New ESPN stat: Net Points

[SouBeachTalents]

https://www.espn.com/nba/story/_/id/44093220/introducing-net-points-latest-nba-metric-amazing-early-findings

https://espnanalytics.com/nba-net-pts

[Xiao Yao You]

Gobert 29th. Trolls will dismiss this stat immediately :lol

[Real Men Wear Green]

Gobert 29th. Trolls will dismiss this stat immediately :lol
Do you think there are 28 more valuable players than Gobert?

[Xiao Yao You]

Do you think there are 28 more valuable players than Gobert?

I would hope so. Hes getting old. Still appears to be most valuable on his own team :cheers:

[Kblaze8855]

I don’t think anyone needs to get that far down to take issue.




https://www.hostpic.org/images/2503070030140360.jpeg

[Kblaze8855]

Actually I suppose The complaints would start with KAT being third though maybe not. I’ve heard vastly different reports on his play this season and haven’t watched enough to contradict them either way. Heard a lot of praise early but last week I read, he might be statistically the worst defensive player in the league who gets big minutes so I don’t know. I acknowledge being uninformed at the moment.

[tpols]

Interesting things I noticed.

-Sengun is the best player by far on defense? That has to be glitch? Trae Young worst defender in the league. I believe that.

-Jokic and SGA are absolutely lapping everybody else.

-Jokic is the GOAT.

[Carbine]

Any stat that supports Jokic having a better season than SGA I support.

[Neal Romer]

Couple things can be true:

Unheralded role players can be much closer in real value to big popular star name guys than typical fans think.

But it's also true that a random metric saying as much isnt what actually makes it true. Metrics are fallible. Human analysis is fallible. A fallible person can often find the fallible metric to validate their erroneous view.

At the end of the day, the eye test is the best tool for intellectually honest and analytically capable people. Even so, there are no "objective" answers to who's best at this or that anyway. There are only arguments that are either more, or less persuasive.

[Kblaze8855]

Interesting things I noticed.

-Sengun is the best player by far on defense? That has to be glitch? Trae Young worst defender in the league. I believe that.

-Jokic and SGA are absolutely lapping everybody else.

-Jokic is the GOAT.

I don’t know enough about the Rockets this year to say, but there have been times lately where a paint locked big who plays with terrible perimeter defenders has much better than expected defensive impact because he’s the only thing preventing an absolute layup line when he’s in the game. He sits and gets replaced by different kind of defender and even if that defender is similar or better overall, the numbers can take a hit. You replaced somebody who just by being in the way turns layups into pull up jumpers with a guy who might be more
mobile and able to switch…

Against some teams, the defense improves, but against one that focuses on collapsing the defense and getting layups or kicking it out after help comes?

They might get scored on more even with a more mobile defender.

again, I’m not speaking on the Rockets specifically, but I watched it happen with Clint Capella a while back.

anything based on different performing lineups is going to come down to who you get to play with.

You replace Duncan with Robinson he looks less impactful than Camby if you replace Camby with Linus Kleiza.

I don’t know how they account for that in some formulas but I’m sure they attempt to correct for it over time.

[Neal Romer]

I don’t know enough about the Rockets this year to say, but there have been times lately where a paint locked big who plays with terrible perimeter defenders has much better than expected defensive impact because he’s the only thing preventing an absolute layup line when he’s in the game. He sits and gets replaced by different kind of defender and even if that defender is similar or better overall, the numbers can take a hit. You replaced somebody who just by being in the way turns layups into pull up jumpers with a guy who might be more
mobile and able to switch…

Against some teams, the defense improves, but against one that focuses on collapsing the defense and getting layups or kicking it out after help comes?

They might get scored on more even with a more mobile defender.

again, I’m not speaking on the Rockets specifically, but I watched it happen with Clint Capella a while back.

anything based on different performing lineups is going to come down to who you get to play with.

You replace Duncan with Robinson he looks less impactful than Camby if you replace Camby with Linus Kleiza.

I don’t know how they account for that in some formulas but I’m sure they attempt to correct for it over time.

Lol.

But yeah the Sengun thing is clearly just some kind of weird data set. Houston's a good defense overall and I dont doubt Sengun is doing something to help contribute to that, but I dont see what he could be doing markedly different from every other center out there to be some crazy individual defensive force. As you said, these things often just come down to circumstance.

[StrongLurk]

Probably another dumb, unnecessary "stat".

With that being said, Jokic seems to be at the top of lists no matter how people are trying to slice stats these last five year.

Jokic is the 3rd best peak Ive seen besides MJ/Lebron.

[Im Still Ballin]

Dean Oliver is the godfather of analytics. I think Net Points has the most value at the team level.

[warriorfan]

Couple things can be true:

Unheralded role players can be much closer in real value to big popular star name guys than typical fans think.

But it's also true that a random metric saying as much isnt what actually makes it true. Metrics are fallible. Human analysis is fallible. A fallible person can often find the fallible metric to validate their erroneous view.

At the end of the day, the eye test is the best tool for intellectually honest and analytically capable people. Even so, there are no "objective" answers to who's best at this or that anyway. There are only arguments that are either more, or less persuasive.

It’s a combo

All metrics, even the good ones, have some holes and some inconsistencies.

A common thing you see is people pointing out one of the anomalies of a metric and try to use that to discredit ALL data from that metric accordingly, which is pretty silly.

For along time now I have always used eye test first but also reference it with a variety of metrics. I feel like that is the best way to try to attempt to get the whole picture.

Using only metrics is problematic, also going only with eye test can be so too at times. Referencing ISB’s thread on Anthony Davis’ defensive impact according to metrics. Where it seemed like the numbers were considerably under of what we would consider they should be given off the eye test of Anthony Davis’ entire career.

Taking bits of information from many different sources and seeing where the common middle grounds align through all of them is the most accurate method.

[Kblaze8855]

A common thing you see is people pointing out one of the anomalies of a metric and try to use that to discredit ALL data from that metric accordingly, which is pretty silly.



I remember having this conversation when people first started talking about wind shares and I think one of the big issues was something like Hakeem not even leading one of his championship teams in win shares in the playoffs. And I remember one of the metric pushers around that time, insisting that 5 Amir Johnson‘s would beat five Kobe‘s because he had better metrics and it was just letting us know how bad our evaluations were.

When I look at this one, I have to ask the same question I asked the people who said the dozens and dozens of outliers were to be ignored then.

How many of the results can be laughable before they aren’t outliers anymore? Glance up to the image I posted up there. It has 15 names. No less than 10 of them have players listed far from where they would rank to a reasonable person. Hell let’s say it’s only 7. That’s still near half.

If even 30 percent of a metrics findings have to be ignored….are those outliers?

Out of 100 how many would be too many to keep calling it a solid list with a few outliers?

Not even saying you’re saying that about this particular list which I’ve not even gone through in full. I’m just speaking generally.

How many Jeff Foster’s can be Elite players on a list before someone needs to go back to the drawing board?

[dankok8]

It doesn't say how they determine the values of individual points, rebounds, assists, steals etc. I'm guessing some sort of regression analysis but those rarely give very good correlation which is why these single metrics are never that accurate. However, you start using multiple metrics all with different methodologies and if a player grades consistently at the top in all of them, then that player is probably pretty good.

[iamgine]

We get 2024 Nurkic better defensively than 2021 Giannis. What is this suppose to measure?