FanPost

Team skills algorithm


I love the geeky title.

Rather than do the work I brought home with me, I decided to have some beer and pizza and pretend football season was starting soon. That of course depressed me so I went back to basketball, and I got further depressed after reflecting on our recent draft. So, in an effort to pull myself out of that tailspin of morbidity (I get verbose when buzzing) I figured I'd geek out with an attempt to quantify the NBA.

I've written a bit on my philosophy for building an NBA team, but I never took it to the next logical step: determining the algorithm that allows for the combination of five seperate players to work as one on the court. It would, in essence, allow one to calculate whether the sum is greater than the parts or not. With that, not insignificant bit of data, one could construct an optimization table to find the best result of players available. Granted, this is all speculation, but I'm buzzing and a math geek at heart so it's cool.

Step 1. The system is the key.

Much like the Patriots in football, the system in place is greater than the players. One guy goes down, and another interchangable player steps up to fill the vacancy. One player wants a big payday somewhere else? Fine. Next! To be sure, that's not nearly so easy in the NBA where only five players from the team play at one time. There's less ability to fully redistribute any load to the other players because they are already operating at near maximum capacity already.

In NBA terms, look at the Jerry Sloan teams of the past. They took players and molded them to fit their style of play. They didn't change their system (too much) to accomodate the player. Some may argure that the Jazz were consistently above average, but never great and should thus not be copied. I'd argue that STARTING your system with an above average base is the only way to achieve greatness. Of course if Jerry Sloan isn't your cup 'o tea, maybe Greg Popovich is. They kept the same system using diferent guys and have how many rings? Yeah. Thought so. Consistency of philosophy is where things start.

Step 2. Identify the algorithm.

Here's my part. I'm going to ramble a bit, but here goes. Each player has their own unique series of abilities that can be calculated using various metrics, but it's how those calculations intermingle with the other four players that really matters. It's in this area I feel we have been deprived the really relevant statistical information on players.

2.1 Traits to guage- The Mental

Proprioception (situational awareness, kinesthesia). http://en.wikipedia.org/wiki/Proprioception. Knowing the ability of a player to understand what is currently going on with his body in relation to those around him. Steve Nash has a high level.

Pattern Recognition. Being able to glimpse an image or series of images and know which image is next. For example, being able to spot when a pick & roll play is being set up.

Cause and effect. In tandem with pattern recognition goes the understanding of what that person's duty is for a given situation. For example, knowing when the PF needs to show hard in order to properly disrupt the pick & roll play.

Decision making speed. Being able to combine pattern recognition and cause and effect into a working combination that allows for the player to be in the right place at the right time is crucial. Being half a step late might as well be not being there at all.

Basket Ball IQ. Takes the last three features and combines them into a constructive whole. Seperately they are nice to have, but for true ability they must be combined by each player. Elite players have such a high level of aptitude that they can actually preempt the oposition by knowing what they will do before they know themselves. See Shane Battier.

2.2 Traits to guage- The physical

Twitch reflex. The time it takes for a decision to be made before the body reacts. Knowing the right thing to do and taking action are not the same thing. Feeling a shot go up (proprioception), recognizing it will miss(pattern recognition), knowing it will bounce to the left corner(pattern recognition), knowing to box out you man(BBIQ), and moving to collect the rebound (Twitch reflex). It's that last motion to action that can mean gaining posession of the ball or not.

Burst Speed. Short distance (10-15ft) speed from dead stop. Similar in football to the snap of the ball, it's the initial burst that can make the play. In basketball it can get the loose ball or make that back door cut.

Lateral speed. Moving side to side allows a player to obstruct an opponent's path, forcing them into lower percentage positions. The goal on offense is to move the ball to a position for a high probability of scoring, and lateral quickness on defense can lower an offense's probability of success.

Globe of influence. The globe of influence combines a player's height, reach, burst speed, lateral speed, and jumping ability to understand the space in which and the degree to which he can influence the outcome of a play. For example, visualize a globe with Nic Batum at the center. At the center of the globe is Nic with a bright green aura where he is very effective. As you travel farther from Batum's core, the green light begins to fade to yellow, and finally red (least effective). The shape and volume of the globe would depend on the given player's physical abilities. This is what many analysts refer to as "length".

2.3 Traits to guage-Phsychological

Mental strength. What stresses has the player been through and how have they responded?

Mental flexibility. What major changes have been required and how have they responded?

Core Self. Despite all changes in their life, what has been the constant?

Leadership ability. In situations that have required him to be a leader, how has he responded? In situations without a clear-cut higherarchy, has he ever stepped up to fill the void? How did that work out? What percent of his career has he been shouldered with leadership responsibility?

Teamwork ability. Does the player try to fit into the system in order to maximize everyone's potential, or does he expect others to mold themselves to his superior abilities? Will he defer to an obviously inferior teammate if it was the right play in the situation?

Basis of motivation. What gets the player to stay late after practice? Inferiority complex? Desire to be the best? Love of the game? Nothing better to do?

These represent the most basic traits that my addled brain could come up with.

Step 3. Functions in the algorithm.

Here's where having too many drinks really messes up this post, so I'll request help from the knowledgable BE readers from here.

What traits am I missing? How would you build the mathematical model (nothing too complex, I'm not ready for that)? Do you envision teams ever evaluation players this way (or are they already and I'm not up to speed)?

Go CHIEFS!

Long Live Greg ODEN! Brandon Roy, the new Ray Allen!

Go BLAZERS!