DATA 621 Blog 2: Predicting NBA Player Efficiency Using Defensive Stats
Intro
Defensive Stats
Model

DATA 621 Blog 2: Predicting NBA Player Efficiency Using Defensive Stats

David Quarshie

Intro

In my last blog I took a look at how scoring plays a role in a NBA player’s overall efficiency. We saw that the total number of points a player gets isn’t a major factor in raising their efficiency rate but having high field goal rates will yield a better efficiency score. In other word, taking and making better shots will make you a better player. But what about on the other side of the ball? What can a player do defensively that will them more efficient? In today’s blog we’ll investigate that.

Defensive Stats

Using our dataset from the last blog that filtered our stats to just the 2017 season we can filter some more to get just the PER and the defensive stats we need.

##       PER              ORB.             DRB.             TRB.      
##  Min.   :-35.30   Min.   : 0.000   Min.   :  0.00   Min.   : 0.00  
##  1st Qu.:  9.60   1st Qu.: 1.800   1st Qu.: 10.30   1st Qu.: 6.20  
##  Median : 12.70   Median : 3.300   Median : 13.80   Median : 8.90  
##  Mean   : 12.73   Mean   : 4.912   Mean   : 15.14   Mean   :10.03  
##  3rd Qu.: 15.55   3rd Qu.: 7.550   3rd Qu.: 19.00   3rd Qu.:13.00  
##  Max.   : 31.50   Max.   :26.300   Max.   :100.00   Max.   :56.40  
##       STL.             BLK.             ORB              DRB       
##  Min.   : 0.000   Min.   : 0.000   Min.   :  0.00   Min.   :  0.0  
##  1st Qu.: 1.000   1st Qu.: 0.500   1st Qu.:  8.00   1st Qu.: 36.0  
##  Median : 1.400   Median : 1.200   Median : 25.00   Median :119.0  
##  Mean   : 1.535   Mean   : 1.685   Mean   : 45.51   Mean   :150.9  
##  3rd Qu.: 1.900   3rd Qu.: 2.300   3rd Qu.: 60.50   3rd Qu.:214.0  
##  Max.   :11.100   Max.   :20.200   Max.   :345.00   Max.   :817.0  
##       TRB              STL              BLK        
##  Min.   :   0.0   Min.   :  0.00   Min.   :  0.00  
##  1st Qu.:  45.0   1st Qu.:  8.00   1st Qu.:  3.00  
##  Median : 151.0   Median : 27.00   Median : 11.00  
##  Mean   : 196.4   Mean   : 34.66   Mean   : 21.51  
##  3rd Qu.: 280.5   3rd Qu.: 52.00   3rd Qu.: 29.00  
##  Max.   :1116.0   Max.   :157.00   Max.   :214.00

For our defensive stats we’ll take offensive, defensive, and total rebounding rates, along with their actual numbers. The rebounding rates are estimates of the percentage of available rebounds a player grabbed while he was on the floor, while the actual numbers are exact numbers of rebounds atained. I decided to look at all rebounds as defensive stats since the act of getting a reabound doesn’t help players’ offensively. In addition to rebounding stats, I’ve also added steal and block rates along with thewir actual numbers.

Model

Once again let’s use R’s lm function to create another linear model for the realtionships between PER and defensive stats.

## 
## Call:
## lm(formula = PER ~ ., data = stats_def)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -41.808  -2.495   0.068   2.460  18.020 
## 
## Coefficients: (1 not defined because of singularities)
##              Estimate Std. Error t value Pr(>|t|)    
## (Intercept)  6.507567   0.590925  11.013  < 2e-16 ***
## ORB.        -0.613222   0.384177  -1.596 0.110985    
## DRB.        -1.264917   0.419477  -3.015 0.002677 ** 
## TRB.         2.353034   0.789708   2.980 0.003006 ** 
## STL.         0.416348   0.219731   1.895 0.058610 .  
## BLK.         0.358059   0.138297   2.589 0.009863 ** 
## ORB         -0.032064   0.009144  -3.507 0.000488 ***
## DRB          0.024923   0.004078   6.112  1.8e-09 ***
## TRB                NA         NA      NA       NA    
## STL          0.039041   0.010946   3.567 0.000391 ***
## BLK         -0.005164   0.013753  -0.375 0.707427    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 4.699 on 585 degrees of freedom
## Multiple R-squared:  0.4043, Adjusted R-squared:  0.3952 
## F-statistic: 44.12 on 9 and 585 DF,  p-value: < 2.2e-16

Our results for how defense correlates with efficiency are telling. While rates and percentages were more important in raising efficiency for offensive stats, for defense there are not many stats that raise PER. The p-values for rates like ORB% and DRB%, are high showing little significance. Meanwhile actual stats like ORB, DRB, and STL, are low showing higher significance. But our results also tell us that increasing stats like true rebound percentage and block percentage will raise PER. I think that defensive percentages don’t have greater significance being that they are estimates while their actual values will be more significant but being able to increase those percentages will help you be more efficient since they show that you are helping your team more.