended well they will be forced into a floater. That makes more sense when explaining a high amount of shots from floater range. This especially makes sense for the Jazz bc of their outstanding ratio of shots at the rim. If shots at the rim are low, they must be shot from somewhere else. The Jazz are not allowing shots at the rim, but it means more shots are taken from floater range. That is the tradeoff.

But what if the Jazz are just random and atypical. What about the rest of the league? Ok well it makes sense to look at the correlation between shot zones. I did that. I provided the correlation coefficients. I explained what each of the values meant, and it turns out that there is a strong negative correlation between shots at the basket and floater range shots. OTOH there is no correlation between floaters and other long 2’s. Again, I already provided what that means, so go back and look if you don’t understand. How I can be accused of explaining, I do not know.

On the second point, how important are these trade offs anyways? I posted the numbers to show that turning a shot at the rim into a floater versus turning a floater into a shot from 10-16 has >25x the impact in regards to eFG allowed. This further backs up the idea that offenses and defense are actually competing over this tradeoff versus the smaller, less impactful tradeoff. But let’s say they aren’t and I’m totally off base. The true battle is taking floaters and pushing them back into even worse shots. That situation has to happen more than 25x more to have the same impact as one rim shot that is pushed into a floater. Technically, ever floater range shot is both a failure to push them back further and a success in preventing a shot at the rim. I choose to believe that the former is not as important as the latter because of the heavy difference in EV.

So at this point I feel confident in saying that 1) the effectiveness of defense is shown in the tradeoff between rim shots and floater range shots. 2) This logically makes sense given how basketball is played and is reflected in the correlation data. 3) Even if this tradeoff isn’t what the data and myself are suggesting, the impact of it is so great that even if it is a less common occasion than pushing a floater range shot further away it is still likely more important.

None of this is new argument or rationale by the way. It is all repeated. I’m sure NAOS will have a problem with this and say no rationale was presented even though all the above rationale is repeated. Or maybe it will be too unclear. I don’t care to explain anymore, especially when there is not really any counter argument. As long as one is investigating I guess there should be no expectation of rationale behind their disagreement. I only expect demands to provide rationale that has already been provided.

Honestly though, the most straightforward argument is still the first one. Everybody wants to take certain shots, so if you allow other shots that’s generally a good thing. Everyone is right when I’m too wordy lol. Nobody gives a **** about correlation coefficients!

I do like the bolding and font change. Maybe I’ll add that to my future posts when I don’t what to say

 

[NAOS_pt2]

....^made it through the first line. Genuinely can't believe that data exploration (and then sharing provisional results) is a difficult concept to grasp. Among all the things that such an exercise is and isn't, it's obviously different from an opinion.

And it's obviously different from a process that applies data transformations to test whether any early signals about patterns/causes are real — so anyone attempting to refute the results of a data exploration with the results of a data transformation better be clear that they are adding additional steps to the process. And they should explain themselves. Right away, preferably (rather than 4.5 hours later).

This is all clear as day to anyone who has spent one hour in a data analytics course. (But I guess I'm/we're the idiots here).