Announcement

Collapse
No announcement yet.

Evaluating players defensively

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Evaluating players defensively

    was tweeting back and forward with Alex Rucker a bit today. Thought I would share the exchange.

    Alex Rucker ‏@Alex_Rucker
    There is almost nothing about a player's on-ball or help defense that you can glean from the box score.

    Timothy Hedden ‏@trhedden
    @Alex_Rucker doesn't blocked shots point towards good help defense?

    Alex Rucker ‏@Alex_Rucker
    @trhedden Yes, it absolutely suggests that. But there are plenty of players who get blocks and are very poor help defenders, esp in the post

    Naveed ‏@NaveedDawg
    @Alex_Rucker Are there better publicly available stats for this? http://82games.com has opponent PER, JJ was ok, anything else?

    Alex Rucker ‏@Alex_Rucker
    @NaveedDawg No...public stats are based on data which does not support accurate defensive analysis.


    Timothy Hedden ‏@trhedden
    @Alex_Rucker what do you look at then in order to evaluate great defenders? Points against, +/-


    Alex Rucker ‏@Alex_Rucker
    @trhedden None of the above. We have orders of magnitude more data than the public sphere.
    For still frame photograph of me reading the DeRozan thread please refer to my avatar

  • #2
    Tom Liston ‏@Liston
    #Raptors James Johnson had the 7th highest turnover rate out of 69 small forwards (min 10mpg, 10g). Hedo 2nd

    Timothy Hedden ‏@trhedden
    @Alex_Rucker casual fans think we gave up a good defender for a 2nd rounder. Being connected 2 the team I'd understand if you can't comment

    in regards to why JJ was shipped out

    Tim Chisholm ‏@timpchisholm
    A major schism between player/team vision of JJ. @trhedden: @timpchisholm did they give a reason? I know there was the flare up in practice.
    For still frame photograph of me reading the DeRozan thread please refer to my avatar

    Comment


    • #3
      This is in line with what Rucker said in the Q&A with Tom Liston (in statophile 35).

      I’d also offer that the media’s perspective of player values is interesting, but it doesn’t always align with pro team’s perspectives – and one of those two enjoys a significant information advantage by virtue of the resources invested into the draft process.
      In my opinion too many people draw too many conclusions from so-called 'advanced stats' like PER, WP, Defensive Rating, etc. All of these stats are making pretty extreme statements about the value of players based on extremely reductionistic versions of the reality on a basketballcourt.

      Comment


      • #4
        And how confident are we all that BC has a guy or a team working on better defensive stats? Fans of Houston, Boston, Dallas, etc would be very confident without hesitation... Toronto fans... not so much.

        May this thread now be merged.

        Comment


        • #5
          Stay away from lightening storms. That tinfoil is dangerous.

          Comment


          • #6
            Lightening storms are more dangerous than darkening storms? Me need a new tinfoil hat.

            By the way Matt, whose idea was it to darken your Username from orange/yellow to black?

            Comment


            • #7
              Hugmenot wrote: View Post
              Lightening storms are more dangerous than darkening storms? Me need a new tinfoil hat.

              By the way Matt, whose idea was it to darken your Username from orange/yellow to black?
              He's not a moderator anymore... Unless that was sarcasm.
              Twitter: @ReubenJRD • NBA, Raptors writer for Daily Hive Vancouver, Toronto.

              Comment


              • #8
                Hugmenot wrote: View Post
                Lightening storms are more dangerous than darkening storms? Me need a new tinfoil hat.

                By the way Matt, whose idea was it to darken your Username from orange/yellow to black?
                Mine.

                Comment


                • #9
                  we get off topic pretty quickly around here
                  For still frame photograph of me reading the DeRozan thread please refer to my avatar

                  Comment


                  • #10
                    Matt52 wrote: View Post
                    Mine.
                    Thanks for the response Matt.

                    And I will raise my glass tonight in honor of your past services to our community! Sincere thanks!

                    Comment


                    • #11
                      Soft Euro wrote: View Post

                      In my opinion too many people draw too many conclusions from so-called 'advanced stats' like PER, WP, Defensive Rating, etc. All of these stats are making pretty extreme statements about the value of players based on extremely reductionistic versions of the reality on a basketballcourt.
                      All of the advanced metrics can be useful in one way or another as long as you understand their limitations and deficiencies. It's one of the things that bothers me so much about Berri and his acolytes. When you are using regressions to analyze very, very, very complicated matters, a little humility and caution seems prudent. Instead, if you question any of their results (no matter how patently ridiculous) they just tell you that regressions are very complicated things that not everyone understands. But for some people all they have are hammers so everything looks like a nail.

                      If you look at MLB (far ahead in the analytics dept) the best teams use a combination of stats and scouts. There is a reason for that....

                      Comment


                      • #12
                        slaw wrote: View Post
                        All of the advanced metrics can be useful in one way or another as long as you understand their limitations and deficiencies. It's one of the things that bothers me so much about Berri and his acolytes. When you are using regressions to analyze very, very, very complicated matters, a little humility and caution seems prudent. Instead, if you question any of their results (no matter how patently ridiculous) they just tell you that regressions are very complicated things that not everyone understands. But for some people all they have are hammers so everything looks like a nail.
                        Absolutely; that's one of the things why I have a little bit of hate for some of these (so-called) advanced metrics. They have titles on articles like "George Karl is an idiot" (really). They have no clue that they are missing a whole lot of information in their truthtelling stats. These guys are, to stay in language they understand, complete morons.

                        In their list of Wins Produced of 2000/2001, the year Philidelpia made it to the finals, they had Iverson as the 388th on their list, about the worst player on the Sixers.

                        Before the 2010 season, they predicted 57 wins for the Golden State Warriors and afterwards some major rationalizations were needed to justify the 36 wins Golden State actually had.

                        Last season Fields was the 8th best player in the NBA according to WP (which btw, is 16 places higher than Kevin Durant; man are we lucky we got Fields instead of Durant this summer).

                        They used to love Josh Childress when he was with Atlanta even though he is a small forward who gets most of his shots at the rim (hence the high efficiency) but can't (and luckily doesn't very often) shoot for shit from 9 feet (he's in the running for worst looking jump shot in the nba, though he probably can't beat MKG). Yet, according to their allknowing formula he's playing on an allstar level. Seriously, by looking at his game they should realize their stat needs drastic change. They simply don't get it and treat their stat like only a fanatic religious sect could.

                        If you look at MLB (far ahead in the analytics dept) the best teams use a combination of stats and scouts. There is a reason for that....
                        I don't know anything about MLB, but what I always read, is that in baseball it's much easier to use these kind of advanced stats and they are much closer to reality.

                        Comment


                        • #13
                          slaw wrote: View Post
                          All of the advanced metrics can be useful in one way or another as long as you understand their limitations and deficiencies. It's one of the things that bothers me so much about Berri and his acolytes. When you are using regressions to analyze very, very, very complicated matters, a little humility and caution seems prudent. Instead, if you question any of their results (no matter how patently ridiculous) they just tell you that regressions are very complicated things that not everyone understands. But for some people all they have are hammers so everything looks like a nail.

                          If you look at MLB (far ahead in the analytics dept) the best teams use a combination of stats and scouts. There is a reason for that....
                          I think Wins Produced is a very good analytical tool. Not perfect, but very good. But the arrogance of the Wages of Wins Journal seems almost limitless. They simply refuse to get into a real debate into the nuts and bolts of there model.


                          No...public stats are based on data which does not support accurate defensive analysis
                          funny part about this, team after team is hiring people from the public sphere to work for their team using those same public stats. Rucker has a very big incentive to discrediting 'public' stats. How important is his job if 'public stats' are more accurate than he leads on?

                          Ofcourse I have to ask the question. If the method he uses, which he didn't release, is so good - why have the Raptors been ranked 30th, 30th and 14th in defense since he got here? Their record over that time is .488,
                          .238, .348.

                          Either Rucker isn't being listened to or his data is yet to show real fruition (or both)

                          Having orders of magnitude and 'more' data does NOT mean better or more accurate data.

                          Comment


                          • #14
                            The stat heads even argue about which metric is the most accurate or relevant for evaluation. To state the obvious, each metric has its own limitations. Each person also has their preference of which metric they find most useful. As the field evolves, the equations derived to arrive at certain figures become more and more complex. Statistics present the facts as outlined by the parameters drawn. Interpretation of these statistics, is another story.

                            I became interested with the whole advanced statistics and I began studying it about 6 months ago. I found it really interesting because it goes beyond the traditional box score & also because I love my Math lol. To tell you the truth, regressions are not that complicated. In fact, you can do it on Excel or a program called SPSS. You can do your own evaluations, accurate or inaccurate. It's fun to do, although I'm pretty sure it's a big waste of time -- on my part.
                            “I don’t create controversies. They’re there long before I open my mouth. I just bring them to your attention.”

                            -- Charles Barkley

                            Comment


                            • #15
                              Craiger wrote: View Post
                              Ofcourse I have to ask the question. If the method he uses, which he didn't release, is so good - why have the Raptors been ranked 30th, 30th and 14th in defense since he got here? Their record over that time is .488,
                              .238, .348.

                              Either Rucker isn't being listened to or his data is yet to show real fruition (or both)

                              Having orders of magnitude and 'more' data does NOT mean better or more accurate data.
                              Maybe this shines some light on what their role is withing the organisation. It's from a Magic blog article (from 2010) I read some time ago about Hedo Turkoglu. The role will be somewhat expanded now, because they joined that season after trades had been made and the draft had taken place. Of course they can't stop Bosh from leaving, so I don't see much of a relation between their influence or performance in the decline from 2009/10 to 2010/11.

                              Part I of the interview with Keith Boyarsky and Alex Rucker
                              Part II

                              I know that you and Keith joined the Raptors in the off-season. Did you have a say in the signing of Turkoglu or did you arrive to the team after that?

                              No, the signing took place before we joined the organization. We had no input on the signing. It was interesting, though, I think it was actually … yeah, the first question that general manager Bryan Colangelo asked us was basically, ‘I made this big decision for the franchise: a.) What do you think? Basically, do some analysis in terms of what impact that will have’ and then b.) was the Jay part, which is ‘how is he best used in our offense?’ or ‘how do you think he’ll fit and what are the roles where he appears to have the most value.’ We did a lot of work on Hedo in the off-season but it was all, kind of, after the fact. That was purely a timing thing.

                              I’m sure you can’t state, specifically, what you’ve done and the numbers you’ve crunched but have you seen some of the suggestions that you’ve given to Jay come to fruition in games? How does that work itself out, in terms of how Jay uses your input?

                              Yeah, there’s a lot of what I call ‘fog of war’ when it comes to that because we … it’s interesting, from reading [the APBRmetrics forum] and some of the other things at the Sloan Conference, my sense is that most of the quantitative analysts, most of their work is with, kind of, the personnel side and with the management side. I think we’re probably most like 70/30 with coach/management stuff so most of the stuff we do is for Jay and the coaches so in terms of that question, the answer is probably yes.

                              I’ll ask questions. We’ll look at things provided for them, give them answers, kind of have an ongoing dialogue about certain things. Yeah, I mean, sometimes I’ll talk to Jay and he’ll ask, ‘hey, can you suggest doing this? Does that work well?’ I think most of it is more just additional data points, in the sense of when coaches or managers make decisions, they consider a lot of things. I think that our job is to make sure that when they consider the quantitative production stuff, they consider it more correct than what they’re going to get at [Basketball-Reference] or whatever. Not that the stuff is wrong. I guess ESPN is a better example. Not that box score stats are wrong, they don’t reveal everything.

                              There’s no question, with the Hedo question, you asked about his role. I mean, yeah, we … possessions that he plays role as creator, not necessarily as ball-handler in terms of bringing the ball up the court but once you’ve gotten into the quarter-court, if he is … I don’t know how you want to characterize it in normal lingo. When he’s the guy running the offense, when it gets to him, he’s the one having to do the shot creation, those are our most valuable possessions. I mean, Jose is a great point guard, Jack is a very good point guard and they have definitely some strengths but Hedo, in the half-court, is a unique player in terms of what he does against defenses. We benefit a lot from that so if you watch Raptors games, I think you’ll see that a fairly significant percentage of our offensive plays when we’re running the half-court will go right through Hedo when he’s on the floor and that’s totally intentional. Now, did Jay do that because of stuff we told him or did he do that because he appreciates Hedo’s talents? I don’t know what the answer is. It’s probably a culmination of the two, I guess.

                              Comment

                              Working...
                              X