Layups: New Thoughts on Positional Designations
Posted by Neil Paine on August 6, 2010
Drew Cannon wrote a very interesting post for Basketball Prospectus on Monday, regarding player positions. His theory is that coaches should break positions down offensively and defensively, since the former deals with certain specific skillsets (scoring, passing, ballhandling, & rebounding) while the latter is concerned with what level of opposing height and speed a player can defend. Instead of worrying about whether a scorer is in the body of someone who can defend a SF or a PG, Cannon contends (and I'm inclined to agree) that as long as you fill all of the necessary roles on both sides of the ball, it doesn't matter who does what job and whether their defensive position matches up with the traditional offensive role of that slot. Anyway, it's a really good read, so check it out and let me know what you think about traditional positions vs. Cannon's idea.
August 6th, 2010 at 10:32 am
Constantly finding and getting to your man after an offensive possession can be a real pain in the butt. But a pro level, it should not be a problem. I think we'll see the Heat doing a fair bit of this in the coming season.
And to throw a little bit of professional credibility behind this (because, as cool as his name is, I have no clue who in the hell Nate Funk is), we saw this work out great for the Bulls in the Finals when they put Pippen (3) on Johnson (1). Or when the Lakers put Bryant (2) on Rondo (1). It's all about getting the good match up.
Still, I feel like this does overlook the fact that there's not a whole lot flexibility with the 4 and 5 position. Those guys need to rebound and provide interior D. And they're not really fast enough to be doing otherwise.
August 6th, 2010 at 10:33 am
I think scorer is too broad a term. There are 3 distinct types of scorers: shooter, driver and low post. Obviously, some players have multiple skills. However, Reggie Miller and Hakeem were both scorers, but a team full of low post guys would be easily stopped. The success of the Magic in 2009 is due to having a great balance of shooters and post presence.
August 6th, 2010 at 11:06 am
Right, good point -- you can be a shooter but not necessarily a scorer (J.J. Redick), you can be a perimeter scorer but not necessarily a great shooter (Dwyane Wade, LeBron James), or you can be an inside scorer who never ventures out of the post (prime Shaq). You probably need to fill at least two of those roles, although maybe at its most elemental, you just need a "shooter" and a "scorer".
August 6th, 2010 at 11:46 am
I'm all for a re-examination of traditional position designation, but I also think the NBA is sort of a different animal from college because of the lack of a pure zone and the overall abilities of the average players.
In theory t's fine to say, "Hey we got a dynamo scoring guard who happens to be 6' 2". Perfect we'll just get another player to be out point, and we're good to go." The problem isn't so much acceptance of the short scoring guard as it is finding complementary pieces to fill in defensively. How many Jason Kidd's are out there? Somebody in your backcourt has to be able to defend bigger guards who can hurt our hypothetical 6' 2" wunderkind scoring machine at the other end, because in the NBA just about any team could exploit that matchup.
Most of your capable point guards are small. Under 6' 4" tend to have the best handle, which, as Kevin notes, puts them in the position of being a playmaker. A quick run through of all the true point forwards who could play a Pippen / Magic type roll and successfully defend a non-point guard at the other end. LeBron. Maybe Roy. Nobody else immediately comes to mind.
August 6th, 2010 at 12:24 pm
Cannon's theory has some major flaws. First, having a cross-match at every position would be chaos defensively. One or two is manageable - as mentioned by the first commenter - but when everyone is running back on D searching for their man, someone is bound to be a step slow. You'd have to have a really cohesive defensive plan for this to work.
Second, his defensive matchups are based on traditional positions. Proposing a theory eschewing traditions that itself is based on traditions will eventually fall apart once enough people adopt that theory.
August 6th, 2010 at 12:27 pm
Great comment, Jason. That crossed my mind as well -- if you play, say, Jason Terry as your SG, it means he has to guard the opposing PG, which means you have to find an offensive PG who can guard SGs. There are big guards out there, sure:
http://www.basketball-reference.com/play-index/tiny.cgi?id=alHxo
But you might also end up with an Emanual Davis type. So yeah, it definitely presents a unique challenge when you play a tweener combo guard as your SG.
August 6th, 2010 at 2:45 pm
Chris - I really don't see where what he's talking about is cross-matching at all. It's just non-traditional. To stick with the Kidd / Terry example, when they play a conventional backcourt like Boston, Kidd covers the bigger guard, Allen, while Terry covers the smaller guard, Rondo. It's not as though Boston then reverses this matchup at the other end. Rondo guards the player who is his size, Terry, while Allen covers the larger Kidd. No one is scrambling to find a different defender when the ball changes ends.
As for the size designations being D1, D2, etc. I didn't really read that as an attempt to normalize to traditional positions so much as an acknowledgement that the traditional positions are largely sized-based. If you looked at a roster and your five best players were all D2s, D3s, and D4s (like the 1996 Bulls for instance), you'd know there could be problems covering very quick players and very big players, but that doesn't necessarily mean your team is missing a point guard or a rebounder. Pippen and Rodman might both fit into the D3 category size-wise and have the ability to comfortably defend D2s and D4s respectively. So while Pip can playmake with the traditional D1s and Worm can outrebound most tradition D5s, you have to consider that defensively they should be, for the most part, in their own basic size range. This is by no means a problem, just a consideration.
August 6th, 2010 at 8:04 pm
Cannon's post hardly amounts to "new thoughts," but rather takes a fairly common idea (that offensive and defensive positions can be split) and dresses it up with a chart and some fancy-sounding abbreviations. He even admits that it is common at the point where notes Jon Scheyer and Nolan Smith. There is not really a new or interesting theory there...
What there is, though, is a potentially very interesting empirical claim (not a theory) that the possibility of splitting offensive and defensive positions is overlooked during recruiting, leading programs to pass on good players. If true, this is very odd...
August 6th, 2010 at 8:57 pm
Is this supposed to groundbreaking? Or maybe it makes no presumption of that but anyway, it's kind of already done. When coaches and GMs talk about fit, they're not necessarily talking about how well a player fits a positional style but how his skills fit into the team, be it as a shooter, a defender, a pick and roll point guard and whatnot.
I look at it teamwide. To win a championship, a team needs to be able to do many things: get to the line, shoot threes, score from the post, defend the post, block some shots, have a perimeter player who can create his own shot and so on. You try to find a group of players who can fulfill all those things. It doesn't matter if it's the PF who gets to the line a lot like Nowitzki as opposed to a SF like Carmelo, just that someone gets there.
August 6th, 2010 at 10:20 pm
Why do we even worry about position? Put out units that put you in a better position to win the game. Maybe a particular unit leaves you with some defensive gaps, but if it creates an offensive situation that far outweighs those gaps, go with it. Same with the inverse. If you have 5 LeBrons, you run them all out there. Maybe Dwight Howard puts up a 40/40 line, but if your team scores 150, who cares?
(I do realize this is oversimplifying a bit, but the idea is sound.)
August 7th, 2010 at 4:28 am
A team with 5 LeBrons would be interesting...not good necessarily, but arguably the most fun team to watch short of 5 Magics in his prime, Steve Nash+whomever, or the 1999-2002 Kings.
But BSK makes a good point: Positions, by and large, are meaningless. Their particular roles depend heavily on the offensive scheme being run and the particular players in any given position.
A couple of days ago I saw someone list all the top power forwards in the NBA, and then ask wtf do these guys have in common? The answer is pretty clear: They are all in the NBA. And that was the list.
August 7th, 2010 at 8:59 am
Nick-
Good point. I mean, suppose the Heat run LeBron at the point. Is he really a point guard at that point? Or are they essentially running a 3 F system where 1 F handles most of the distributing? What does it even MEAN to be a PG? We have a typical definition, but we are increasingly seeing guys who transcend traditional positioning. I saw one guy advocate for a simple delineation between wing and post players. You still have some freaks (like LeBron and Durant) who don't fit neatly into either of these, but it seems much more descriptive of how they actually play, as opposed to predetermined "roles" on the court.
August 7th, 2010 at 11:39 am
I read that article a few days ago, and I'm not sure how much it really accomplishes. The traditional positions are still a very useful quick and dirty point of reference.
I'm sure that when most people think about a player, they quickly go beyond positional designations. In fact, I think that the defensive designations proposed by Cannon are a big part of what people already think about. One of the first questions I ask when thinking about a player is whether he can guard his position on the court. Ben Wallace was noteworthy because even though he was undersized, he could guard larger players, which weighs in favor of counting him as a center.
So basically 2 points: (1) the 5 positions are already useful, people already go beyond them, and (2) any alternative scheme has a high risk of being more confusing and less brief.
Additionally, it may also be the case that the 5 positions are so deeply rooted in the teaching of organized basketball that they will never go away or be replaced by something else.
August 7th, 2010 at 4:18 pm
I like that JLK1. When somebody says, "I wanna see LeBron play the 4 more" what they are really say is "post your 265 lbs ass UP, LeBron!" It's more a description of the tasks at hand than anything else.
August 8th, 2010 at 10:58 am
@P Middy
LeBron playing the four also means he is guarding the opposing team's power forward, and it increases the chance that he will be on the floor with three other shooters (for instance, Mo, Delonte, and Anthony Parker last year).
I think a number of NBA teams are adapting this positional flexibility in order to better take advantage of a market that improperly values certain characteristics. Orlando is the most obvious example of this, but Portland also fits the idea to some extent.
August 8th, 2010 at 11:17 am
I think the idea contained within this article can be extrapolated to think about lineups that lack particular traditional skills - for instance, Phoenix and Golden State this season and their lack of a 'Rebounder'.
A statistical system that could predict the potential efficiency of a given lineup (as a function of their respective TS%, USG%, 3pt%, Reb%, Ast%, DefRate, ect) would bring this analysis a step further, by explaining the relative merits of a lineup of all shooters, or with two strong rebounders, or the d'antoni fantasy of a team composed entirely of 6'8" combo-forwards with a variety of different skills.
At the beginning of the 09/10 season, phoenix was roundly criticized for their lack of rebounding. They managed to more than offset this 'deficiency' by playing alternative lineups that were incredibly offensively efficient, making up for the lack of an ability/role deemed 'essential'.
August 8th, 2010 at 2:09 pm
Is it just me or did anyone else read that and think immediately of the Bad Boy Pistons or maybe even the 2004 Pistons?
Seems you can flip this completely around and say really that if you can find D1-D5 and rebound that you can find ways to score enough in a given game.
Thinking about someone like Pippen and Rodman. Between those two guys you could guard any position on the floor with high effectiveness. Rodman could guard 3-5 and Pippen could guard 1-3. Add MJ from those Bulls teams and you have another guy that can guard 1-3.
The ability to defend multiple positions and sizes seems much more pivotal than having a scorer and creator/handler. This is where the Bad Boys and the Bulls were so impressive. Since defense is ultimately a reaction to other teams it can't be controlled with any certainty like offense. If you get guys in the NBA with athleticism, work ethic, and common sense a good coach and shape an offensive system of reasonable effectiveness for them. The key is to have enough variability in defensive talent to guard the other 29 teams of varying sizes and skill sets.
So in building a team I would never recruit for offense personally. I would look for guys that could defend multiple posistions effectively.
In today's NBA you'd think of guys like Noah, Varejao, Battier, Anthony Randolph, LeBron, etc.
In recruiting, I think teams are stuck in the traditional roles too much and really recruit to a system far more than necessary. Why not recruit to talent and build an offensive system around that talent? There are as many offensive systems as there are skill sets.
The Bad Boy Pistons had a creator/handler in Isiah, but no one else that would really fit the "scorer" mold after Dantley left. They had great rebounders and multi-faceted defenders. To me that is more important. Don't recruit to system or to fill offensive roles. Recruit to defensive variability and athleticism and build around that as best you can.
August 8th, 2010 at 9:28 pm
Good idea but mnemonically abstruse.
Why not skip the "D" and simplify even further?
Shorter shorthand.
1,2,3,4,5 = offensive role.
a,b,c,d,e = defensive role.
In Cannon's example, rather than Scheyer the 1/D2 or Smith the 2/D1, Scheyer becomes 1b, Smith 2a.
Likewise Young (2a), Funk (1b), Crouch (3c), Wilson (5d), Calathes (4e).
Same thing, less syllables. less numeric, less confusion.
Even Coach K can be a little too concrete sequential for his own good.
Tomorrow, color-coded skillsets.
August 9th, 2010 at 2:22 pm
Can LeBron James defender centers? Has anyone ever seen this? I remember a couple years ago he mentioned that he outweighed Dwight Howard (not sure if that was true then or if it would still be true now). If he could Oakley-up on big men and leave Bosh free to block shots and crash boards, they could run a small, crazy-fast lineup with a point-center.
August 9th, 2010 at 2:25 pm
He could do it. Though I think it would be a waste of his defensive talents and skills.
August 9th, 2010 at 4:39 pm
The way I tend to consider positions and the value they add to a team is drastically different between offense and defense. Let me explain....
I am sure everyone has heard the phrase, "only a strong as it's weakest link." Well that phrase is true for defense but not necessarily on offense. On defense a team does not (typically) get to decide where the offense gets to attack them. The offense will likely attack the weakest defensive player and either exploit the match-up for an easier score or force the other defenders to help creating other opportunities on the floor. So the defense truly is "only as strong as the weakest link".
On offense though a team is not "only as strong as the weakest link" because the team doesn't have to utilize the weakest link at all. It is their choice whether or not to include the player. The offense can decide to go through the strongest "links" first. It is all about exploiting a strong mismatch.
This gets to the heart of the difference between offensive and defensive team success. If I were to build a super team then I offensively I would want a couple of players who were dominant scorers inside, a couple of great shooters, a couple of guys who can penetrate and create opportunities, and a couple of slashers to finish plays. I don't necessarily need a single guy who can do all of those things (although that would be best), as long as I have a single mismatch somewhere I can exploit it so it really boils down to having differet players who are each bove average at a certain skill. Basically I need a "typical" PG, SG, SF, PF, C.
On defense however I need 5 guys who each have very few weaknesses that can be exploited. I need guys who can all move their feet, have length to disrupt passes and shots, and can rebound. Basically I don't need each of a PG, SG, SF, PF, and C but rather a bunch of guys that can each guard multiple positions (which is especially helpful in pick-and-roll when a team can switch).
So I think on offense, the positions are important with regard to skills (not necessarily size) but on defense it is not important what your typical position is but whether you have the ability to guard multiple positions.
August 9th, 2010 at 10:36 pm
He could probably defend against some of the smaller, less athletic centers, but I wouldn't put him up against, say a Dwight.