Jump to content


Toggle shoutbox Shoutbox Open the Shoutbox in a popup

@  08huangj : (26 October 2014 - 01:11 PM) For those of you that care: Happy birthday Parsons!!!
@  Cooper : (25 October 2014 - 01:35 PM) Good thing they wasted their first round pick on a guy they sent directly to the dleague.
@  feelingsuper... : (25 October 2014 - 11:14 AM) The Thunder have lost Morrow for about a month and a half. They will be digging themselves out of a hole most of the season.
@  2016Champions : (25 October 2014 - 06:04 AM) I'm a little hyped right now, holding myself back from going full retard.
@  2016Champions : (25 October 2014 - 06:03 AM) Call me crazy but I think if Rockets get the Thunder in the 1st round, then the Spurs in the 2nd, that would be more ideal than any road to the finals where Terrence Jones will be guarding the likes of Blake Griffin, Aldridge, Lee, Dirk, Zbo ect.
@  2016Champions : (25 October 2014 - 05:59 AM) Jeremy who? Lol
@  thejohnnygold : (25 October 2014 - 02:25 AM) That was definitely an awesome way to finish the game
@  Willk : (25 October 2014 - 02:22 AM) im on the canaan bandwagon
@  Losthief : (25 October 2014 - 12:43 AM) Keep Gettin Them CHecks All-Star!
@  Buckko : (24 October 2014 - 03:57 AM) That's an understatement Cooper.
@  rockets best... : (24 October 2014 - 01:16 AM) Nash should have quit 4 years ago
@  Cooper : (24 October 2014 - 01:15 AM) suns really sold high on him.
@  Willk : (24 October 2014 - 12:03 AM) sad to see nash's career end like this
@  cointurtlemoose : (23 October 2014 - 07:16 PM) Philly "doing it right" is a catch-22 when doing it right is completely sacrificing competitiveness for almost entire seasons. I see what you're saying, and it's not Philly's fault.
@  2016Champions : (23 October 2014 - 05:59 PM) I need more to go on to write something up. Just brainstorming right now. Monroe is more of a defensive plus at the 4 than the 5; I think the reason he played the 5 early on his because he couldn't shoot--no longer the case.
@  08huangj : (23 October 2014 - 12:28 PM) He just doesn't seem like a four to me, he started his career as a center. I also don't think that he'd be happy to be a backup and what does a very reasonable price mean. Maybe you can create a thread.
@  2016Champions : (23 October 2014 - 10:14 AM) Monroe would not only improve our starting 4 spot but also serve as a backup center.
@  2016Champions : (23 October 2014 - 10:12 AM) Should we go after Greg Monroe? He seems to have improved his midrange shot, and considering the likelihood he walks, we might be able to acquire him for a very reasonable price.
@  Buckko : (23 October 2014 - 04:31 AM) I have no problem with philly tanking. If you're gonna do something, do it right. They should be an extremely entertaining team in a few years.
@  Willk : (23 October 2014 - 01:23 AM) 2016 - I agree. I would rather see Canaan play than Ish

Photo

The Rockets starting line-up plays better with D-Mo


  • Please log in to reply
12 replies to this topic

#1 2016Champions

2016Champions

    Veteran

  • Members
  • PipPipPipPipPipPip
  • 3,172 posts
  • LocationVirginia, USA.

Posted 22 October 2014 - 09:10 AM

One look at the stats and the highlights and it's clear that Terrence Jones has more potential than Donatas Motiejunas. Jones had a PER of 18.9 while D-Mo's was a mere 11.3. Coach McHale seems convinced too seeing as the Beverley-Harden-Parsons-Jones-Howard line-up saw 619.2 minutes last season, while in contrast Motiejunas was given only 67.6 minutes with that starting line-up in Jones's steed. The 5-man unit with D-Mo actually performed better but we can't draw any conclusions from that given the small sample size, however, at the very least I think it validates the question: are the Rockets a better team with D-Mo in the starting line-up? Jones is probably the better player, but the better player and the better fit is not necessarily mutually exclusive. 


The Rockets starting line-up performed better with Donatas Motiejunas in Terrence Jones's steed

The starting line-up with Casspi at power-forward also performed better, but given Casspi's lack of size, obviously that was something the Rockets could only do if the match-ups allowed it. Is that the case with D-Mo as well? He's not the fastest guy in the world laterally, maybe those 67.6 minutes came against opposing PFs who aren't great at putting the ball on the floor.

There is so much that goes on defensively, so many different aspects of defense to consider, that it's hard for even experts to put a value on a player's defensive impact. I've seen many fans in NBA discussion groups attempt to measure defensive impact with blocks and steals, but most of us are aware by now that method is a flawed science at best--Dennis Rodman is a testament to that. After discussions with some very smart people familiar with NBA analytics (via twitter and online forums), I have come to the conclusion that RAPM is the best single tool for measuring defensive impact, even ESPN have developed their own version (I'm not yet sure what the difference is) called RPM. According to RPM, Donatas is a plus defender. This metric does have a margin of error, and those like Donatas who didn't play a whole lot of minutes would have a much larger margin of error than a starter would. Obviously, we should take the fact Donatas has a positive dRPM with a grain of salt, but I don't think we should dismiss it completely. Donatas's lateral movement might look poor in contrast to the likes of Paul Millsap and Draymond Green, but it's not exactly Al Jefferson-bad. At the end of the day he's a legit 7 footer who tries, and for whatever it's worth, defense has been his focus of improvement this summer. Besides, Terrence Jones's dRPM is atrocious, ranking 80th among power forwards last season which doesn't surprise me one bit. Seemingly every power forward in the league put up career highs against Terrence Jones. At the very least, the defense won't get any worse if Donatas was to start over Terrence, and I have reason to believe the offense could get astronomically better. I'm going to explain why I think the Rockets offense could be better with Donatas, just bear with me for a minute.

 

According to SynergySports, the two least efficient plays the Rockets ran last season were the Dwight Howard post-up and the James Harden isolation, and for reasons unknown these were also the two most frequently ran plays. In contrast, the pick-and-roll was the most efficient play, and it was used on less than 13% of Rockets possessions. 

 


The Rockets ranked 1st in pick-and-roll efficiency but used it on less than 13% of their possessions

 

Interestingly, the Mavericks used the pick-and-roll more than any other team in the league, and they arguably gave the Spurs the most trouble in the playoffs this year pushing them to 7 games. However, the fact the Mavericks have the best floor spacing power forward in the league might have something to do with how frequently they were able to run the pick-and-roll. Afterall, many defenses are putting more emphasis on packing the paint these last several years--something Zach Lowe believes has come as a result of Coach Thibodeau's influence. 


An example of how teams pack the paint against the pick-and-roll

 

I believe the Rockets will be able to operate the pick-and-roll with a higher level of frequency with a starting power-forward who stretches the floor, and I believe Donatas is a much better shooter than Jones is. Don't let the stats fool you, despite the poor percentage Donatas shot from outside last season, defenses still respect him out there to a certain degree--a degree I am optimistic can rise given Donatas's good looking stroke. In contrast, Terrence gets blatantly left wide open out there. 

 

That about sums up my thoughts on why the Rockets starting line-up might perform better with D-Mo. I could be wrong, but considering all the information which points towards the possibility in question, I only hope McHale gives D-Mo a chance. Whether you like D-Mo or not, there's no denying that a 7 footer with his skill-set is extremely rare, and I would hate to see D-Mo leave the Rockets and blow-up somewhere else (especially the Mavs, that would kill me). 

 


Edited by 2016Champions, 22 October 2014 - 09:17 AM.

  • 0

"We're not satisfied with where we're at, but we're optimistic on where things will go from here" - Daryl Morey


#2 marbony81110

marbony81110

    Newbie

  • Members
  • Pip
  • 33 posts
  • LocationGermany

Posted 22 October 2014 - 11:14 AM

Based on the stats and watching the games I honestly don't think it matters if it is D-Mo or Terrence Jones in the starting lineup. D-Mo offers more size which is a benefit on D, but does have the explosiveness and athleticism Jones shows on the offensive end. The starting lineup minutes and results comparison is too limited to accurately judge D-Mo's performance as you already stated, but the effectiveness of that unit surprised me a bit.

 

Both players are poor 3-pt shooters. I think Jones gets more open looks because players play off him since he is a good dribble-penetrating PF. D-Mo does not have that ability in his game so defenders stick closer to him. If I was guarding Jones, I would much rather give him the open 3 than be crossed up, especially as a PF that is typically slower than Jones.

 

Both players have their merits and faults. Bottom line is that if one of them does not step up their game and consistency, an upgrade is going to be needed. That goes for the backup PG as well. If Jones doesn't perform well in the starting lineup, I'm sure Mchale will make the switch quick. At least I think he will. This is McHale I'm talking about here...sigh...


  • 0

#3 bboley24

bboley24

    Junior Member

  • Members
  • PipPipPip
  • 356 posts

    Posted 22 October 2014 - 12:49 PM

    DMO is one of the strangest ballers that ive seen.  You look at his game and say, WOW.  It rarely translates though.  It seems that he gets really nervous at times when he goes to get a shot up down low.


    • 0

    #4 Alituro

    Alituro

      Junior Member

    • Members
    • PipPipPip
    • 451 posts

      Posted 22 October 2014 - 01:00 PM

      I think the 3's and layups philosophy is what has been holding D-mo back the most, and the same applies for Papa this preseason also, from what I've observed. These guys have spent their entire basketball career building muscle memory to shoot the Euro 3. If they weren't expected by directive to take that extra step back and shoot from the spot they've been most comfortable from their entire lives, I think they can be more effective offensively. To build the muscle memory to shoot the NBA 3 at a high rate could take a long time. Jones, on the flipside, even though the college 3 is shorter, like many others, dreamed of an NBA career and has spent many hours practicing his shot from that outer arch on the court. This is where I think the analytics movement falls a little short. While, in a vacuum, the 3 is the most efficient longer range shot, is it necessarily for certain types of players? How much would their eFG% improve if they just took the shot they're comfortable with rather than forcing one they're not? I'm not sure what factors into dRAPM but if the performance of the unit affects an individuals score, then you have to realize that a lot of Jones statistical incompetence could have everything to do with the sieve that was our wing defense last season. Always having to leave his man to try and stop dribble penetration, would result in his man being open and scoring a bunch, but was it really Jones' fault? Is D-mo's elevated because he played with a stronger defensive unit?

       

      I don't really have a preference on which one starts, they both have plusses and minuses that cancel each other out., But like Marbony81110 said, both will have to pick it up with consistency or expect to be replaced.


      • 0

      #5 2016Champions

      2016Champions

        Veteran

      • Members
      • PipPipPipPipPipPip
      • 3,172 posts
      • LocationVirginia, USA.

      Posted 22 October 2014 - 04:59 PM

      RPM attempts to iron out that flaw in +/-, and for the most part I think it does a pretty good job. Not a perfect job, hence the fact it has a margin of error, but in large sample sizes that margin of error becomes minimal.

      http://m.espn.go.com...818&src=desktop
      • 0

      "We're not satisfied with where we're at, but we're optimistic on where things will go from here" - Daryl Morey


      #6 thejohnnygold

      thejohnnygold

        Veteran

      • Moderators
      • 3,544 posts
      • LocationAustin, TX

      Posted 22 October 2014 - 06:25 PM

      RPM attempts to iron out that flaw in +/-, and for the most part I think it does a pretty good job. Not a perfect job, hence the fact it has a margin of error, but in large sample sizes that margin of error becomes minimal.

      http://m.espn.go.com...818&src=desktop

       

      I like how that "explanation" of RPM goes.  They just "yadda-yadda-yadda'd" right past the part where they explain what it actually is.

       

      elaine%20yada%20yada.gif

       

      They completely destroy any remnant of actual data by running it through so many progressions and adjustments that the progressions and adjustments need progressions and adjustments.  Further, they reach their conclusion not by accounting for the team impact, but by ignoring it completely.

       

      Vinegar, on it's own is disgusting.  Vinegar in food is amazing.  Garlic, on it's own is harsh and can be bitter.  Garlic in food is amazing.  Could you imagine a team of statisticians trying to calculate the impact of vinegar or garlic in food this way?  That's not how this works.  It is mathematical wizardry of the highest order (which is cool), but I caution against putting too much stock in it because I believe it is fool's gold.

       

      For instance, just to be a hypothetical jerk, if we took the top 5 DRPM PG's and put them on the floor together one could logically argue that this team would be excellent on defense.  Obviously, that would not work--because who is out there matters.  The combination matters.  

       

      Sauerkraut is good on bratwurst and Reuben sandwiches.  That doesn't mean it will be good with everything.  Don't take my word for it though.... ;)

       

      RPM reminds me of the infamous transmogrifier...

       

       

      I think my thoughts on the matter can best be summed up by Mr. Hobbes.

       

      fffduplicator.jpg

       

      I'm not saying "don't use RPM".  I'm just saying people should get educated about what it is and isn't before accepting anonymous testimonials from "very smart people".  Someone told me long ago that "RPM isn't meant to find answers, but to lead people to the right questions".  So far, my concern that people will not do that and will just use the numbers to help support their opinions has been very, very true.

       

      Good luck to everyone with the xRAPM, RAPM, dRAPM, RPM, and the rest.  Caveat emptor.  By the way, here is a link that explains it all in layman's terms: LINK.


      • 0

      #7 Sir Thursday

      Sir Thursday

        Senior Member

      • Members
      • PipPipPipPipPip
      • 1,026 posts
      • LocationUnited Kingdom

      Posted 22 October 2014 - 07:32 PM

      Personally I use RPM as a reference because it is easy to access and is pretty close to RAPM, which I know the details of. I would prefer to be using RAPM, but I don't know of a good source for it - any ideas?

       

      Also JG, I think you're being a little facetious when you say things like this:


      For instance, just to be a hypothetical jerk, if we took the top 5 DRPM PG's and put them on the floor together one could logically argue that this team would be excellent on defense.  Obviously, that would not work--because who is out there matters.  The combination matters. 

       

      One cannot 'logically argue' any such thing. This class of statistic has not, nor has ever purported to be, the be all and end all of analysis of defensive players. It does not measure the best defenders, but rather the players who have the biggest impact within their schemes. Just because it is possible to use a statistic for something other than its intended purpose does not make the statistic itself invalid.

       

      I guess what I'm saying is, I agree with being critical of the provenance of statistics and how they are used, but I don't think we need to dismiss statistical tools out of hand because occasionally someone uses them wrong. Are you saying you don't think 2016 was using R*PM correctly here?

       

      The one thing I would say is that generally I'd like to see a more serious treatment of error bars in NBA statistics and their use. The handwavy 'small sample size!' is thrown around a lot and then dismissed, often without too much thought. Understanding how accurate (and therefore useful) a measurement actually is should be a key part of its use - 2016 mentions in his post that the lack of playing time might make D-Mo's RPM unreliable...well how unreliable? I want to know so that I understand whether it's actually worth taking into account.

       

      I've got a good idea for an article on this, will see if I can find the time to write it up at some point.

       

      ST


      • 1

      #8 thejohnnygold

      thejohnnygold

        Veteran

      • Moderators
      • 3,544 posts
      • LocationAustin, TX

      Posted 22 October 2014 - 10:26 PM

      Personally I use RPM as a reference because it is easy to access and is pretty close to RAPM, which I know the details of. I would prefer to be using RAPM, but I don't know of a good source for it - any ideas?

       

      Also JG, I think you're being a little facetious when you say things like this:

       

      One cannot 'logically argue' any such thing. This class of statistic has not, nor has ever purported to be, the be all and end all of analysis of defensive players. It does not measure the best defenders, but rather the players who have the biggest impact within their schemes. Just because it is possible to use a statistic for something other than its intended purpose does not make the statistic itself invalid.

       

      I guess what I'm saying is, I agree with being critical of the provenance of statistics and how they are used, but I don't think we need to dismiss statistical tools out of hand because occasionally someone uses them wrong. Are you saying you don't think 2016 was using R*PM correctly here?

       

      The one thing I would say is that generally I'd like to see a more serious treatment of error bars in NBA statistics and their use. The handwavy 'small sample size!' is thrown around a lot and then dismissed, often without too much thought. Understanding how accurate (and therefore useful) a measurement actually is should be a key part of its use - 2016 mentions in his post that the lack of playing time might make D-Mo's RPM unreliable...well how unreliable? I want to know so that I understand whether it's actually worth taking into account.

       

      I've got a good idea for an article on this, will see if I can find the time to write it up at some point.

       

      ST

       

      I was being over-the-top facetious when I put forth the idea of the 5 top RAPM PG's--I thought I had made that clear, but perhaps not.  I was simply pointing out the complete lack of focus it has.  These numbers get published for public consumption without much, or any, direction at all.

       

      As far as I know, ESPN's RPM is closely based off of xRAPM which can be viewed here at it's creator's site.  I did not find a sortable table, but there is lots to see there.  Here is an article that tries to explain all the different advanced metrics including xRAPM.  LINK

       

      I agree that the statistic is not without its usefulness; however, I have found it's usefulness largely limited to validating the "eye test".  I have also lobbied very hard against it in the past for it's high number of anomalous scores and rankings, its favoritism towards big men, the massive amount of number filtering, and other things that get swept under the rug.

       

      Like I said above, I'm not prohibiting anyone from using/referencing RPM stats.  I am encouraging people to get educated on them.

       

      My response regarding 2016 is partially due to how it was presented.  Taking DRPM and cloaking it in nonchalance, then buttressing it with a combination of bandwagon/celeb endorsement--basically giving us the "all the cool kids are doing it" routine.  He freely admitted he did not fully understand the very stat he was citing; yet, amongst all the links he posted none had as much information as the sites I found in 15 seconds of google-fu.

       

      Secondly, yes I do think it is being mis-used above.  It is his leading argument for making D-Mo the starter at PF.  Further, he fails to ever mention any exact data regarding D-Mo's score referring to it only as "a plus" while getting very specific with Terrence Jones' "atrocious" 80th rank.  I know a snow job when I see one.  You can't just dismiss everything that sweeps the legs from your argument with a simple "handwave"; yet, that paragraph is littered with such waves of the hand--"take it with a grain of salt" and "for what it's worth".  A grain of salt?  More like the entire salt mine. 

       

      D-Mo had 20 DNP's last season.  He played in 62 games and started 3 averaging 15 mpg.  A look at his splits shows some ridiculous numbers especially regarding his +/- (which I am guessing is adjusted to 100 possessions because how can you get a +50.8 in 16 minutes against Charlotte?).  xRAPM includes box score stuff--and lots of it--so a guy like D-Mo who finds himself in blow-outs, garbage time, and often against 2nd and 3rd teamers is going to fare a lot better than Terrence Jones who deals with arguably the deepest position in the league on a nightly basis.  D-Mo's defense is so amazing that Coach McHale couldn't will himself to put him in against Aldridge for a single minute when we were desperate for big man defense in last year's playoff series.  But sure, I'll somehow convince myself that D-Mo is a plus defender because DRPM says so.   :unsure:

       

      (I do think D-Mo will be better this year than last, but I think the same about Jones.)

       

      2016 and I have been around this topic more than once (as I am sure most recall).  I am not interested in preventing it's use, but as a moderator here part of the duty entails keeping the group on the level and as informed/educated as possible.  We like to claim the higher standard and I feel part of maintaining that is to keep proper perspective on things like RPM.  It is incredibly complicated and somewhat shrouded in secrecy.  (this also makes me a tad dubious of it's value/validity)  If it is going to be used, I think it is important to have a proper perspective on it.  Giving it the "pay no attention to the man behind the curtain" treatment just doesn't fly with me.

       

      I agree with you that the accuracy of the stat is important to its usefulness (seems obvious enough) and I am curious to see what you have in mind regarding your potential article.

       

      This link that 2016 posted alluding to the "very smart people" who support it is less of an endorsement and more of a discussion group on its flaws, short-comings, and ideas on how to improve it.  The conversation begins back in 2011 and seems to have fizzled out back in Oct. of 2013.

       

      I didn't mean to poo-poo 2016's post.  It is a good post all-in-all.  I just want people to use caution and get educated when it comes to any kind of RPM use.  It's fine to use it in future posts, but you won't fool me putting lipstick on a pig.

       

      lipstick-pig_600x600-400x400.jpg


      • 0

      #9 timetodienow1234567

      timetodienow1234567

        Veteran

      • Members
      • PipPipPipPipPipPip
      • 2,446 posts
      • LocationAlabama

      Posted 22 October 2014 - 10:37 PM

      I disagree. PG is the deepest position IMO.
      • 0

      Why so Serious? :D


      #10 thenit

      thenit

        Advanced Member

      • Members
      • PipPipPipPip
      • 581 posts

        Posted 23 October 2014 - 04:59 AM

        DMo won't succeed for some reason and leave us next year, somehow he will end up being another euro guy ending up at the spurs becoming an allstar.
        • 0

        #11 thejohnnygold

        thejohnnygold

          Veteran

        • Moderators
        • 3,544 posts
        • LocationAustin, TX

        Posted 23 October 2014 - 06:15 AM

        I disagree. PG is the deepest position IMO.

         

        I know you're just being contrary, but I did say "arguably".

         

        Counting the guys who spend lots of time at PF, here is a short-ish list: (based on 2013 positions since that is what we're talking about)

         

        LeBron James/Chris Bosh (depending on line-up)

        Carmelo Anthony (Stoudemire doesn't count anymore, does he?)

        Kevin Love

        Blake Griffin

        LaMarcus Aldridge

        Dirk Nowitzki

        Tim Duncan

        Anthony Davis

        Zach Randolph

        Paul MIllsap

        Nene

        David West

        David Lee

        Pau Gasol

        -----------------

         

        That's 15 all-stars right there.  These guys will light you up every single night.  Guys who didn't make the list?

         

        Serge Ibaka, Thad Young, Derrick Favors, Markieff Morris, Kenneth Fareid, Amir Johnson, Greg Monroe, Taj Gibson, Josh McRoberts, Channing Frye, Boris Diaw, Ryan Anderson, Josh Smith, and Carlos Boozer.  Here are another 14 guys who can all light up their opponent on any given night.

         

        I agree that PG is deep, and possibly deeper than PF.  Don't let that diminish the point I was making.  Playing defense in the NBA at the PF position sucks just about every single night.  The difference between defending these guys and their back-ups (or back-ups back-ups) is HUGE.

         

        This is why seeing Patrick Beverley on that DRPM list at #6 (should be 5, but Nate Wolters snuck in at that spot) is awesome.  We know he plays tough assignments every night and the numbers back up what we see on the floor.  Beverley is #5 for total RPM because apparently Patrick "Everyone says I suck on offense" Beverley sports a better ORPM than John Wall, Kyrie Irving, Jose Calderon, Rajon Rondo, Jrue Holiday, Brandon Jennings, Eric Bledsoe, Jeremy Lin, and Jeff Teague.....wait a minute....what does this number mean again???????   :wacko:

         

        Ish Smith was ranked 81st in ORPM (out of 83)--which after watching him this pre-season I believe.


        • 0

        #12 Steven

        Steven

          Senior Member

        • Members
        • PipPipPipPipPip
        • 1,813 posts

          Posted 23 October 2014 - 02:09 PM

          Hopefully Morey releases Ish Smith.
          • 0

          #13 Willk

          Willk

            Junior Member

          • Members
          • PipPipPip
          • 226 posts

            Posted 23 October 2014 - 07:18 PM

            Hopefully Morey releases Ish Smith.

            everybody has to agree with this
            • 0




            0 user(s) are reading this topic

            0 members, 0 guests, 0 anonymous users