Sloan Sports Conference

[UNO Gracias]

Has anyone else besides BallSoHard and myself been following this conference?

A little Pelicans tidbit here (for better and worse)
25. Opponents do not shoot when Chris Paul is guarding them. The Clippers star faces the fewest attempts of anyone in the league, according to Goldsberry, and reduces shot frequency from every spot on the floor. In contrast, James Harden faces the most attempts of any wing defender in the league. To give you an idea of how dominant Paul is defensively, Goldsberry compared him to Deron Williams in this visualization. The colored dots display the frequency of opponents' attempts.

​26. Paul allows the fewest points of any backcourt player in the NBA. Who is on the opposite side of the spectrum? The Pelicans’ Jrue Holiday, who edges Shaun Livingston, Jarrett Jack and Mo Williams. The best wing defender using the same metrics, surprisingly, is Mike Dunleavy while the worst is Jodie Meeks.

While sad for Jrue, who is #2 behind CP3? Norris Cole.

http://www.si.com/nba/2015/03/01/mit...icetype=tablet

[BallSoHard]

I thought the Jrue thing was interesting. It makes me wonder if we rely on reputation a little too much.

[UNO Gracias]

You notice who is the 3rd least scored on wing in the NBA? Puddin, aka Mr. Glass, aka Puddle of blub, aka I left my heart in Pheonix but my wallet in the car.

[DroopyDawg]

You notice who is the 3rd least scored on wing in the NBA? Puddin, aka Mr. Glass, aka Puddle of blub, aka I left my heart in Pheonix but my wallet in the car.

[rowen12]

John Salmons is in that top 5 category. Won't pay much attention to this lol

[UNO Gracias]

Why wouldn't you pay much attention to the list? Did Salmons play bad defense when he was here?

[AKpelicanfried]

Why wouldn't you pay much attention to the list? Did Salmons play bad defense when he was here?

Boom!

i don't know the answer to that... i hardly remember him playing much to tell.

[BallSoHard]

Well what's not fair about this list is it doesn't point to who was being defended. In other word, if Jrue is defending the best perimeter player every game, then the numbers aren't exactly comparing apples to apples. It is a start though.


Sent from my iPhone using Tapatalk

[UptownFuz504]

This goes to show how our biases dictate how we evaluate players. The numbers don't lie. Wait.....did I say that?

[NMThreeMVP]

Any analytics list or measurement that gives off the impression Gordon and Dunleavy are better defenders than Jrue Holiday should not only be discounted, but shat on, lit on fire, and thrown by Charles Barkley at Daryl Morey's front door.

[UNO Gracias]

So Cole and Gordon aren't good defenders?

[greewe]

Greivis Vazquez is an elite defender by that metric, as are Eric Gordon and the corpse of John Salmons.

Jrue and MKG are two of the worst defenders in the league.

I applaud the attempt to create better defensive analytics, but this one doesn't really pass the sniff test.

[BallSoHard]

Any analytics list or measurement that gives off the impression Gordon and Dunleavy are better defenders than Jrue Holiday should not only be discounted, but shat on, lit on fire, and thrown by Charles Barkley at Daryl Morey's front door.

I can't really agree abnd interestingly enough this was one of the quotes in the conference:

“When you say ‘I don’t believe the numbers,’ you’re really saying, ‘I don’t believe in the guy that’s telling me.’” That bit of wisdom came from Celtics assistant GM Mike Zarren, a Sloan mainstay and one of the most candid speakers year-in and year-out.

I think when people are dismissive about numbers it's either they're not understanding what's being measured, or they're not willing to admit that what they perceived is wrong. Otherwise, they're just numbers and they're being framed.

[SportsPillowTlk]

Here is the problem with those statistics. If Jrue covers the best perimeter offensive player on the opposing team of course the player will shot/score more. There needs to be a weighted average depending on which player you are guarding. I'm pretty sure Mike Dunlevey or Vasquez is not guarding James Harden during crunch time. I'm a huge supporter of anaylitics, but this needs to be revised, and I don't believe that Jrue is a terrible perimeter defender, maybe not great as some of us think he is, but he is better than average.

[SportsPillowTlk]

There should be a weight of the offensive player that the person is guarding and that should be accounted into this ranking.

[UNO Gracias]

So Cole and Gordon aren't good defenders?

[SportsPillowTlk]

I'm not saying these particular defensive statistics are completely wrong, but they are flawed. If I'm the worst defender on the floor, I'm sure I would be guarding the worst offensive player on the floor, if he never shoots, that doesn't mean I'm a good defender.

[NMThreeMVP]

I can't really agree abnd interestingly enough this was one of the quotes in the conference:

“When you say ‘I don’t believe the numbers,’ you’re really saying, ‘I don’t believe in the guy that’s telling me.’” That bit of wisdom came from Celtics assistant GM Mike Zarren, a Sloan mainstay and one of the most candid speakers year-in and year-out.

I think when people are dismissive about numbers it's either they're not understanding what's being measured, or they're not willing to admit that what they perceived is wrong. Otherwise, they're just numbers and they're being framed.

That quote is asinine and not true at all. Numbers are numbers. I'm not telling you I don't believe in numbers or the the guy that's telling me. What I'm saying is I think the conclusion you're drawing based on your numbers is wrong. I think your interpretation of what the numbers actually mean is flawed.

Numbers can always be manipulated to produce different results. That's all analytics is. Something as simple as who is the best scorer in the league can give you a different answer based on different calculations.

I agree with Barkley to some extent that numbers are a way for guys who never played the game to get involved with the game. But without people who actually know what they're seeing on the court, numbers tell you stupid stuff like Gordon and Dunleavy are better defenders than Jrue Holiday. I think James Harden and Russell Westbrook would disagree. I think Popavich and Carlisle would disagree.

[BallSoHard]

That quote is asinine and not true at all. Numbers are numbers. I'm not telling you I don't believe in numbers or the the guy that's telling me. What I'm saying is I think the conclusion you're drawing based on your numbers is wrong. I think your interpretation of what the numbers actually mean is flawed.

Numbers can always be manipulated to produce different results. That's all analytics is. Something as simple as who is the best scorer in the league can give you a different answer based on different calculations.

I agree with Barkley to some extent that numbers are a way for guys who never played the game to get involved with the game. But without people who actually know what they're seeing on the court, numbers tell you stupid stuff like Gordon and Dunleavy are better defenders than Jrue Holiday. I think James Harden and Russell Westbrook would disagree. I think Popavich and Carlisle would disagree.

What conclusion is that? Is it possible you drew the wrong conclusion? I don't think anyone is saying unilaterally that those guys are better defenders, just like PER doesn't mean you're a better player than anyone else. The metric says that Jrue allows a top 5 amount ranking amongst back court players. That's it.

[N.O.Bronco]

That quote is asinine and not true at all. Numbers are numbers. I'm not telling you I don't believe in numbers or the the guy that's telling me. What I'm saying is I think the conclusion you're drawing based on your numbers is wrong. I think your interpretation of what the numbers actually mean is flawed.

Numbers can always be manipulated to produce different results. That's all analytics is. Something as simple as who is the best scorer in the league can give you a different answer based on different calculations.

I agree with Barkley to some extent that numbers are a way for guys who never played the game to get involved with the game. But without people who actually know what they're seeing on the court, numbers tell you stupid stuff like Gordon and Dunleavy are better defenders than Jrue Holiday. I think James Harden and Russell Westbrook would disagree. I think Popavich and Carlisle would disagree.

I think people who try and dismiss analytics as manipulated voodoo are the people who have never taken an advanced statistics course in their life.

It was exactly what a certain political party did twice when famed statistician Nate Silver predicted the presidential elections almost perfectly twice using weighted polling data. Or how baseball experts laughed at him when he used analytics to predict the Rays would win the World Series.

There are very real limitations to current analytics but there is a reason that the teams that seem to embrace a lot of what analytics has concluded are teams with winning records right now. Played a role in why Dallas and San Antonio have won championships.

For all intents and purposes jrue has not been a consistently good defender this season. He has had moments but how soon do people forget the constant threads and posts whining about his defensive output and players have great nights against him? Or threads complaint that our rim protection is in larger part because of poor defenders at the point of attack. Which puts jrue when in the crosshairs of that argument.

I think another poster pointed out a meaningful flaw and that is I don't think the current formula accounts for who a player is guarding. I may be wrong on this though. Which is why a guy like Cole, Dunleavy and salmons may look good here. Because a lot of time they are playing backups and a guy like salmons had a very low sample size.

[BallSoHard]

I think people who try and dismiss analytics as manipulated voodoo are the people who have never taken an advanced statistics course in their life.

There are very real limitations to current analytics but there is a reason that the teams that seem to embrace a lot of what analytics has concluded are teams with winning records right now. Played a role in why Dallas and San Antonio have won championships.

For all intents and purposes jrue has not been a consistently good defender this season. He has had moments but how soon do people forget the constant threads and posts whining about his defensive output and players have great nights against him? Or threads complaint that our rim protection is in larger part because of poor defenders at the point of attack. Which puts jrue when in the crosshairs of that argument.

I think another poster pointed out a meaningful flaw and that is I don't think the current formula accounts for who a player is guarding. I may be wrong on this though. Which is why a guy like Cole, Dunleavy and salmons may look good here. Because a lot of time they are playing backups and a guy like salmons had a very low sample size.

Well what's not fair about this list is it doesn't point to who was being defended. In other word, if Jrue is defending the best perimeter player every game, then the numbers aren't exactly comparing apples to apples. It is a start though.


Sent from my iPhone using Tapatalk

<3

[MichaelMcNamara]

The hardest part with defensive analytics is determining who is responsible for a guy on the other team scoring. There is no gray area on the other end, but with defense there are different schemes, rotation responsibilities, etc.

[NMThreeMVP]

I think people who try and dismiss analytics as manipulated voodoo are the people who have never taken an advanced statistics course in their life.

It was exactly what a certain political party did twice when famed statistician Nate Silver predicted the presidential elections almost perfectly twice using weighted polling data. Or how baseball experts laughed at him when he used analytics to predict the Rays would win the World Series.

There are very real limitations to current analytics but there is a reason that the teams that seem to embrace a lot of what analytics has concluded are teams with winning records right now. Played a role in why Dallas and San Antonio have won championships.

For all intents and purposes jrue has not been a consistently good defender this season. He has had moments but how soon do people forget the constant threads and posts whining about his defensive output and players have great nights against him? Or threads complaint that our rim protection is in larger part because of poor defenders at the point of attack. Which puts jrue when in the crosshairs of that argument.

I think another poster pointed out a meaningful flaw and that is I don't think the current formula accounts for who a player is guarding. I may be wrong on this though. Which is why a guy like Cole, Dunleavy and salmons may look good here. Because a lot of time they are playing backups and a guy like salmons had a very low sample size.

I've taken advanced statistics. I've also played basketball. And I've been watching everything ...Biddy, HS, AAU, college, and the NBA ...since 85!

I think sometimes analytics can be dismissed as manipulated voodoo. Sometimes they're a valuable tool that can confirm or challenge what is skillfully observed. But it's also true that not everyone has the same level of observation skills. And people who favor analytics tend to be new to the game of BBall or have no extensive playing/coaching experience, and so therefor rely more heavily on analytics, the way that Daredevil relies heavily on his hearing to fight crime!