Statistical Analysis. Humor. Knicks.

Saturday, November 1, 2014

Knicks Morning News (2014.07.28)

  • [New York Times] Sports Briefing | Pro Basketball: James Will Wear No. 23 With the Cavaliers (Mon, 28 Jul 2014 03:35:53 GMT)
    LeBron James will go back to wearing jersey No. 23 in his return to the Cleveland Cavaliers.

  • [New York Times] In Joining Lakers, Lin Narrows Focus and Sheds Status as a Sensation (Mon, 28 Jul 2014 01:27:56 GMT)
    Jeremy Lin is more concerned with finding a starting spot and growing in Los Angeles than with rekindling his banner run.

  • [New York Times] National Basketball Association Roundup (Mon, 28 Jul 2014 01:03:53 GMT)
    Byron Scott said in a television interview that he is the new coach of the Los Angeles Lakers but the team insisted that an agreement has not yet been reached.

  • [New York Times] Sports Briefing | Pro Basketball: James Will Wear No. 23 With the Cavaliers (Mon, 28 Jul 2014 03:35:53 GMT)
    LeBron James will go back to wearing jersey No. 23 in his return to the Cleveland Cavaliers.

  • [New York Times] In Joining Lakers, Lin Narrows Focus and Sheds Status as a Sensation (Mon, 28 Jul 2014 01:27:56 GMT)
    Jeremy Lin is more concerned with finding a starting spot and growing in Los Angeles than with rekindling his banner run.

  • [New York Times] National Basketball Association Roundup (Mon, 28 Jul 2014 01:03:53 GMT)
    Byron Scott said in a television interview that he is the new coach of the Los Angeles Lakers but the team insisted that an agreement has not yet been reached.

  • 74 comments on “Knicks Morning News (2014.07.28)

    1. Brian Cronin

      The Clippers sale is now official. Great news.

      Although it would be sort of funny to see what Woody would have done if Sterling kept the team and Rivers quit.

    2. Kahnzy

      Maybe I’m being selfish here (actually, there’s no “maybe” about it), but I was kinda hoping the sale would be stalled long enough that the league would give the Clipper players an out in their contracts and that we would be able to work some sort of zen voodoo to get Chris Paul in a Knicks jersey.

      Pipe dream though it was, it was still a pleasant one. Oh well. Bring on the Spanish Duo! (Yes, I am aware that Priggy Smalls is Argentinian, but he played in Spain so it’s close enough for me)

    3. ephus

      The Clippers sale is now official. Great news.

      Almost. Not quite yet. Donald Sterling will certainly try to get an emergency stay from the appellate court of the ruling under Section 1310(b) that empowers Shelly Sterling to act as if no appeal was filed. He will also seek an injunction from the federal and state courts where his other lawsuits are pending. I do not expect any of these requests to be granted.

      I have submitted a wrap-up column to CNBC.com. Will post the link if (when) it is published. It would be awesome to get picked up in Knicks Morning News.

    4. JD & the 21-Foot Turnaround Fadeaway Carmelo Shot

      ephus,

      You are the humblest. Picked up by CNBC, hope for a spot in an automated news post.

      Sincerely,

      The Scarcely-Published Cock Jowles

    5. GoNyGoNYGo

      @Ephus – It’s nice knowing we’ve got a legal eagle on this blog.

      It seems like Sterling’s motivation here is to cause as much havoc as possible for as long as he can. What do you think is a reasonable time frame that we can expect for all of his challenges to get resolved? Could he take this into training camp which begins in October? Does the high profile of the case accelerate the wheels of justice?

    6. ess-dog

      Can this work?

      Phoenix gets: Harrison Barnes, TH2, 2 2nd rounders

      Golden State gets: Bargnani, Shump

      NYK gets: David Lee, newly signed Eric Bledsoe

      Sure we then have a logjam at pg, but getting Bledsoe would be worth it. A starting 5 of Bledsoe, JR, Melo, Lee, Cole/Dalembert could be nice. I’d love to upgrade JR of course, but maybe he’ll pull it together?

    7. ephus

      What do you think is a reasonable time frame that we can expect for all of his challenges to get resolved? Could he take this into training camp which begins in October? Does the high profile of the case accelerate the wheels of justice?

      Unless Donald Sterling pulls a rabbit out of his hat and gets an emergency stay from the appeals court or an injunction from the federal district court or the California Superior Court (all highly unlikely), this should be over and done by August 15. Pierce O’Donnell (Shelly’s attorney) has already said tonight that he expects the deal closed by August 15 – the deadline under the contract with Ballmer.

      Once the deal closes, Donald Sterling can continue his lawsuit, but he can never get the team back. At most, he can get money damages from Shelly, Adam Silver and/or the NBA. And – in reality – he cannot get that because Shelly agreed to indemnify the NBA (and Silver) for any damages, including attorney’s fees, from Donald’s lawsuits. Since California is a community property state (meaning the husband and wife own all property in common), it is as if Donald Sterling is suing himself.

    8. ephus

      Can this work?

      Phoenix gets: Harrison Barnes, TH2, 2 2nd rounders

      Golden State gets: Bargnani, Shump

      NYK gets: David Lee, newly signed Eric Bledsoe

      No – as a matter of salary cap mechanics. Because the Knicks are over the Apron, they cannot receive a newly-signed player in a sign and trade. So, Eric Bledsoe could not be sent to the Knicks. That is before thinking about if the salaries match, which I do not think would work either.

    9. GoNyGoNYGo

      Z-Man, I think that both Hardaway and Shumpert are decent pieces but we need to see growth. Shumpert has to find an offensive game and Hardaway needs to become a strong defender and contribute to ball movement. I like that he seems to be developing a mid-range jumper and I would love to see him use that athleticism to drive to the rim.

    10. KnickfaninNJ

      Can this work?

      Phoenix gets: Harrison Barnes, TH2, 2 2nd rounders

      Golden State gets: Bargnani, Shump

      NYK gets: David Lee, newly signed Eric Bledsoe

      I have to ask this: Why in the world would Golden State think it’s getting good value by trading David Lee and Harrison Barnes for Bargnanin and Shump?

    11. lavor postell

      The fear is real. The gurus at Stats LLC, the company behind the SportVU cameras, have developed two previously unreleased metrics designed to measure the amount of attention an offensive player gets from defenders when he doesn’t have the ball.

      The first, dubbed “gravity score,” measures how often defenders are really guarding a particular player away from the ball. Korver had the fourth-highest score, behind only Kevin Durant, Carmelo Anthony, and Paul George.

      http://grantland.com/features/kyle-korver-nba-atlanta-hawks/

      I thought this was cool. The whole piece is awesome.

    12. Z-man

      GNYGNYG,
      I may be in the minority opinion here, but I think concerns about Shump’s offense are overblown. Maybe I’m overly committed to him because I advocated strongly for drafting him, but I think he has enough offensive tools to be an asset on that end in the right system with the right role. His biggest weaknesses are shot selection and dribble-drive penetration/finishing. However, he shot 41% from 3 over an extended period after coming back from ACL surgery. I think he depends more than most on being paired with a strong offensive PG, which he clearly hasn’t yet. I am hoping that we get a chance to see what he can do playing with Calderon, and even Larkin. Prigs is efficient, but so unlikely to shot anything but a wide-open 3 late in the clock that he does nothing for Shump. Alas, a trade is always lurking.

      As to Hardaway becoming a “strong” defender, I’d settle for adequate (same for rebounding and passing). He was as much of a one-trick pony as Novak last year (unless you count finishing as a second trick!) He definitely looked physically stronger in summer league. PJ Hairston seemed to win the physical battle with TH2, but Hairston looks like a NBA-ready body physically. I think the USA team experience will be great for him in this regard. He’s trying to distinguish himself, and is working with some great coaches, including Thibs.

    13. Hubert

      Dang, what’s not to love about this kid? I think he’s gonna make a big impact this year, he’s obviously smart and committed to improving his overall game. The rebounding, defense, ball handling and passing all need major improvement, but he has the athleticism, intelligence and drive to make for a very promising wing player.

      Talk about answering your own question!!

      As nice (and surprising) as his shooting was last year, I think we are a long way from Tim Hardaway Jr being a player that a good team would want to carry in their rotation. There are just too many areas he is deficient in, and he is extremely deficient in them.

      Fortunately for him, he doesn’t play on a good team!

    14. flossy

      I think we are a long way from Tim Hardaway Jr being a player that a good team would want to carry in their rotation.

      Really? So I suppose Coach K and Thibs invited him to practice against Team USA for what, comic relief? Out of pity?

      I think the “Tim Hardaway sucks at everything except volume shooting” meme is way overblown. Not that many rookies, particularly ones drafted at the end of the first round, can stand out from day one as a bright spot in an otherwise horrible season/dysfunctional team, and demonstrate–even with minimal (or even detrimental) coaching–clear cut above-average NBA caliber skills in any area of the game.

      Sure, I think it would have been better if he’d come out as a superb all-around player who can score, pass, board, defend, etc., but that’s just not realistic expectations of the #25 overall pick in his rookie year. Aside from his good rookie year, he’s got all the tools–smarts, worth ethic, body–to continue to grow. Yet many here are ready to consign him to the realm of one-dimensionality forever, while still championing Shumpert, who has spent 3 years showing he is injury prone and bordering on zero-dimensional (save for the 45 games of good spot-up shooting that subsequently disappeared, and the sporadic displays of great one-on-one defense).

    15. DRed

      As nice (and surprising) as his shooting was last year, I think we are a long way from Tim Hardaway Jr being a player that a good team would want to carry in their rotation. There are just too many areas he is deficient in, and he is extremely deficient in them.

      Fortunately for him, he doesn’t play on a good team!

      Yeah, I’m not super high on Timmy, but this team isn’t winning anything next year, so it’s not a bad idea to see if he can raise his passing, rebounding, and defense from “uniquely terrible” to “merely bad”. If he can, and his efficiency improves a tick, you’ve got yourself a nice player. And one that would be an attractive trade piece for a lot of teams.

    16. ephus

      One of my takeaways from Zach Lowe’s awesome (as in it inspired awe in me) on Kyle Korver is that any wing player with an NBA body and a high basketball-IQ can become a competent defender with effort and practice. I think that Tim Hardaway Jr. has both the smarts and the (improved) body to become a competent defender if he makes it a priority. He also could become a one-way player like James Harden, but at a much lower level.

      This is one area where coaching will matter. Hardaway is young enough that he is still learning new habits. If he absorbs the Phish defensive system, he could be a real asset.

      _______

      On a separate note, I am astounded that in 2003, the Nets sold off Korver for $125,000 in order to raise the capital needed to have a summer league team (and a photocopier). This was in the last months of the YankeesNets ownership of the team, which means that Steinbrenner/Vanderbeek/Goldman Sachs would not contribute $125k of capital to fund the summer league.

      ________

      The Sportsvu stats that Lowe cites in his article – “gravity score” and “distraction score” – are the first attempts I have seen to quantify how much a player contributes to “spacing.” [Paging THCJ] Gravity score measures how closely a player is guarded off the ball. Distraction score measures how often a defender cheats off a player to help guard the ball. The players with the four highest gravity scores last year were (1) Kevin Durant, (2) Carmelo Anthony, (3) Paul George and (4) Kyle Korver. Korver had the lowest distraction score (low means no one cheats off you) last year.

      _________

      I know that it is not cool to praise Bill Simmons around here, but his hire of Zach Lowe away from Sports Illustrated was truly inspired. Since then, he has given Lowe the bandwidth (literally) to write articles that greatly expand my knowledge of basketball.

    17. johnno

      ” The players with the four highest gravity scores last year were (1) Kevin Durant, (2) Carmelo Anthony, (3) Paul George and (4) Kyle Korver. ”
      Ephus — As I am sure you are aware, any stat that mentions Melo as being anywhere in the same conversation as Durant will be mocked, ridiculed and dismissed as invalid ’round these parts — especially if he ranks above LeBron.

    18. johnno

      “I know that it is not cool to praise Bill Simmons around here”
      For the record, I really like Simmons. I sometimes disagree with his opinions, but I almost always enjoy what he has to say (especially his columns and mailbags; not as much on TV).

    19. Brian Cronin

      I think the “Tim Hardaway sucks at everything except volume shooting” meme is way overblown. Not that many rookies, particularly ones drafted at the end of the first round, can stand out from day one as a bright spot in an otherwise horrible season/dysfunctional team, and demonstrate–even with minimal (or even detrimental) coaching–clear cut above-average NBA caliber skills in any area of the game.

      Pretty much every rookie who played any significant minutes at all in the NBA’s history contributed more than Hardaway Jr did outside of shooting. It is not that he is not a good rebounder, defender or passer. It is that he was historically bad last year in areas other than scoring. Not just bad, but in the bottom rung of all NBA rookies ever. Whenever the term “historically bad” can be accurately applied to you as a player, that’s pretty much a guarantee that criticism of you is not “overblown.”

      But he certainly can shoot and that’s definitely something. Seriously, it is a very good thing. And if the Knicks were stocked with lots of great two-way players (like, say, a good national team should be) then the current form of THJ would fit fine. Since they’re not, the current version of THJ had better quickly become a much-improved version of THJ.

    20. Brian Cronin

      Regarding gravity scores, that was pretty much the whole reason why the ISO offense was so moronic. Melo obviously draws a lot of attention, so late in the game come up with something where a player can get open, dumb coach!

    21. ephus

      I think the Knicks inability to feed Carmelo Anthony in the post when he was fronted was a large part of the reason for the iso-Melo down the stretch. While an iso-Melo was not efficient, it was less likely to lead to a turnover than Felton trying to force the ball into a fronted Carmelo.

      A much better solution would have been to swing the ball to the weak side and feed Carmelo going back door. That is what I expect to see often as the Knicks run the Triangle.

    22. DRed

      It’s interesting to know “[t]he players with the four highest gravity scores last year were (1) Kevin Durant, (2) Carmelo Anthony, (3) Paul George and (4) Kyle Korver”, but we don’t even know what gravity score is. I like Zach Lowe a lot, but that’s a terrible way to introduce a stat.

    23. Hubert

      Really? So I suppose Coach K and Thibs invited him to practice against Team USA for what, comic relief? Out of pity?

      I think the “Tim Hardaway sucks at everything except volume shooting” meme is way overblown. Not that many rookies, particularly ones drafted at the end of the first round, can stand out from day one as a bright spot in an otherwise horrible season/dysfunctional team, and demonstrate–even with minimal (or even detrimental) coaching–clear cut above-average NBA caliber skills in any area of the game.

      I don’t think we’re saying anything different, flossy. I believe he showed promise and agree that he was a bright spot and outperformed his draft position. I just think he needs a LOT of work on his deficiencies and I don’t expect miracles. I think within 4-5 years he could round himself out and become a top 8 player on a very good team. That’s why I said I think we’re a long way away.

      One of my takeaways from Zach Lowe’s awesome (as in it inspired awe in me) on Kyle Korver is that any wing player with an NBA body and a high basketball-IQ can become a competent defender with effort and practice. I think that Tim Hardaway Jr. has both the smarts and the (improved) body to become a competent defender if he makes it a priority

      I agree with this, too, but it also supports my feelings on TH2. Kyle Korver is heading into his 12th NBA season. It took him a while.

    24. stratomatic

      The fear is real. The gurus at Stats LLC, the company behind the SportVU cameras, have developed two previously unreleased metrics designed to measure the amount of attention an offensive player gets from defenders when he doesn’t have the ball.

      The first, dubbed “gravity score,” measures how often defenders are really guarding a particular player away from the ball. Korver had the fourth-highest score, behind only Kevin Durant, Carmelo Anthony, and Paul George.

      The Korver argument can be made (to a lesser degree and for more reasons) to a guy like Novak. Even when he was having one of his games where he didn’t score much, that was only because the opposing team game planned to make sure he never got an open look beyond the arc. There has to be some value in that.

      I saw one study a few years back that looked specifically at the value of “spacing”, but I don’t think I’ll be able to locate it. The whole spacing argument is also related to this.

    25. stratomatic

      Ephus — As I am sure you are aware, any stat that mentions Melo as being anywhere in the same conversation as Durant will be mocked, ridiculed and dismissed as invalid ’round these parts — especially if he ranks above LeBron.

      OK. I’ll do the honors even though I have no idea at all how it is calculated. :-)

      I’m going to guess the stat does not control for the other scoring options on the court with the player in question.

      James doesn’t get as much attention as Melo because he’s on the court with Wade and Bosh. That’s 3 high usage efficient scorers on the court at the same time. Melo is on the court with Prigioni, Chandler, Shumpert, Martin, Bargnani, JR Smith, and Felton. Not much to fear there. So you can focus heavily on Melo. About the only other distraction would be Amare, but only when he’s healthy…which is almost never.

      That also probably explains Paul George’s very high rank. The Pacer’s also lacked enough high caliber scoring options.

    26. ephus

      The Korver argument can be made (to a lesser degree and for more reasons) to a guy like Novak. Even when he was having one of his games where he didn’t score much, that was only because the opposing team game planned to make sure he never got an open look beyond the arc. There has to be some value in that.

      Korver commands much more attention than Novak because:

      1. Korver is so good at the catch and shoot off of the curl, while Novak is really only dangerous off of the stationary catch and shoot.

      2. Korver has a much quicker release than Novak, which means that defenders have more time to recover to Novak.

      3. Korver can put the ball on the floor and hit the mid-range floater, while that is not in Novak’s arsenal. As a result, teams can close out much harder on Novak, which means they can have defenders cheat further away from him. Korver also has become a much better passer off of the bounce, which means he makes teams pay for sending a second defender if he gets past the hard closeout.

      When the Heat eliminated the Knicks in 2012, they were able to eliminate Novak’s effectiveness without keeping a defender in close contact. Wade, LBJ and Battier were able to close out hard and force Novak to reset the offense with the shot clock winding down.

      Novak certainly has value as a floor spacer. But Korver is at a different level.

    27. KnickfaninNJ

      Bill Simmons reminds me a bit of Hunter S Thompson in the sense that they both write stuff that’s fun to read, but you couldn’t imagine yourself actually thinking like they do.

    28. lavor postell

      OK. I’ll do the honors even though I have no idea at all how it is calculated. :-)

      I’m going to guess the stat does not control for the other scoring options on the court with the player in question.

      James doesn’t get as much attention as Melo because he’s on the court with Wade and Bosh. That’s 3 high usage efficient scorers on the court at the same time. Melo is on the court with Prigioni, Chandler, Shumpert, Martin, Bargnani, JR Smith, and Felton. Not much to fear there. So you can focus heavily on Melo. About the only other distraction would be Amare, but only when he’s healthy…which is almost never.

      That also probably explains Paul George’s very high rank. The Pacer’s also lacked enough high caliber scoring options.

      What about Durant? He plays with Westbrook and Ibaka who teams do have to pay attention too.

      I think there’s definitely truth to what you’re saying and without knowing the specifics it’s hard to say this is a comprehensive stat. That being said I do think it illustrates that Melo does command the attention of opposition defenses even when he isn’t directly involved in a play.

      Would be a lot more interesting to discuss if we knew the methodology behind this statistic as well as if the information stretched back beyond this past season.

    29. Brian Cronin

      Would be a lot more interesting to discuss if we knew the methodology behind this statistic as well as if the information stretched back beyond this past season.

      It irks me so much when we’re not given the methodology about stats. Even the great Nate Silver did that the other week with a new form of basketball PECOTA he came up with.

    30. mokers

      an enterprising sports journalist or interested fan could probably email stats llc to see if they would give more info on gravity and distraction. Lowe’s article had a footnote saying it was calculated using baseline tendencies, so there might be some correction going on. Definitely would be cool to know more.

    31. Hubert

      Re:

      It’s interesting to know “[t]he players with the four highest gravity scores last year were (1) Kevin Durant, (2) Carmelo Anthony, (3) Paul George and (4) Kyle Korver”, but we don’t even know what gravity score is. I like Zach Lowe a lot, but that’s a terrible way to introduce a stat.

      and

      It irks me so much when we’re not given the methodology about stats. Even the great Nate Silver did that the other week with a new form of basketball PECOTA he came up with.

      In fairness to Lowe, only NBA teams who purchased SportsVU have access to its information. And it’s information is actually extremely raw, so they have to make it into something coherent themselves. So Gravity Score is most likely a proprietary stat developed by a specific NBA team in order to gain a competitive advantage. If they share the methodology so Zach can publish it, the advantage they gave themselves is gone.

      I’d say full credit to Zach for developing sources that offer us even a glimpse of something that seems like it’s pretty far ahead of what we have available. Let’s understand his limitations a little bit.

    32. Hubert

      Regarding how raw the data is, this is how SportsVU works according to another Lowe article in 2013:

      “The most important innovation in the NBA in recent years is a camera-tracking system, known as SportVU, that records every movement on the floor and spits it back at its front-office keepers as a byzantine series of geometric coordinates. Fifteen NBA teams have purchased the cameras, which cost about $100,000 per year, from STATS LLC; turning those X-Y coordinates into useful data is the main challenge those teams face.”

      http://grantland.com/features/the-toronto-raptors-sportvu-cameras-nba-analytical-revolution/

    33. Farfa

      James doesn’t get as much attention as Melo because he’s on the court with Wade and Bosh. That’s 3 high usage efficient scorers on the court at the same time.

      I would also say that James has the ball in his hands at the beginning of possessions much more ofter than those guys (Melo included). So that stat probably doesn’t apply to him because to be guarded far from the ball, there has to be someone else bringing the ball down.

    34. DRed

      The problem with the stat is that we can only speculate because Lowe told us basically nothing about it. Maybe James is 5th. Maybe he’s 50th. Maybe Carmelo and Durant are an order of magnitude above George and Korver. Who knows? We don’t even know what scale the score is on, or what these guys scored on it.

      Melo and Durant and Korver are all great 3 point shooters. Durant, Melo and George are all high usage guys who play on teams that, at least anecdotally, did not run sophisticated offenses. The 4 don’t really have anything in common.

      And finally, is having a high gravity guy good? It seems like it would be, but what’s the effect of it? Does even SportsVu think they know? If not, why are you even bringing it up? Lowe is writing an article about how good Kyle Korver is, and he’s got an exotic stat that makes Kyle Korver look really good (something, by the way, I totally agree with), so I understand why it’s in his article, but it really doesn’t tell us anything. The fact that the stat is mysterious gives it an illusion of authority.

    35. Kevin Udwary

      I agree completely DRed. Unfortunately SportsVU data isn’t public, so the only insight all of us hobbyist analysts are able to glean from it is through sportswriters who have no statistics training, or the rare academic papers. For us fans it’s going to muddy the picture more than to help clarify the game.

    36. Z-man

      I agree. Clearly some players get more attention than others but the offends still has to capitalize posession by possession. If the result is the more closely guarded guy shooting anyway, who cares? Ironically, the Spurs wouldn’t likely have anyone at the top of the list because their offense is totally based on creating spacing.

    37. stratomatic

      Novak certainly has value as a floor spacer. But Korver is at a different level.

      I think that was pretty much what I said also. :-)

    38. Hubert

      The problem with the stat is that we can only speculate because Lowe told us basically nothing about it. Maybe James is 5th. Maybe he’s 50th. Maybe Carmelo and Durant are an order of magnitude above George and Korver. Who knows? We don’t even know what scale the score is on, or what these guys scored on it.

      I don’t think his inclusion of the stat was designed to say anything about who the best players in the league are, and I’m surprised to see him getting flak. I think it makes sense that Melo would draw more defenders than LeBron. He holds the ball for a long time and constantly takes shots when double teamed. And I’m not sure how where he ranks compared to LeBron is relevant.

      I think it’s pretty remarkable that Kyle Korver draws nearly as much attention as guys like Durant and Melo, and that was the only reason he introduced the information into the article

    39. EB

      @ Hubert

      I think the problem with gravity is that we don’t actually know what the statistic is saying therefore we can’t make any inferences with gravity. If I say name the best scorer on the Knicks then the lay fan would name, without a thought, Carmelo Anthony, but DRed and THCJ would name Tyson Chandler. Best scorer is confusing because it could mean a raw point total or it could mean efficiency or it could mean a combination of both. The semantics of the statistic when looked at in a technical way will likely differ from the way in which Lowe presents them in a popular article. Without insight into what the statistic is actually measuring then we can’t make any claims about what it means, or really use it to add worth to a player.

    40. Brian Cronin

      Especially since Lowe had already done such a great job describing Korver’s importance, why even throw in the new stat?

    41. Kevin Udwary

      I can think of a couple questions with respect to this Gravity score off the top of my head. I imagine it’s calculated as a percentage of time a defender is a certain distance from the player, but how far? Does it have to be the players primary defender, or any opposing player being that distance away? Is it real-time calculated, or calculated in chunks of time? Is it only calculated for perimeter players? I would think that the players with the highest gravity scores would be post players, but that’s not what they’re claiming. There are a ton of questions, but none can be answered because the methodology of this stat is a complete mystery.

    42. johnno

      “but DRed and THCJ would name Tyson Chandler. ”
      Which is why it is so stunning that Melo scores so high in the “gravity ratings.” One would think that NBA coaches and players would realize that Chandler was the best player on the Knicks — by far — and that they would realize that Chandler is the ONE PLAYER on the team that you have to watch at all times because he is such a deadly efficient scorer. Amazing how stupid they all are guarding Melo when he doesn’t have the ball. What a waste of time and energy…

    43. DRed

      Melo should absolutely be guarded when he doesn’t have the ball. He’s a remarkably good open shot maker.

    44. DRed

      I don’t think his inclusion of the stat was designed to say anything about who the best players in the league are, and I’m surprised to see him getting flak. I think it makes sense that Melo would draw more defenders than LeBron. He holds the ball for a long time and constantly takes shots when double teamed. And I’m not sure how where he ranks compared to LeBron is relevant.

      I think the two stats he mentioned are supposed to measure how defenses react to players who don’t have the ball. It’s supposed to show how much defenses fear Kyle Korver. Or maybe it shows that Kyler Korver sucks at getting away from his defenders. Or something else. While I understand Zach probably can’t legally explain the stat, saying “The methodology is complex, incorporating specific location data from every recorded NBA possession and factoring in baseline player tendencies” doesn’t tell us anything.

    45. Hubert

      I think the problem with gravity is that we don’t actually know what the statistic is saying therefore we can’t make any inferences with gravity. If I say name the best scorer on the Knicks then the lay fan would name, without a thought, Carmelo Anthony, but DRed and THCJ would name Tyson Chandler.

      But we do know what it’s saying:

      “gravity score,” measures how often defenders are really guarding a particular player away from the ball. Korver had the fourth-highest score, behind only Kevin Durant, Carmelo Anthony, and Paul George.

      It’s a fact, a piece of information.

      What we don’t know is how it can be used to factor into the endless debates we have about Carmelo Anthony.

      But the article doesn’t give a shit about whether or not Melo is better than Chandler, it’s only focused on making a point about Korver, and it used that information extremely well to make a revelatory conclusion.

      At no point did he say, here’s this newfangled stat and here’s what it says about all the players in the league and how good they are. He’s just making a point about Kyle Korver, and he used evidence that no other writer had access to in a credible manner. Again, I think it’s astounding that Korver draws as much attention as Melo & Durant, and that’s really all he was trying to communicate.

    46. Hubert

      While I understand Zach probably can’t legally explain the stat, saying “The methodology is complex, incorporating specific location data from every recorded NBA possession and factoring in baseline player tendencies” doesn’t tell us anything.

      DRed, that footnote pertains to “distraction score”, not gravity score.

      There are two stats he introduced, and he only mentioned how Korver ranks in gravity score, which he did explain.

      The gurus at Stats LLC, the company behind the SportVU cameras, have developed two previously unreleased metrics designed to measure the amount of attention an offensive player gets from defenders when he doesn’t have the ball.

      The first, dubbed “gravity score,” measures how often defenders are really guarding a particular player away from the ball. Korver had the fourth-highest score, behind only Kevin Durant, Carmelo Anthony, and Paul George. The second — “distraction score” — is a related attempt to measure how often a player’s defender strays away from him to patrol the on-ball action. Korver had the lowest such score in the league.1

    47. Kevin Udwary

      It’s a fact, a piece of information.

      No no no no no no no no no no! We don’t know that at all, because we have no visibility on the methodology.

    48. DRed

      For all we know, Carmelo, KD and Paul George could have a 99 in gravity score, while Kyle Korver is 4th in the league with an 78.2, and is followed by 9 guys with a 78.1. Now that you know Kyle was 4th in gravity score last season, what can you tell me about his game that I didn’t already know? Did his being 4th in gravity score make his teammates better? Did it make him better? Worse? If I’m a GM, how much more or less should I want to pay a guy based on his gravity score? Does it carry over from year to year?

    49. Hubert

      Kevin, the existence of gravity score and where Kyle Korver ranks in gravity score are facts regardless of whether or not you know the methodology behind gravity score.

      Again, the point of the article, and the inclusion of that statistic, seems like it’s being ignored in this conversation, and I don’t know why. If I’m coming off as Zach Lowe’s RuRu, I apologize, but I think its inclusion was insightful and well used in the context it was intended for and don’t understand the criticism.

    50. Kevin Udwary

      Kevin, the existence of gravity score and where Kyle Korver ranks in gravity score are facts regardless of whether or not you know the methodology behind gravity score.

      But for all we know, gravity score could be calculated with a dice roll. I’m not saying that it is a bullshit statistic, but I am saying we have no idea because the methodology is hidden. It’s best not to draw conclusions from unverified data.

    51. Hubert

      Did his being 4th in gravity score make his teammates better? Did it make him better? Worse? If I’m a GM, how much more or less should I want to pay a guy based on his gravity score? Does it carry over from year to year?

      All irrelevant to the point he was making in his article.

      The point is Kyle Korver attracts a lot of attention. The inclusion of gravity score wasn’t necessary to make his point. He showed some clips of him attracting attention, included a few quotes from coaches talking about how much attention he attracts, some photos of how his positioning creates space. Then he offered a glimpse of a burgeoning new source of data that tells us something we might not have imagined. It wasn’t integral to his argument, it was just a little treat to give us a peak behind the scenes of some more things that may be coming. It was fascinating to know that they’re working on things like that.

      But unfortunately Carmelo Anthony ranked high on it, so we have to take it apart, and we can’t, so it sucks, I guess.

    52. Hubert

      But for all we know, gravity score could be calculated with a dice roll. I’m not saying that it is a bullshit statistic, but I am saying we have no idea because the methodology is hidden. It’s best not to draw conclusions from unverified data.

      No, we know it’s calculated using SportsVU data. We know SportsVU tracks the position of players on the court. We know that something that tracks the position of players on the court has a fair chance of being accurate about how close a player is guarded. If I’m his editor, I’m saying throwing that in there adds value and is worth keeping in.

      Again, it was ONE piece of information he included among many things supporting his premise, and it was arguably the one he gave the least amount of weight to.

      If the article was “Kyle Korver is an offense to himself because he ranks fourth on gravity score”, I’d see your issue. But it wasn’t.

    53. Kevin Udwary

      If the article was “Kyle Korver is an offense to himself because he ranks fourth on gravity score”, I’d see your issue. But it wasn’t.

      I’m not criticizing the article, I’m criticizing people who are trying to draw conclusions from this unknown stat. I think it was a good article, on the whole, but the inclusion of this stat was completely unnecessary.

    54. Hubert

      I’m criticizing people who are trying to draw conclusions from this unknown stat.

      Fair enough. I think johnno was just trying to stir shit up but making this into a Melo v Chandler thing. It has no place in that discussion.

      But you said

      but the inclusion of this stat was completely unnecessary

      and BC said

      Especially since Lowe had already done such a great job describing Korver’s importance, why even throw in the new stat?

      Wasn’t it pretty awesome to know that shit like this is being developed? Isn’t that enough reason to include it? Lowe is a different kind of writer than most. He’s a pluralist, whereas most of the guys we like to discuss here are dogmatists.

    55. Kevin Udwary

      Hey, SportsVU is a potentially awesome tool for analysis of the game. I’m just pissed the data isn’t public, and anything we’re going to see from it is going to be black box statistics, like gravity score.

    56. Unreason

      incorporating specific location data from every recorded NBA possession and factoring in baseline player tendencies

      This gives hints that allow for an educated guess at the general approach.

      I suspect the gist was to:
      1. Develop a predictive model of each defender’s man-to-man distance during off-ball defense (i.e. each defender’s baseline tendency) using everything except who was being guarded as predictors
      2. Calculate the mean difference between the model-predicted and actual observed man-to-man distance for each guarded player across all defenders who guarded him

      To do that, you’d need a rule for classifying who is covering who and for when off-ball coverage begins and ends. Then for each player, calculate the mean distance from their man across all guarding episodes. Then fit that distance measure in a hierarchical regression using any and all potentially relevant measures at their disposal as predictors.

      Since it’s not a theory-building analysis, they’d be able to throw in everything plus the kitchen sink as predictors: distance from basket, whether a screen was set, minutes played so far (fatigue), current score, shot clock, etc., plus interactions. No need for valid betas, just an optimized prediction of each defender’s mean man-to-man distance, given all those conditions, based on all players they guarded.

      With that they could calculate the size of the difference between defender X’s predicted mean distance from a hypothetical average player and his actual mean distance when guarding player X. That difference would indicate how much closer/further defender Y stayed to player X than he would have stayed to a hypothetical average player after taking all other measurable factors into account. Player X’s gravity score would just be the average difference from predicted mean distance across all defenders who covered him.

      I’d be surprised if it isn’t some variation on that general approach.

    57. EB

      My problem and my point before was that “really guarding” is a completely meaningless term to me. What does “really guarding” actually look like on a court? The problem is the same as when we say best scorer. What does one quantify best scorer? As I pointed out before there are several ways, similarly there are likely several ways to quantify “really guarding”, therefore we don’t actually know what the statistic means. Is it how often a defender is within arm’s reach of him, how often the player has his eyes on him vs. the ball, or how often a player is within closeout distance from a player in an area which the player is a dangerous scorer? If you can tell me which one of those we are talking about, then I will concede that we know what gravity and distraction are.

    58. DRed

      The point is Kyle Korver attracts a lot of attention. The inclusion of gravity score wasn’t necessary to make his point

      But does it even show that? Like I said, maybe it shows Kyle Korver is bad at getting open, and defenders are closer to him than to other shooters are because the other guys are better at finding open spots than Kyle. Maybe some of the other guys in the top 10 are bad offensive players (I’d doubt it, but since we’ve only seen the top 4, who knows?) How does he stack up against other guys known for their long range shooting? I think we all can imagine that teams try to closely guard guys who are deadly marksmen. I thought that before I’d even heard of gravity score.

    59. Kevin Udwary

      Since it’s not a theory-building analysis, they’d be able to throw in everything plus the kitchen sink as predictors: distance from basket, whether a screen was set, minutes played so far (fatigue), current score, shot clock, etc., plus interactions. No need for valid betas, just an optimized prediction of each defender’s mean man-to-man distance, given all those conditions, based on all players they guarded.

      But with the raw data that SportVU gives, all that is not at all trivial. Check out this paper on automatically recognizing an on-ball screen. I mean you are dealing with just raw positional data of 10 guys. It would be awesome data to have, though…

    60. johnno

      ” I think johnno was just trying to stir shit up but making this into a Melo v Chandler thing.”
      Nah. My comment @22 was meant to stir shit up. The Melo v. Chandler comment was a lame attempt at humor.

    61. johnno

      By the way, I read two rumors recently — the Knicks are considering Michael Beasley and they are trying to somehow swing a deal that nets them Greg Monroe. Not sure how I feel about either.

    62. Unreason

      I’m with Hubert on this. It’s cool to get a glimpse however fleeting at the more informative stats that are on the horizon. The context for citing the stat determine it’s appropriateness. It’s in a well-written mass audience article. He gives an appropriately concise description of what it tries to capture. He doesn’t make grand claims about it’s comprehensiveness or say it should determine player value. He uses it as a tantalizing bit of ancillary evidence that backs up expert opinion and visual illustrations. I came away thinking I’d love to find out more, not feeling like he should have omitted it if he wasn’t going to lay out the details.

    63. DRed

      How is something like gravity score “more informative” when we literally don’t know what it is or how it effects the game?

    64. DRed

      Beasley’s TS% the last five seasons: .505, .510, .502, .462, .559. If he’s somehow changed his game, or if there’s something in the way Miami used him that Phil thinks we can replicate, and he’s super cheap, than he’s not a bad guy to have on the bench.

      Monroe is big and young and not terrible, so he’s also not a bad guy to have on the team, but it would all depend on what we’re giving up and what we’re paying him.

    65. Kevin Udwary

      Monroe is a FA, right? And we’re over the cap so we can’t take anyone back in a sign-and-trade? I’m pretty sure the Knicks are the default team that agents throw out as “showing interest” in their clients in order to try and stimulate the market.

    66. EB

      Monroe was killing it his first two years. I’d love to have him, though I’m curious why he fell off. I’d guess it’s because Drummond takes up the paint.

    67. Unreason

      But with the raw data that SportVU gives, all that is not at all trivial. Check out this paper on automatically recognizing an on-ball screen. I mean you are dealing with just raw positional data of 10 guys.

      I agree that measure development has to be done thoughtfully. I assume it is. I just meant that it probably wasn’t too hard to guess the gist of what they did.

      The importance of measurement imperfections depends on the precision you need though, don’t you think? Getting things nearly perfect is seldom trivial but in this case my hunch is you don’t need anything close to perfect. You’re not even dealing with sample statistics. For lots of the things you’d want to measure the raw data the measure is built from captures the entire universe of the phenomena you want to draw inferences about in the entire population. That cures a whole lot of measurement ills right there.

      Even if you had to rely on samples of data instead of the whole population, I’d guess that a PPV of 80% for classifying on-ball screens is highly serviceable for this purpose. Don’t you think? I could be wrong, of course, but unless your algorithm produces large systematic errors in measuring a very important and distinct predictor, I’d guess that in general a bit of work will get you measures that are good-enough to capture the main drivers of man-to-man distance.

      It would be awesome data to have, though…

      It sure would. Wasn’t there some plan to make at least some of it publicly available in the hopes that enthusiasts and academics would drive new methods?

    68. Unreason

      How is something like gravity score “more informative” when we literally don’t know what it is or how it effects the game?

      I agree it’s only loosely informative, but then again we don’t know the details of what goes into expert opinions either, right? Or believe that box score-based stats are the end-all be-all; Or that YouTube clips reveal the unvarnished truth. We agree that basketball is not an area where there’s much precisely detailed evidence-based knowledge. So the brief description of what it’s supposed to measure and the related conjectures about how it might be valuable are just the kind of loosely informative things that make the hoops talk world go round, IMO. Given the context, I don’t feel the need to hold it to a higher standard.

    69. johnno

      “Monroe is a FA, right? And we’re over the cap so we can’t take anyone back in a sign-and-trade?”
      I’m not sure how it would work but the Knicks would have to trade away enough salary to get under the “apron” so that they could take back someone in a sign and trade. My biggest concerns with Monroe is what the Knicks would have to give up and that he’s bound to be overpaid (as almost all tall non-complete-stiffs are).

    Comments are closed.