NBAplayoff2019
Damian Lim
2019 6 2
2019 NBAplayoff
NBA(National Basketball Association) has gained a lot of popularity with the heated atmosphere of NBA Finals 2019 between the Toronto Raptors and Golden State Warriors. nba.com provides a lot of specific data about teams and players, from basic stats to unconventional hustle stats. Data from https://stats.nba.com/teams/scoring/?sort=W&dir=-1 This is a practice session of dplyr package and I will look at a glimpse of this dataset.
Data Demonstration
I scrolled essential information from the website and saved it into a csv file named “nbaplayoff.csv”.
It’s a data frame with 16 columns(observations) with 21 variables.
Team name / Conference(East/West) / games won / games lost / minutes / 2point attempts% / 3point attempts% / points from 2 point shots out of total points / points from mid range 2 pointers out of total points/ points from 3 point shots out of total points / percentage of fastbreaks / percentage of free throws / percentage of points off turnover / percentage of points in paint / assisted 2pt percentage / unassisted 2pt percentage / assisted 3pt percentage / unassisted 3pt percentage / total assisted point percentage / total unassisted point percentage
If you need more demonstration, you can go to https://stats.nba.com/teams/scoring/?sort=W&dir=-1 and search for “glossary”.
Things to mind
I am using a data updated right after the game 1 of NBA Finals. People who saw after the game 2 will see a different data from mine.
Playoff Data has more variance. (not many samples) I’m aware of this fact and an analysis from the regular season can be different from mine.
1. Data Preparation
nbaplayoff2019=read.csv("nbaplayoff.csv",fileEncoding="UTF-8-BOM")
library(dplyr)##
## Attaching package: 'dplyr'## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, unionlibrary(ggplot2)
nbaplayoff2019=tbl_df(nbaplayoff2019) #represents which class each variable isnbaplayoff2019 %>%
glimpse() #better structured than head()## Observations: 16
## Variables: 21
## $ Team <fct> Toronto Raptors, Golden State Warriors, Milwaukee B...
## $ conference <fct> East, West, East, West, West, East, West, East, Wes...
## $ game <int> 19, 17, 15, 16, 14, 12, 11, 9, 7, 6, 5, 5, 5, 5, 4, 4
## $ win <int> 13, 12, 10, 8, 7, 7, 6, 5, 3, 2, 1, 1, 1, 1, 0, 0
## $ lose <int> 6, 5, 5, 8, 7, 5, 5, 4, 4, 4, 4, 4, 4, 4, 4, 4
## $ min <dbl> 48.5, 48.6, 48.7, 49.6, 49.4, 48.0, 48.5, 48.0, 48....
## $ att2 <dbl> 59.3, 62.9, 56.9, 63.7, 69.2, 66.9, 49.0, 60.8, 76....
## $ att3 <dbl> 40.7, 37.1, 43.1, 36.3, 30.8, 33.1, 51.0, 39.2, 23....
## $ pt2 <dbl> 48.4, 51.8, 47.9, 50.4, 55.6, 55.2, 39.7, 47.7, 63....
## $ midrange2 <dbl> 10.6, 14.0, 6.9, 10.7, 9.3, 11.5, 2.7, 14.4, 18.8, ...
## $ pt3 <dbl> 33.9, 30.8, 34.0, 32.3, 26.3, 26.0, 43.2, 33.3, 19....
## $ fastbreak <dbl> 15.4, 14.1, 18.5, 9.4, 10.1, 10.5, 10.5, 15.3, 9.3,...
## $ freethrow <dbl> 17.6, 17.4, 18.1, 17.3, 18.1, 18.8, 17.1, 19.0, 17....
## $ offto <dbl> 16.9, 16.5, 12.3, 12.4, 10.5, 14.0, 15.9, 12.7, 9.1...
## $ pitp <dbl> 37.9, 37.8, 41.0, 39.7, 46.3, 43.7, 37.0, 33.3, 44....
## $ assist2 <dbl> 50.2, 64.6, 51.9, 43.4, 53.2, 52.5, 41.4, 46.6, 46....
## $ unassist2 <dbl> 49.8, 35.4, 48.1, 56.6, 46.8, 47.5, 58.6, 53.4, 53....
## $ assist3 <dbl> 81.1, 75.0, 85.9, 65.6, 82.1, 92.0, 59.9, 89.2, 89....
## $ unassist3 <dbl> 18.9, 25.0, 14.1, 34.4, 17.9, 8.0, 40.1, 10.8, 10.6...
## $ assist <dbl> 60.0, 67.5, 62.8, 50.1, 60.1, 61.9, 49.1, 60.1, 53....
## $ unassist <dbl> 40.0, 32.5, 37.2, 49.9, 39.9, 38.1, 50.9, 39.9, 46....nbaplayoff2019[,"min"]= NULL #don't need about data of minutes per game. The value will be 48 for teams that didn't go overtime.2. mutate
This is good for adding new columns in a data frame
nbaplayoff2019=nbaplayoff2019 %>% mutate(ratio=win/game*100)
#however, for adding rows, base function rbind is better
str(mtcars)## 'data.frame': 32 obs. of 11 variables:
## $ mpg : num 21 21 22.8 21.4 18.7 18.1 14.3 24.4 22.8 19.2 ...
## $ cyl : num 6 6 4 6 8 6 8 4 4 6 ...
## $ disp: num 160 160 108 258 360 ...
## $ hp : num 110 110 93 110 175 105 245 62 95 123 ...
## $ drat: num 3.9 3.9 3.85 3.08 3.15 2.76 3.21 3.69 3.92 3.92 ...
## $ wt : num 2.62 2.88 2.32 3.21 3.44 ...
## $ qsec: num 16.5 17 18.6 19.4 17 ...
## $ vs : num 0 0 1 1 0 1 0 1 1 1 ...
## $ am : num 1 1 1 0 0 0 0 0 0 0 ...
## $ gear: num 4 4 4 3 3 3 3 4 4 4 ...
## $ carb: num 4 4 1 1 2 1 4 2 2 4 ...head(mtcars)## mpg cyl disp hp drat wt qsec vs am gear carb
## Mazda RX4 21.0 6 160 110 3.90 2.620 16.46 0 1 4 4
## Mazda RX4 Wag 21.0 6 160 110 3.90 2.875 17.02 0 1 4 4
## Datsun 710 22.8 4 108 93 3.85 2.320 18.61 1 1 4 1
## Hornet 4 Drive 21.4 6 258 110 3.08 3.215 19.44 1 0 3 1
## Hornet Sportabout 18.7 8 360 175 3.15 3.440 17.02 0 0 3 2
## Valiant 18.1 6 225 105 2.76 3.460 20.22 1 0 3 1mtcars %>%
rbind(mean=colMeans(mtcars)) %>%
tail()## mpg cyl disp hp drat wt qsec
## Lotus Europa 30.40000 4.0000 95.1000 113.0000 3.770000 1.51300 16.90000
## Ford Pantera L 15.80000 8.0000 351.0000 264.0000 4.220000 3.17000 14.50000
## Ferrari Dino 19.70000 6.0000 145.0000 175.0000 3.620000 2.77000 15.50000
## Maserati Bora 15.00000 8.0000 301.0000 335.0000 3.540000 3.57000 14.60000
## Volvo 142E 21.40000 4.0000 121.0000 109.0000 4.110000 2.78000 18.60000
## mean 20.09062 6.1875 230.7219 146.6875 3.596563 3.21725 17.84875
## vs am gear carb
## Lotus Europa 1.0000 1.00000 5.0000 2.0000
## Ford Pantera L 0.0000 1.00000 5.0000 4.0000
## Ferrari Dino 0.0000 1.00000 5.0000 6.0000
## Maserati Bora 0.0000 1.00000 5.0000 8.0000
## Volvo 142E 1.0000 1.00000 4.0000 2.0000
## mean 0.4375 0.40625 3.6875 2.81253. filter
picking out specific rows or columns that fit certain conditions
nbaplayoff2019 %>%
filter(game>7) #filter teams that advanced to the 2nd round## # A tibble: 8 x 21
## Team conference game win lose att2 att3 pt2 midrange2 pt3
## <fct> <fct> <int> <int> <int> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 Toro~ East 19 13 6 59.3 40.7 48.4 10.6 33.9
## 2 Gold~ West 17 12 5 62.9 37.1 51.8 14 30.8
## 3 Milw~ East 15 10 5 56.9 43.1 47.9 6.9 34
## 4 Port~ West 16 8 8 63.7 36.3 50.4 10.7 32.3
## 5 Denv~ West 14 7 7 69.2 30.8 55.6 9.3 26.3
## 6 Phil~ East 12 7 5 66.9 33.1 55.2 11.5 26
## 7 Hous~ West 11 6 5 49 51 39.7 2.7 43.2
## 8 Bost~ East 9 5 4 60.8 39.2 47.7 14.4 33.3
## # ... with 11 more variables: fastbreak <dbl>, freethrow <dbl>,
## # offto <dbl>, pitp <dbl>, assist2 <dbl>, unassist2 <dbl>,
## # assist3 <dbl>, unassist3 <dbl>, assist <dbl>, unassist <dbl>,
## # ratio <dbl>nbaplayoff2019 %>%
filter(assist<50) #filter teams that a huge proportion of points earned are unassisted## # A tibble: 4 x 21
## Team conference game win lose att2 att3 pt2 midrange2 pt3
## <fct> <fct> <int> <int> <int> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 Hous~ West 11 6 5 49 51 39.7 2.7 43.2
## 2 Utah~ West 5 1 4 58.6 41.4 51.5 3.3 28.8
## 3 Detr~ East 4 0 4 65.6 34.4 54.1 13.3 33.7
## 4 Indi~ East 4 0 4 64.4 35.6 51.2 10.9 32.7
## # ... with 11 more variables: fastbreak <dbl>, freethrow <dbl>,
## # offto <dbl>, pitp <dbl>, assist2 <dbl>, unassist2 <dbl>,
## # assist3 <dbl>, unassist3 <dbl>, assist <dbl>, unassist <dbl>,
## # ratio <dbl>nbaplayoff2019 %>%
filter(assist<50&ratio>50) #unassisted and successful## # A tibble: 1 x 21
## Team conference game win lose att2 att3 pt2 midrange2 pt3
## <fct> <fct> <int> <int> <int> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 Hous~ West 11 6 5 49 51 39.7 2.7 43.2
## # ... with 11 more variables: fastbreak <dbl>, freethrow <dbl>,
## # offto <dbl>, pitp <dbl>, assist2 <dbl>, unassist2 <dbl>,
## # assist3 <dbl>, unassist3 <dbl>, assist <dbl>, unassist <dbl>,
## # ratio <dbl>nbaplayoff2019 %>%
filter(att2>60|freethrow>20)## # A tibble: 11 x 21
## Team conference game win lose att2 att3 pt2 midrange2 pt3
## <fct> <fct> <int> <int> <int> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 Gold~ West 17 12 5 62.9 37.1 51.8 14 30.8
## 2 Port~ West 16 8 8 63.7 36.3 50.4 10.7 32.3
## 3 Denv~ West 14 7 7 69.2 30.8 55.6 9.3 26.3
## 4 Phil~ East 12 7 5 66.9 33.1 55.2 11.5 26
## 5 Bost~ East 9 5 4 60.8 39.2 47.7 14.4 33.3
## 6 San ~ West 7 3 4 76.7 23.3 63.3 18.8 19.5
## 7 LA C~ West 6 2 4 64.5 35.5 49.4 9.6 30.1
## 8 Broo~ East 5 1 4 62.8 37.2 49.2 6.1 29.6
## 9 Okla~ West 5 1 4 63.5 36.5 52.1 11.8 29.7
## 10 Detr~ East 4 0 4 65.6 34.4 54.1 13.3 33.7
## 11 Indi~ East 4 0 4 64.4 35.6 51.2 10.9 32.7
## # ... with 11 more variables: fastbreak <dbl>, freethrow <dbl>,
## # offto <dbl>, pitp <dbl>, assist2 <dbl>, unassist2 <dbl>,
## # assist3 <dbl>, unassist3 <dbl>, assist <dbl>, unassist <dbl>,
## # ratio <dbl>nbaplayoff2019 %>%
filter(rank(-att3)<=6 & rank(-ratio)<=6) #We think rank 1, the best/great/big... To fit this convention, I applied -.## # A tibble: 4 x 21
## Team conference game win lose att2 att3 pt2 midrange2 pt3
## <fct> <fct> <int> <int> <int> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 Toro~ East 19 13 6 59.3 40.7 48.4 10.6 33.9
## 2 Milw~ East 15 10 5 56.9 43.1 47.9 6.9 34
## 3 Hous~ West 11 6 5 49 51 39.7 2.7 43.2
## 4 Bost~ East 9 5 4 60.8 39.2 47.7 14.4 33.3
## # ... with 11 more variables: fastbreak <dbl>, freethrow <dbl>,
## # offto <dbl>, pitp <dbl>, assist2 <dbl>, unassist2 <dbl>,
## # assist3 <dbl>, unassist3 <dbl>, assist <dbl>, unassist <dbl>,
## # ratio <dbl>#filter teams that throws 3point a lot and win 4. select
Selecting column names
select(nbaplayoff2019,contains("2"),contains("3")) #select all columns that has 2 or 3 in column names## # A tibble: 16 x 9
## att2 pt2 midrange2 assist2 unassist2 att3 pt3 assist3 unassist3
## <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 59.3 48.4 10.6 50.2 49.8 40.7 33.9 81.1 18.9
## 2 62.9 51.8 14 64.6 35.4 37.1 30.8 75 25
## 3 56.9 47.9 6.9 51.9 48.1 43.1 34 85.9 14.1
## 4 63.7 50.4 10.7 43.4 56.6 36.3 32.3 65.6 34.4
## 5 69.2 55.6 9.3 53.2 46.8 30.8 26.3 82.1 17.9
## 6 66.9 55.2 11.5 52.5 47.5 33.1 26 92 8
## 7 49 39.7 2.7 41.4 58.6 51 43.2 59.9 40.1
## 8 60.8 47.7 14.4 46.6 53.4 39.2 33.3 89.2 10.8
## 9 76.7 63.3 18.8 46.3 53.7 23.3 19.5 89.4 10.6
## 10 64.5 49.4 9.6 53.5 46.5 35.5 30.1 88.4 11.6
## 11 62.8 49.2 6.1 50.4 49.6 37.2 29.6 49.1 50.9
## 12 63.5 52.1 11.8 47.4 52.6 36.5 29.7 92.3 7.7
## 13 58.1 47.4 10.4 49.5 50.5 41.9 33.9 78.8 21.2
## 14 58.6 51.5 3.3 57.1 10.1 41.4 28.8 21.2 15.6
## 15 65.6 54.1 13.3 50 14.3 34.4 33.7 18.3 13.9
## 16 64.4 51.2 10.9 57.4 13 35.6 32.7 18.7 16.7nbaplayoff2019 %>%
select(starts_with("att")) #select all columns that starts with "att"## # A tibble: 16 x 2
## att2 att3
## <dbl> <dbl>
## 1 59.3 40.7
## 2 62.9 37.1
## 3 56.9 43.1
## 4 63.7 36.3
## 5 69.2 30.8
## 6 66.9 33.1
## 7 49 51
## 8 60.8 39.2
## 9 76.7 23.3
## 10 64.5 35.5
## 11 62.8 37.2
## 12 63.5 36.5
## 13 58.1 41.9
## 14 58.6 41.4
## 15 65.6 34.4
## 16 64.4 35.65. Plotting
nbaplayoff2019%>%
ggplot(aes(x=assist,y=ratio,col=conference,size=ratio)) + geom_point() + geom_text(aes(label=Team),cex=3,hjust=-0.1,vjust=0,col="gray47") + scale_x_continuous(breaks = seq(35,75,5)) + scale_y_continuous(breaks=seq(0,70,10)) + xlim(c(35,75)) #Quite strong relationship between assisted points percentage and win ratio. Correlation Coefficient of 0.71## Scale for 'x' is already present. Adding another scale for 'x', which
## will replace the existing scale.
nbaplayoff2019%>%
ggplot(aes(x=freethrow,y=ratio,col=conference,size=ratio))+geom_point() + geom_text(aes(label=Team),cex=3,hjust=0.35,vjust=-0.65,col="gray47") +xlim(c(11,23))+scale_x_continuous(breaks = seq(12,22,by=2)) #Freethrow is one of the 4 factors which are crucial for a team's win## Scale for 'x' is already present. Adding another scale for 'x', which
## will replace the existing scale.
#Correlation very low (0.22)
nbaplayoff2019%>%
ggplot(aes(x=offto,y=ratio,col=conference,size=ratio)) +geom_point()+ geom_text(aes(label=Team),cex=3,hjust=0.35,vjust=-0.65,col="gray47") #not any correlation between points off turnovers and win ratio