Back to Square Zero

Posted by Rob Mahoney on August 5, 2010 under Commentary | 63 Comments to Read

Screen shot 2010-08-05 at 3.59.35 AM

Positional certainty has never been a luxury the Dallas Mavericks could afford during the Dirk Nowitzki era. Yet year after year, the team’s flaws are diagnosed according to the standards of a conventional lineup. Dallas needs a better center. A better shooting guard. A better point guard. Hell, anything that isn’t power forward. Dirk has been the one constant, and despite his unconventional and unique talents, the success of his team is ultimately measured by way of an antiquated tradition.

No longer. Or at least as minimally as possible in this space.

It may be naive to think that the mainstream basketball audience will soon abandon the five conventional positions, but that doesn’t mean those of us in this corner of the universe can’t strive to be better, smarter basketball fans. I’m ready to take a hop (more than a step, but well short of a leap) in the way we classify players. With that, I’ll cue Drew Cannon of Basketball Prospectus:

But what do you really need from a lineup?

On defense, you need to be able to guard your opponents. This means you have to be ready for speeds and heights of all kinds. You need to have a player capable of guarding each of the five traditional C-PF-SF-SG-PG positions. We’ll call the players capable of defending each position “D1” through “D5,” respectively, with speed/athleticism on the x-axis and height/strength on the y-axis:

100802_positions

And on offense what do you need to be successful? You need to be able to make shots (from the field or free throw line), avoid turnovers, and clean up the offensive glass–at the very least to the point where you aren’t handing over points by doing the opposite. This means that you need someone who can take care of the ball, someone who can put it in the basket, someone who can get the ball to that guy, and someone who can get the ball back when someone misses. We’ll call these four characters the Handler, the Scorer, the Creator, and the Rebounder.

Quick point. The Creator and the Handler have to be the same guy. Because you can’t have your Creator losing the ball all the time before he can feed your Scorers, and you can’t have your Handler with the ball all the time but unable to get it to the Scorers.

…It boils down to this: On defense, you have to be ready for whatever the offense throws at you. But on offense, you really just need to rebound and protect the ball enough to let your scorers go to work (or protect the ball just enough that your dominant rebounding can keep putting points on the board despite below-average scoring, etc.). Really, how you put points on the board is your business. The defense is just reacting.

This is more than just a quaint idea.

I’m sure Cannon’s model isn’t a perfect one, but it doesn’t have to be. It’s a start, and nothing more. Just as the traditional formula yielded point-forwards (or even point-centers…word up the the Antoine Walker experiment), combo guards, and other atypical cogs, I’m sure that this framework will allow for a few more offensive player designations yet. What matters is that we move away from a nondescript and misleading method of classifying players in favor of something — anything — that actually manages to advance basketball discourse.

To those still clinging to what they know, I’d ask this: what’s a power forward? What characteristics link Dirk Nowitzki, Tim Duncan, Rashard Lewis, Lamar Odom, Reggie Evans, Tyrus Thomas, and J.J. Hickson? Not rebounding. Not scoring. Not skill set. Not height relative to their teammates. Not even the spaces they occupy on the floor. I’m at a total loss as to the criterion that would group that bunch together, which makes the assessment “Player X isn’t a real power forward” pretty much worthless. I think I know what it means, but without the ability to define the contemporary power forward, how could I really know for sure?

Conceptually, this is nothing new. Players like Dirk have been bending positional bounds for years, and the basic tenets of fluid positionality have been preached by a number of NBA scribes. Yet this system makes enough intuitive sense to work, and gives the thought a more practical and literal application.

If you’d like to join me on this little adventure, I’d love the company. If not, that’s fine, too. This post isn’t meant to convert, but primarily to do two things:

  1. Inform as to what the hell I’m talking about when I write that “Jason Kidd is a D2,” in the future.
  2. Bring the idea to the forefront. Even if you’re not ready to buy into an overhaul of positional classifications, I hope this at least gets you to think about what those classifications mean (or don’t mean).

This could be fun, but I’m going to need a lot of help. Here are the initial offensive and defensive positions for all of the current Mavs according to my own assessment, but they’re not infallible. Are there offensive profiles that aren’t represented? Is it fair to list Shawn Marion strictly as a rebounder? Or Jason Terry as a D2? Let me have it. Rip this idea to pieces. Tear it down so we can build it back up with stronger and smarter ideas, making our collective analysis that much better in the process.

Alexis Ajinca – D?, Large body
J.J. Barea – D1, Scorer-Creator/Handler
Rodrigue Beaubois – D1, Scorer
Caron Butler – D3/D2, Scorer
Tyson Chandler – D5, Rebounder
Brendan Haywood - D5, Rebounder
Dominique Jones – D2/D1, Scorer
Jason Kidd – D2/D1, Creator/Handler
Ian Mahinmi – D5/D4, Rebounder
Shawn Marion – D3/D2/D4, Rebounder
Dirk Nowitzki – D4, Scorer-Rebounder
DeShawn Stevenson – D2, Abe Lincoln tattoo
Jason Terry – D2 (I guess?), Scorer

  • Kirk Henderson

    I like it. New discussions tend to be helpful in the long run for getting a better understanding of why something does and doesnt work.

    I really like the D# system. I think it’s great. My suggestions as for how they pertain to the Mavs:

    -Butler is strictly a D3. He cannot guard quick guards. He is who he is.
    -Terry has to be qualified as a D0. I don’t mean it disrespectfully, but he can’t guard anyone anymore. I could be very wrong, but I would venture that over longer stretches the Mavs are worse defensively with Terry on the floor and he is usually the weak link.

    • http://www.thetwomangame.com Rob Mahoney

      @Kirk I agree that Butler is much better at guarding 3s than 2s, and originally had him listed as a D3. I guess the point of contention is how effective a defender has to be to qualify as a particular designation; Butler isn’t great at guarding 2s, but he’s not awful, y’know? I also like the ‘D0′ classification for Terry and in general.

      • http://www.thetwomangame.com Rob Mahoney

        @Nick: They’re hybrid defenders. I think LeBron might be a D3/D4/D2/D1/D5, honestly.

    • http://www.thetwomangame.com Rob Mahoney

      @Kirk: I agree that Butler is much better at guarding 3s than 2s, and originally had him listed as a D3. I guess the point of contention is how effective a defender has to be to qualify as a particular designation; Butler isn't great at guarding 2s, but he's not awful, y'know? I also like the 'D0' classification for Terry and in general.

  • http://basketballoutsider.wordpress.com Rafael Uehara

    What if we start separating players by their functions.
    for example: Steve Nash is not only a ball handler but a playmaker. Not just by pushing the ball or running the high pick-and-rolls, he’s his team main way to generate offense.
    Kyle Korver is a wingman. He constantly moves around the perimeter to find an opportunity to shoot not to drive.
    Dirk & Pau Gasol are low post presences because their teams isolate them there and let them work.
    Andrew Bynum is not. He’s cleanup guy and a rim protector. You hardly ever seen the Lakers isolate for Bynum unless it’s a screaming matchup advantage. He’s there to complete Gasol.
    The PG, SG, SF, PF, C way to characterize players is quash. Hopefully we can start separating them by their purposes on the floor.

    • http://www.thetwomangame.com Rob Mahoney

      @Rafael I like the idea of creating Shooter as a subset for Scorer, and Korver would for there’d rather comfortably. However, I’m a bit reluctant to describe players as post-up guys or whatnot in this analysis just because I feel our most natural instincts would limit their application. Kobe is one of the best post-up players in the game, but are you really going to list him as such?

      Otherwise, Nash is a Handler/Creator (the Creator tag seems like it encompasses your points nicely), Dirk and Pau are both Scorer/Rebounders, Bynum is a Rebounder, and Korver a Scorer (Shooter). I see where you’re going, but I think these labels cover many of the same bases.

    • http://www.thetwomangame.com Rob Mahoney

      @Rafael Uehara: I like the idea of creating Shooter as a subset for Scorer, and Korver would for there'd rather comfortably. However, I'm a bit reluctant to describe players as post-up guys or whatnot in this analysis just because I feel our most natural instincts would limit their application. Kobe is one of the best post-up players in the game, but are you really going to list him as such?

      Otherwise, Nash is a Handler/Creator (the Creator tag seems like it encompasses your points nicely), Dirk and Pau are both Scorer/Rebounders, Bynum is a Rebounder, and Korver a Scorer (Shooter). I see where you're going, but I think these labels cover many of the same bases.

  • http://Www.premiersportstalk.com Rob Howe

    I hear you. The Mavs have been struggling with the concept of positional fluctuation for some years now, and it would appear that they have the right pieces yet again to make a solid push, except for one problem.

    Even in a D# system comprised of unorthodox lineups, to win a championship your best scorer cannot be your worst 1-on-1 defender, which Dirk is at times (when Terry isn’t on the floor). The thought has been that by surrounding him with more defensive minded players (Haywood, Chandler, Kidd, Butler, Marion) that they would be able to cover his flaws. And it hasn’t worked so far. The Mavs will make a good run, but then get beat by anther unconventional lineup in the OKC Thunder

    • http://www.thetwomangame.com Rob Mahoney

      @Rob Howe: Why can’t your best scorer be your worst one-on-one defender (as long as he’s not a complete liability)? I just see any explicit or implicit reason why that has to be the case. Plus, Dirk isn’t a horrible defender. He’s right around average, so it’s not as if his weaknesses are dooming the Mavs on their own.

    • http://www.thetwomangame.com Rob Mahoney

      @Rob Howe: Why can't your best scorer be your worst one-on-one defender (as long as he's not a complete liability)? I just see any explicit or implicit reason why that has to be the case. Plus, Dirk isn't a horrible defender. He's right around average, so it's not as if his weaknesses are dooming the Mavs on their own.

      • Lala

        Dirk is overrated. He is a loser and if he is the best player on any team the team will never win a title.

        Mavs fans are biased and pretty stupid overall. Face the facts. Your time is over.

        • Andy

          How do those words taste now, son?

      • Quetzpalin

        I don't see this either. Paul Pierce has arguably been the Celtics worst defender and best scorer over the last 3 years, but it's worked for them.

  • John

    Your non-synopsis of the power forward position makes a case for what I have always thought. With Dirk as the focal point, the Mavericks are a team best suited to have two PF’s on the floor at the same time. But a player with a different skill set than Dirk’s. I know a lot of people would disagree, but I believe a player with skills similar to Tyrus Thomas would be a great fit alongside Dirk. That is a fast, athletic, shot blocker with an attitude that could cover a 3,4 and some 5′s, as well as pull down a few offensive rebounds and occasionally contribute offensively with some big dunks and put backs.
    Also, I think you could put JJ into that (I guess) defensive category along with Jet. And I’d like to see you make an exception on that Creator/Handler category for Roddy and make his designation as Creator/Scorer.

    • http://www.thetwomangame.com Rob Mahoney

      @John: I think a player like Tyrus Thomas could fit in well next to Dirk, provided you don’t ask Nowitzki to be a D5. Slot in Tyrus as a D3, and it could definitely fly.

      As for Beaubois, he’s not a creator yet. I like where you’re going with the idea, but his playmaking skills aren’t quite strong enough to grab that label just yet, in my opinion. Then again, maybe you nailed it. Maybe it’s Beaubois’ handle that’s holding him back, not his creativity or vision. Something to think about, for sure.

    • http://www.thetwomangame.com Rob Mahoney

      @John: I think a player like Tyrus Thomas could fit in well next to Dirk, provided you don't ask Nowitzki to be a D5. Slot in Tyrus as a D3, and it could definitely fly.

      As for Beaubois, he's not a creator yet. I like where you're going with the idea, but his playmaking skills aren't quite strong enough to grab that label just yet, in my opinion. Then again, maybe you nailed it. Maybe it's Beaubois' handle that's holding him back, not his creativity or vision. Something to think about, for sure.

  • econlibVA

    I agree that the positions we currently use are wrong, but I still think that these divisions aren’t what’s most important. Instead, we should think about offense in terms of specific skills and defense in terms of abilities to stop those skills. On offense, I think the skills are:

    - shooting
    - low-post scoring
    - (offensive) rebounding
    - screening
    - passing
    - driving

    Defensive skills would then be described as the ability to stop those skills.

    Note that there are a lot of ways to put together an effective offense. Orlando is my favorite example, as they generally use one shooter/driver/passer (Jameer Nelson), three shooters, and one low-post/screener/rebounder (Dwight Howard). This is damn near unstoppable, and Orlando was the second best offensive team in the league last year.

    Another example is Phoenix, who uses one shooter/driver/passer (Steve Nash), three shooters, and one shooter/screener/rebounder (Amare Stoudamire). They were the best offensive team in basketball last year.

    We really need to study what combinations work and don’t work on offense. In some ways, we waste the talents of Nowitzki by not putting the best players around him. My gut tells me that we need to look for more shooters to put around Nowitzki. If it were up to me, I’d have reacquired Bass and signed Morrow with the idea that we’d have the best shooting team in the league. What we’re currently doing isn’t going to get us to a championship, so why not try something different?

    • http://www.thetwomangame.com Rob Mahoney

      @econlibVA: I think using that skeleton for defensive skills would be a bit too complex for what’s supposed to be a short, positional designation. Something like D5 may not tell you everything, but we also don’t want to mistake this as a method for thorough and complete skill analysis. It’s basic by definition.

      Definitely considering Screener as another designation, though I think there’s some overlap with Rebounder. Obviously they’re very different skills, but most players who do one do both. In the name of trying to keep this as simple as possible, labels like that may be the casualties.

    • http://www.thetwomangame.com Rob Mahoney

      @econlibVA: I think using that skeleton for defensive skills would be a bit too complex for what's supposed to be a short, positional designation. Something like D5 may not tell you everything, but we also don't want to mistake this as a method for thorough and complete skill analysis. It's basic by definition.

      Definitely considering Screener as another designation, though I think there's some overlap with Rebounder. Obviously they're very different skills, but most players who do one do both. In the name of trying to keep this as simple as possible, labels like that may be the casualties.

  • Pingback: Finding position in a positionless world | Daily Thunder.com

  • Nick

    What about guys like LeBron, who is a 5 at speed/athleticism, and probably a 4 or 5 for height/strength?

    Over the curve means defensive versatility and under the curve means defensive liability?

    • http://www.thetwomangame.com Rob Mahoney

      @Nick: They're hybrid defenders. I think LeBron might be a D3/D4/D2/D1/D5, honestly.

  • finzent

    I’m aware of all the caveats you added, I welcome the discussion and am a big fan of your power forward example. With that said: I really don’t think the D# is any more valuable than just using traditional positions. The reason for that is that it classifies defenders by their ability to defend the classical positions in their most orthodox interpretation. I don’t see where the added value of doing this would be.

    One way of making it a bit more flexible: get rid of the numbers, scale the x and y axis and classify defenders by the range of points on the grid the are expected to be able to defend.

    • http://www.thetwomangame.com Rob Mahoney

      @finzent: I thought the same thing about the defensive designations at first, but try this:

      Don’t think about D1, D2, etc. As being able to defend the 1, 2, etc. Think about it as groupings based on size, speed, and athleticism. It’s similar, but not quite the same. This system is basically a less complicated and quantifiable version of the model you’re suggesting.

    • http://www.thetwomangame.com Rob Mahoney

      @finzent: I thought the same thing about the defensive designations at first, but try this:

      Don't think about D1, D2, etc. As being able to defend the 1, 2, etc. Think about it as groupings based on size, speed, and athleticism. It's similar, but not quite the same. This system is basically a less complicated and quantifiable version of the model you're suggesting.

      • finzent

        Yeah, but I think that when we normally talk in terms of traditional positions and say things like “Kidd looks bad against most PGs, but he does OK against most SGs” we mean nothing very different from “Kidd's a D2″; I think what the D# system does is simply spelling out what usually determines our talk about positions (maybe there is a slight difference, but I strongly suspect we don't gain much from the change in expression).

        I had another point in mind but now I noticed an ambiguity in the description of D# above; does, for example, D2 mean that a defensive player has (roughly) that combination of speed/athleticism and height/strength, or does it mean that a defensive player can defend offensive players with that combination. I automatically took it to be the latter, but now I see that the former might also be correct. (There are problems with both interpretation, but I didn't want to make my point and then look stupid).

        • http://www.thetwomangame.com Rob Mahoney

          I agree, saying “Kidd's a D2,” isn't all that different from saying he's better against most SGs but looks bad against most PGs. However, the value of this system is in establishing cross-matches. Kidd is a point guard, but how often does he actually guard point guards? That designation is very misleading in terms of a typical defensive assignment (obviously there would be tweaks on the basis of specific match-ups), whereas the DX listing is a bit more descriptive of actual defensive roles.

          Good call on the ambiguity. I'd say that a defensive position indicates that a player can defend offenisve players with that combination, as you mentioned. Chuck Hayes may have the size of an NBA shooting guard or small forward, but he can legitimately defend those at the high end of the height axis. If I explained anywhere that it was the former, it was in error.

          • finzent

            Ah, ok, I think that's the better reading, too. You didn't imply it was otherwise, it just wasn't totally clear from Cannon's description.

            The point I wanted to make was just that, even if the D# talk is better/different from old fashioned positional talk, it still relies on a somewhat rigid taxonomy; my impression was that the whole point of trying to break up the old way of seeing positions was to account for the fact that there are all kinds of varieties of players. So, there might be “pure” D2s on offense just like there might be “pure” SGs, but they certainly are the minority. I think the real fun starts only when you do something like Phil downthread or myself above suggested: try to order defensive players by the “area” of player types they can be expected to cover.

            I don't want to grumble, of course, the fact that this is even discussed in this way is just great.

  • Aleks

    I’ve been waiting for a discussion like this for a long time.

    The one additional thing I want to say is this: This ‘taxonomy’ of players might be much better than the traditional one, but it is still a static system. We need to keep in mind that players are not static and can change on timescales of months or years. There are many examples for this, Marion has basically abandoned his shot in Dallas and was a scorer just two years ago. Kidd’s surprising revival as scorer, etc. Also, what is not static are coaches and roles which are available in their system. The triangle offense, for example, doesn’t really need a specific creator.

    This is important when trying to figure out what type of players to put around Dirk. Over the years Dirk has developed more and more into a halfcourt, high post player, for various reasons that would be interesting to look at. At this point he doesn’t need a a lot of ‘creating’ anymore, which is the reason the Kidd trade didn’t have a huge effect. Since the Avery-Johnson era, Dirk plays a game that is best complemented by a group of defensively strong shooters/slashers. Not surprisingly, the Mavs got the best results with a comination of Terry/Howard/Harris/Stackhouse in 06 and 07.

    That team, of course, had other issues which are not covered by the positional scheme anyway. The current Mavs team is so weirdly constructed that I have no bloody idea what the hell they are going to do this year.

    • http://www.thetwomangame.com Rob Mahoney

      @Aleks: I think an adoption of this system should come with an acknowledgement that these roles are more dynamic than traditional positions. Like you mentioned, players constantly evolve in their skills and responsibilities, and while most would still be hesitant to shift Player X from power forward to small forward accordingly (after all, with little meaning to those designations, what would be the impetus for the change?), that doesn’t mean we have to be. Beaubois could be a Handler/Creator by February, or maybe a slimmed down Butler would be a more convincing D2.

      As for Kidd’s impact as a creator, I think the idea stems from a desire to create for the non-Dirk Mavs. Nowitzki can get a shot when he wants, but Josh Howard, Jason Terry, and even guys like Erick Dampier stood to benefit most (and in some ways have) from Kidd’s arrival.

    • http://www.thetwomangame.com Rob Mahoney

      @Aleks: I think an adoption of this system should come with an acknowledgement that these roles are more dynamic than traditional positions. Like you mentioned, players constantly evolve in their skills and responsibilities, and while most would still be hesitant to shift Player X from power forward to small forward accordingly (after all, with little meaning to those designations, what would be the impetus for the change?), that doesn't mean we have to be. Beaubois could be a Handler/Creator by February, or maybe a slimmed down Butler would be a more convincing D2.

      As for Kidd's impact as a creator, I think the idea stems from a desire to create for the non-Dirk Mavs. Nowitzki can get a shot when he wants, but Josh Howard, Jason Terry, and even guys like Erick Dampier stood to benefit most (and in some ways have) from Kidd's arrival.

  • John

    Right on Rob. Tyrus, or a similar player, and Dirk switching rolls on offense and defense. Along with a D5 on the floor.

    I believe that’s the direction they’ll be headed with Kidd and Beaubois. But also with Kidd as a coach on the floor training Roddy to be more of a floor leader and facilitator.

    As far as Roddy handle, I think he’s got one of the best in the league. It’s his decision making that’s holding him back, kind of like keeping the brakes on an Indy car around the track.

    Hopefully by the time Kidd’s contract runs out, the facilitator part of the game will come more naturally to Mr. Beaubois.

    All that being said, I still think the Mavs have one of the deepest teams in the league. And I hope they give Mahinmi a chance to develop at the 4 before the playoffs come around. I liked the toughness and ability he showed against Cousins and think he could be a valuable piece off the bench in the post-season.

  • econlibVA

    Rob – OK, I think I’m on board with your defensive designations. D1-D5 is simple, even if it doesn’t quite get what you need for skill analysis.

    That said, screener and rebounder are very different skills. Some guys are good rebounders and AWFUL screeners – I think that was part of Drew Gooden’s problem here in Dallas. Also, part of screening is the ability to roll to the basket, catch the ball, and make the layup after the screen. Dampier actually got good at that, but there are a lot of good rebounders who are kind of useless because they can’t shoot and they can’t catch. Conversely, there are some good screeners that aren’t good rebounders – it is harder for me to give a good example here as I don’t watch enough basketball games for close details like that.

    • http://www.thetwomangame.com Rob Mahoney

      @econlibVA: I hear you. No doubt that it's a different skill that some players excel at and others don't, I'm just trying to figure out if it's worthy of its own designation. I agree in principle, but just wonder if it's a necessary addition to the position pool.

  • http://www.thetwomangame.com Rob Mahoney

    @econlibVA: I hear you. No doubt that it’s a different skill that some players excel at and others don’t, I’m just trying to figure out if it’s worthy of its own designation. I agree in principle, but just wonder if it’s a necessary addition to the position pool.

  • J

    The problem is this:
    If your Scorer is a D5′s are mostly Centers, and D2′s are mostly Scorers and Shooting Guards. When you look on your example, you see that every Scorer is either a D1/D2/D3, so he’ll guard somebody in the Backcourt. So it’s still way easier to just say “Shooting Guard” because you know that they’ll mostly be D2′s and Scorers.

    • http://www.thetwomangame.com Rob Mahoney

      @J: I don't think that's true at all. Dirk is the Mavs primary scorer and he clearly isn't defending a “guard” or “wing.” Likewise, the difference between a D1 and D3 is pretty substantial, so to just say “shooting guard” ignores the incredible variance between those defensive positions. A D1 Scorer is not a shooting guard, and neither is a D3 Scorer. Shooting guard may be easier to say, but the problem is that the term means way too little .

  • Aleks

    I guess my main problem with this positional scheme is the handler/creator part. All other parts are certainly essential for a successful NBA team; over the course of a season it’s impossible to be successful without D1-D5 and obviously without scoring/rebounding. But ‘creating’ in the sense of a floor general who runs plays is not really necessary, depending on the other players and the type of system they are playing. A versatile roster can easily be successful without dedicated handler/creator (the Heat are going to prove this).

    The second possible refinement is to put more emphasis on the type of scoring, as suggested already. A balanced roster needs to be able to score from all areas on the court, I suppose, so, for example, one could distinguish between shooter and finisher, or something like that.

    By the way, with the current Mavs roster it’s pretty clear that the problems are going to be D1 and scoring.

    • http://www.thetwomangame.com Rob Mahoney

      @Aleks: Even if it's not essential in your eyes to have a creator/playmaker, most team models have one. It may not be absolutely mandatory, but I think the Creator is common enough that it needs to be accounted for. However, I think your contention about the Heat is very, very wrong. LeBron and Wade are both incredible Creator/Handlers, even if neither is ever plugged in as a “PG.”

      But yeah, working on the more focused scoring categories.

  • http://thenullhypodermic.blogspot.com/ Brian Tung

    Interesting proposal. And I can see a spectrum of player notations depending on whether your criterion is precision or concision.

    For instance, if I were devising a system for use in computer analysis, I might characterize players by their speed, acceleration, reaction time, vision (both at 1 s and 10-100 ms time scales), shooting accuracy (as a function of distance), finishing at the rim, hand speed, vertical leap, distractability (again at different time scales), etc. In other words, the same sort of system that you might use as a video game developer. Those parameters would define a space of possible players. Defense would be a kind of function over that player space, or possibly player-cross-action, denoting how well the defender can stop that player from performing that action.

    Such a system, when cleaned up, would undoubtedly furnish higher fidelity than any existing system, and could be used to guide GMs in their personnel decisions, and coaches in their strategic ones. However, it would obviously be tremendously unwieldy as a succinct player notation.

    Nonetheless, I think the idea of characterizing players as having a certain offensive mold, and being more or less able to defend that or other molds, is a good approach, even if you give less detail. Might need a more flexible design if you want to incorporate zone defensing on an ongoing basis, but it’s a start.

    • finzent

      “For instance, if I were devising a system for use in computer analysis, I might characterize players by their speed, acceleration, reaction time, vision (both at 1 s and 10-100 ms time scales), shooting accuracy (as a function of distance), finishing at the rim, hand speed, vertical leap, distractability (again at different time scales), etc. In other words, the same sort of system that you might use as a video game developer. Those parameters would define a space of possible players. Defense would be a kind of function over that player space, or possibly player-cross-action, denoting how well the defender can stop that player from performing that action.”

      Someone else made a similar proposition above, and the defensive part of that bugs me. I certainly see the merits of something like that on offense, but I think analysing defensive aptitude as the ability to stop certain offensive skills is severely flawed because it totally disregards the respective positions (in a neutral sense) of the players. Seeing defensive skill in that way would mean that, for example, Player X has a certain value in defending certain skills when a better way to see it would say that Player X has a certain value in defending certain player types. You can't say “X defends jump shots well, passing ok and penetration average”, because in reality Player X will only defend certain jump shooters, passers and penetrators, namely those who play his position (where “play his position” may be seen as “is close to his combination of speed/athletics and height/strength”, like in D#).

      • http://thenullhypodermic.blogspot.com Brian Tung

        @finzent: I may not have made this clear, but I'm proposing (in a very rough sense) defense as a function not of action, but of PLAYER x ACTION. In other words, how well does the player defend a particular player type, attempting a sort of action? So both aspects are covered by this. As I say, this would make the player characterization very cumbersome as a medium of conversation, but would be perfectly acceptable to, say, a computer application that evaluated a player's fit for a particular team.

        • finzent

          Actually, I think you were pretty clear in your first comment and I just wasn't reading as closely as I should have. I guess I reacted more to something said above, and just kind of assumed you were saying the same.. My bad.

  • Hhammer

    I agree with how you classified the offense. People should be defined by universal characteristics such as playmaker, shooter, post etc. The problem I have is defining the defense in terms of their defensive assignments. If we switch to the a non-conventional offensive naming system then who does a D2 cover when there no longer is a SG. Who would a team’s starting D2 cover, the person whose height they match up with or the person’s whose talents they match up with. Rather the defense needs to be named by what height and what skills that person is able to defend. Though I understand that it would be very complicated to have players designated as say a 6’0 to 6’6 perimeter defender, in your system you might as well simply called the D1 a PG and a D2 a SG for defensive purposes. Can’t change up the offensive system while keep the defense the same because they should ideally relate to each other.

    Good read though.

    • http://www.thetwomangame.com Rob Mahoney

      @Hhammer: Mentioned this in one of the comments earlier, but D1 does not mean “able to defend 1s.” It means the player is able to defend players in a certain size/speed range. It's not perfect. It's not exact. But it's also not the same as saying “so-and-so can defend point guards, but not shooting guards.” There's a basic difference there that makes this system passable, in my mind.

  • Aleks

    I thought this should be mentioned here:
    http://freedarko.blogspot.com/2009/04/dream-ang

    I like the Magic Z better, because it's visually more attractive and has some of the additional things mentioned here (post-scoring, threes, drives). The approach for defense is here the same as four offense – split it up in specific abilities, instead of the less accurate D1-D5 scheme. It would be great to develop Z's for all Mavs players and then coadd the Z's for the entire roster, to see in which areas the Mavs are great and where do they have problems. In theory, this should then determine the type of game they should play. So, if someone has too much time…

    • http://www.thetwomangame.com Rob Mahoney

      @Aleks: A personal favorite FD venture of mine. Tom Ziller is a madman.

      It's cool to imagine a world where the Z is readily adopted, but it's not the easiest tool for actual discussion. It's a great visual descriptor, but not all that easy to talk about/describe. I think it absolutely has a place in this conversation, I'm just not quite sure where it fits in quite yet.

  • Aleks

    Imagine, we could have a Magic Z next to the player's photograph at nba.com. And one for each team. That would be awesome.

  • Dan

    Interesting viewpoint but I still think the similar problems exists here as with the traditional classification. First, the classification of a D2, D3, ect assumes that offensively all 2′s or 3′s are equally skilled. You face the same problem you’re trying to solve with the new classifications. Jason Kidd may be predominately a D2 but not always. He can’t guard Arron Brooks but he can guard Chauncey Billups or Derick Fisher. You can find examples of this all over the league. Why wouldn’t Chandler be a D4? He can certainly guard 4′s like Amare Staudimire and Pau Gasol who play a traditional post position but would have more problems with Dirk and Odom who tend to face the basket more. All shooting guards don’t have to be able to shoot well and all small forwards don’t have to be bad at rebounding. What you do need is a balance of size, speed, shooting, and defense. So if a 2 guard isn’t a good shooter but is a great defender(Thabo Sefolosha) he will fit best on a team with a 3 who can shoot. Having a team who has the right balance of the necessary skills is much more important than a certain position possessing a skill that their position stereotypically is good at.

    • http://www.thetwomangame.com Rob Mahoney

      I don't think D2 necessarily means a player is actually able to defend all players that fit into that size/speed classification, just as being a shooting guard doesn't mean Player X can defend all shooting guards.

      Not a flawless system, so still working out the kinks.

  • Ron

    From a more simplistic view… each traditional position is primarily defined by height/weight (and quickness). The biggest guy is the center, the smallest guy is a point guard. Small forwards and shooting guards are often interchangeable. Power forwards and centers are as well.

    It all comes down to matchups. Take Tim Duncan and Pau Gasol, for example… both are considered PF or C depending on who they are matching up against.

    Taking it further: a players position label can change on offense vs. defense, but again, depending on who they’re matched up against.

    So, I would argue that positional labels are mostly a matter of conveniece for conversation.

  • Kaanon

    I think the league has been a league of 1 point, 2 wings, and 2 bigs for a while now.

  • Phil

    Hey Rob, first of all, I love the statistic-analytical approach! In a game that continues to adjust to offensive arsenals rather than (relatively stiff) position attribution, this is a great way to specify team needs and (defensive) depth. I

    think it has to be adjusted in terms of particular offensive and defensive skillsets (as already mentioned [screens, shooters, maggette-type-slashers]).

    but first, I have to agree with inzent: Why “sticking with d1-5 (even if understood not as the positions but rather the specific attribution)”? How about saying that kidd is a 2/2 (meaning he is relatively strong for a pg, but also very slow) and butler as a 3/3, Roddy had to be a 5/1 (or even 6/1) and so on… the way you do it, you have barea (who in this case is a 4/1 at best) listed the same as roddy. Just rate and classify them within the coordinate system leaving us with a more accurate approach.
    keep it up, Phil

  • http://twitter.com/KirkSeriousFace Kirk Henderson

    This has been an awesome re-read first thing in the morning. The fluidity of basketball is why its so appealing.

  • Pingback: Heard It Through the Grapevine | The Two Man Game

  • GrimBaller

    intresting idea, but its really pointless when u think about it.

    traditional position already tell u who does what on offense.

    pg = takes care of the ball

    sg/sf = score form perimeter/slash to the basket

    pf/c = score from inside/grab off rebounds

    on d its called defensive matchups n it has nothing to do with players positions. guarding otehr teams pg doesnt make u a pg. it means u have the ability to do guard that position.

    when kobe guards rondo n fisher allen, does that mean that kobes pg n fisher is sg? or does it just mean, fisher cant keep up with rondo, so kobe took that matchup?

    now for the “whats a pf” example:

    thats just the style of play.

    just like theres pass 1st pg like nash, n then there are score 1st pg like parker. their styles are different, but they play the same position.

    same with dirk, lewis n to some extent duncan. some like mid range game, others like to bang inside.

  • Pingback: Positional revolution: Examining roles and positions on defense | The Two Man Game

  • Pingback: Heard It Through the Grapevine | The Two Man Game

  • Pingback: Positional revolution: Kobe Bryant envisions a basketball world with no positions | The Two Man Game

  • Pingback: Rotaciones Conferencia Este (y III) « Manager Fantástico

  • Pingback: A thesis on point guards | 48 Minutes of Hell

  • Pingback: The Position Revolution and the Myth of Height in Basketball Brian McCormick Basketball

  • Pingback: The Difference: Dallas Mavericks 116, Houston Rockets 109 | The Two Man Game