Only read this if you have a high IQ

[Guapo]

If this is true.. Basketball has just evolved!


Does Basketball Actually Have 13 Positions?

13 positions in basketball? Muthu Alagappan makes the argument with topology


Back in March, after the Sloan Sports Analytics Conference was held at MIT, I wrote about a paper by Kirk Goldsberry called “Court Vision: New Visual and Spatial Analytics for the NBA.” Goldsberry persuasively argued that NBA teams could use maps to improve, for instance, their scoring efficiency. The maps he produced distilled huge amounts of information into quickly understandable visual guides, and struck me as a notable new tool for savvy NBA teams.


On Monday, Wired published a piece about another paper from the same conference. In “From 5 to 13: Redefining the Positions in Basketball,” Muthu Alagappan, using data from last season, groups players according to the points, rebounds, assists, steals, rebounds, blocks, turnovers, and fouls they accumulated per minute, then creates topological maps that render these groupings visually. Guided by these visual “clusters,” Alagappan concludes that there are 13 positions in basketball, rather than five. His paper won a conference award for “best Evolution of Sport.”


If none of this makes sense yet.. Give the video below a few minutes:



Basketball has 13 positions, not just 5MAY 01 2012

Muthu Alagappan used topological data analysis to group NBA players into thirteen different player types, including Role-Playing Ball-Handler, Paint Protector, All-NBA 1st Team, and One-of-a-Kind.

[choppsboy]

Damn. I should have followed your instructions and not read this.

[Mr.MULLETINO]

I thought I was looking at some new constellations.

[Money4Nothing]

I've always believed that the definitions of player positions are over-simplified.

The only reason we have 5 positions, is because only 5 players can be on the court at a time. But there are many types of players, and choosing which types to play when is the trick.

We're all familiar with the Tim Duncan predicament, whether we classify him as a power forward or center. Obviously, it doesn't matter how we classify him in terms of the 5 positions. He's a unique type of player.

The two things that the author of this study probably failed to understand:

1. It doesn't so much matter which of the 13 positions a player falls into, as how effective he is at that position. For example, Jason Terry and Tony Parker might be the same type of player, but Parker is far better at it than Terry is.

2. The study fails to take into account matchups. If a "scoring rebounder" is matched up with a "paint protector", who helps their team succeed more? You might get the perfect balance of player positions on your team, but if you don't match them up correctly against the players on the other team, you won't get very far.

Again, if you are measuring the value of a player, and some of those bench players seem to have higher point totals per minute than Tim Duncan, you need to remember that they are usually matched up against other bench players, and rarely match up against the players that they share a node with.

However, I do like this positional analysis a lot, and I think it could be very useful for teams to balance lineups and substitutions and maybe even find some of those "diamonds in the rough".

$

[nbahoops21]

[WILLTHETHRILL]

Is Bonner one of the paint protectors??? lol

[gwilly101]

Defensive Ballhandler? Now that's an oxymoron!

[Jose_TheGenius]

"one of a kind"

[JuanCaca]

nice neurona picture!

:P

[KVas]

Defensively I don't know how accurate it is when you consider that steals and blocks doesn't make a great defensive player. I would like to see it done with defensive efficiency stats or PPP allowed for different plays. Though I can't imagine how difficult that would be to conjure up those kind of categories into a cluster. I assume it would create more defensive positions though.

[Money4Nothing]

Originally Posted by KVas

Defensively I don't know how accurate it is when you consider that steals and blocks doesn't make a great defensive player. I would like to see it done with defensive efficiency stats or PPP allowed for different plays. Though I can't imagine how difficult that would be to conjure up those kind of categories into a cluster. I assume it would create more defensive positions though.

Right,

this math is a good start, but they would have to use more effective metrics to really get a sense of how good players are to get any useful predictive value out of it.

However, I do agree with the principle, that the "5 positions" are really oversimplistic and not really relevant anymore.

$

[PHANTOM21]

So when u say high IQ what u really mean is geeky dweeb of the highest order?

[Uwe Blab]

[KAD]

This is a great study and well presented. I only wish he had prescribed an ideal team positions? One of each (13) and 2 backups at key slots for the 15 that the NBA allows on a roster.

[spursr#1cl]

I won't question their math skills and ability to create formulas and algorithms there at MIT. however there is no statistic for intangibles, such as hustle play, and pure will to win.. Ahem.. IE Micheal Jordan playing through the flu or the amount of times a player saves the ball from changing possession.

I can't really believe this would work until I saw it done in the NBA. Maybe Micheal Jordan can try this with the bobcats. Couldn't hurt

[Guapo]

It almost feels as if the Spurs had this formulation or some sort of similar formula in play already. It is a game changer and every GM in basketball should be trying to license it!

[clovisnmspurfan]

Huh??

[zipper-boy]

This guy has been channeling Pop!

[SilverSpur]

This will change the way GMs draft and build teams through trades. No doubt Cuban is already pushing this in Dallas.