March 02, 2015

Basketball Efficiency Rating and Analysis

Basketball Efficiency Rating and Analysis

I'm a keen Basketball player and in the past have coached quite a bit. All coaches want to know how their team is performing, as individuals and as a team, so they can address training in those areas which need attention. There are any number of readily available formulas out there which may be of help, the one I like the most being either a linear weighted....

Player Efficiency Rating ( PER )

or a basketball version of VORP ( Value Over Replacement Player ):-

Here is a simple VORP = (pts+0.75*orb+0.25*drb+0.5*ast+stl+0.33*blk)/(1.5*(fga+(.44*fta)+tov + ((0.75*orb+0.25*drb+0.5*ast+stl+0.33*blk)/2)+(mp/4))).

Most people will be familiar with PER, though it can be more easily dealt with if you take each playing element and give it a weighting, then calculate your total. I do it that way, here is someone who has done it also, though his weightings you may like to change as they appear incorrect ( at least at my examination ):-

The link to explain this methodology

[ FGM x 85.910+ Steals x 53.897+ 3PTM x 51.757+ FTM x 46.845+ Blocks x 39.190+ Offensive_Rebx 39.190+ Assists x 34.677+ Defensive_Reb x 14.707- Foul x 17.174- FT_Miss x 20.091- FG_Miss x 39.190- TO x 53.897 ] x (1 / Minutes).

These calculations are worked from readily available game statistics. But they don't tell the whole story as they tend to be offensively biased and don't accurately measure defensive specialists, who we'll all agree are pretty important. Thinking about this I've come up with my own system and weightings for each factor. I generally work it as a ratio not a raw number, but this is the way I look at the game:-



Offense

Individual

Field Goals
Offensive Rebounds
Open Court Dribbles
Close Court Dribbles
Passes Attempted ( includes weighting for Assists, I value screens on and off the ball more )
Ball Receipt No Shot ( turnovers are accounted for, same for the three above )
Screens and Basket ( worth more than Assists - see Passes Attempted above )
Spatial Awareness and Court Spread
Total Time on Offense

Team

Inbounding and Transition
Consistency vs Predictability
Free Throws
Fast Break & Game Pace
Time Outs
Fouls Drawn

Defense

Individual

Defensive Rebounds
Steals
Screens Fought Over ( less for sliding )
Blocks
Close Guarding Perimeter
Guarding Paint and Helping
Spatial Awareness and Communication
Tactile Interference and Timidity
Total Time on Defense

Team

Zone or Man-to-Man Defense
Fouls ( not necessarily negative see below also, compared to team foul situation )
Number of Players
Time Outs
Turnovers Created by Pressure.
Positive Fouls.

Ok, so I measure each factor using game monitoring statisticians, and multiply the score by each weighting. The whole game is 100%, or in FIBA 4 x 10 minute quarters. So the game is scored out of 100. You need to decide your weighting for each measure and what proportion of the game is offense and defense ( and not 50/50 guys c'mon ).

(ie) Let's say you think offense is 60% of any game and Field goals make up 50% of any offensive effort. Your player shoots 50% from general play, thus; .50 x .50 = .25 x 60 = 15% from shooting alone.

Did your player do this ? In a 100 total point game, did she/he score 15 points ? ( just for arguments sake ).

I can tell you now that offense is not 60% of all game effort and field goals ARE NOT worth 50% of all offensive effort, but hey, I'm using this to illustrate.

Measure and weight each factor, then find your score, in theory your rating should equal actual game score, and to win the game you must be more efficient. How to do this ? To measure you'll need eyes on the game, at least five statisticians following each of your players, or ten if you want to follow the other team as well. You have separate offense and defense sheets, and each statistician follows one player and tallies the qualities and activities you are recording; such as rebounds, baskets, but also ball receptions, dribbles, etc. Your statistical people are following their player all the time whether with or without the ball. You are measuring in-game activity and off-the-ball play is critical to winning basketball.

If we use my earlier example of weighting ( it's nonsense, I'm just using the example to illustrate ), offense was 60% of all game effort and within that shooting was 50% of all offensive effort. That means across the whole game shooting is 30% if the player achieves 100% accuracy ( he never misses ). See ? There are nine individual offensive ratings, along with six team qualities, meaning the remaining 50% of offensive effort will be shared across 14 other results, an average of 3.5714% for each. You can see then that shooting is not the be all and end all, though it is the most important.

Bear in mind that when scoring each area of activity, you need to not just count the thing being done, but compare that thing with something else relevant. So in the simplest terms, Field Goals Made are divided by Field Goals Attempted.

Continuing with our example; on defense there are nine individual and six team measures ( the same number as for offense ). We must now say that 40% of our effort will be defensive ( 60 + 40 = 100 ). If rebounds are reckoned to represent 40% of all defensive effort, then that leaves 4.2857% for each other defensive measure if distributed evenly. If a player plays many minutes perfectly defensively and takes many rebounds without missing many chances, he may score highly, in our case up to 16% across the entire game in rebounds alone. But if you increase the proportion of defensive effort in our formula, then you also increase the defenders score, it comes down to how you see a basketball game, how much is offense and how much defense and what are the relative importance of each ?

Our theoretical player

Based on our 60/40; 15 + 16 = 31% in field goals and rebounds, to get a pass mark, or a winning effort the player has to do a lot else apart from grab boards and score. Oh, and I've grossly inflated this just to make the point, making it even harder to get a pass mark in reality. The only place success comes before team is in the dictionary, as they say.

Philosophy


I coach contact and you can see it in the framework. If we look at assists, I give it a lower rank than say, an on-the-ball screen leading to a basket, or an off-the-ball screen leading to a basket. Take a simple play for example, the point guard has a screen set for her ( on-the-ball ) while another player cuts to receive a pass from the point guard off a screen set for that player ( off-the-ball ). The basket is made and the point guard is credited with an assist, only made possible by the two screens. In simple game statistics only the assist and the basket mean anything. In my world, the hierarchy is; basket, off-the-ball screen followed by basket, on-the-ball screen, pass assist; close court dribble, open court dribble and then farther back whatever started the move whether inbounds or rebound.

I receive criticism - that my formula promotes dirty play. Well, no, I don't accept that but I concede it doesn't automatically penalise for fouls as they play an important part in games, fouls do count and not always negatively. Same with time outs, was the time out reactive or initiating, what about the roster, is the skill evenly distributed and the whole roster can wear the opposition down ? I try to measure effectiveness, and I don't simply deduct all negatives without something balancing it (ie) a team with depth can lose a player to fouling out, while those same fouls may stall the opposition advance, rather like a rear-guard action before the final battle.

Anyway have at it.


No comments:

Post a Comment