Initial Data Analysis
Eduardo Martinez
5/3/2021
The purpose of this project was to predict which current NBA players will get their jersey numbers retired in the future. Historic NBA player statistics were analyzed and used to forecast future outcomes. Rather than include excessive amount of predictors, the best subset was chosen via statistical methods.
library(Hmisc)
library(tab)
library(knitr)
library(kableExtra)
library(tidyverse)
library(caret)
library(caTools)
select <- dplyr::selectThe Data
There are two separate data sets that are considered throughout the project.
- Team Data: If a player played at least 3 seasons with more than one team, the Career Data is split by each of such teams. For instance, Kareem Abdul-Jabbar played for the LA Lakers and Milwaukee Bucks, so he will occupy two rows in the Team Data.
- Career Data: Statistics of each player’s entire career. Accordingly, each player occupies only one row.
teamAvgs1 <- read_csv("Team_Avgs")
totalAvgs1 <- read_csv("Total_Avgs")
teamHistory <- read_csv("TeamSummaries.csv")outliers <- teamAvgs1 %>%
filter(nSeasons < 3 & Jersey_Retired == 1)teamAvgs <- teamAvgs1 %>%
filter(nSeasons > 2) %>%
mutate(Total_Titles = replace_na(Total_Titles, 0), Team_Titles = replace_na(Team_Titles, 0))
totalAvgs <- totalAvgs1 %>%
filter(Player %in% teamAvgs1$Player) %>%
mutate(Total_Titles = replace_na(Total_Titles, 0))CareerVarsXy <- c("Total_Titles", "AllStar_Selections", "nSeasons", "GP", "GS", "PPG", "TOV", "PER", "USGp", "WS", "BPM", "VORP", "Jersey_Retired")
TeamVarsXy <- c("Team_Titles", "nSeasons", "GP", "GS", "PPG", "TOV", "PER", "USGp", "WS", "BPM", "VORP", "Jersey_Retired")Training and Validation Sets
testSetTotal <- totalAvgs %>%
filter(Seasons %in% grep(pattern = "2017$", x = totalAvgs$Seasons, ignore.case = TRUE, value = T))
testSetTeam <- teamAvgs %>%
filter(Player %in% testSetTotal$Player)nonTestTotal <- totalAvgs %>%
filter(Player %nin% testSetTotal$Player)
nonTestTeam <- teamAvgs %>%
filter(Player %nin% testSetTeam$Player)set.seed(9711)
sampleSplit <- sample.split(Y = nonTestTotal$Jersey_Retired, SplitRatio = 0.7)
trainSetTotal <- subset(x = nonTestTotal, sampleSplit == TRUE)
valSetTotal <- subset(x = nonTestTotal, sampleSplit == FALSE)trainSetTeam <- teamAvgs %>%
filter(Player %in% trainSetTotal$Player)
valSetTeam <- teamAvgs %>%
filter(Player %in% valSetTotal$Player)Best Subset Model (Training Set)
library(leaps)
library(bestglm)
select <- dplyr::selecttrainTotal_Xy <- trainSetTotal %>%
select(all_of(CareerVarsXy))set.seed(11)
trainTotalBestGLM <- bestglm(Xy = as.data.frame(trainTotal_Xy), family = binomial, IC = "BIC", method = "forward")trainTeam_Xy <- trainSetTeam %>%
select(all_of(TeamVarsXy))set.seed(11)
trainTeamBestGLM <- bestglm(Xy = as.data.frame(trainTeam_Xy), family = binomial, IC = "BIC", method = "forward")Performance on Training Sets
trainTotalPreds <- ifelse(trainTotalBestGLM$BestModel$fitted.values > 0.5, 1, 0)
trainTeamPreds <- ifelse(trainTeamBestGLM$BestModel$fitted.values > 0.5, 1, 0)Table1A <- as_tibble(table(Predictions = trainTotalPreds, `True Outcome` = trainSetTotal$Jersey_Retired, deparse.level = 2)) %>%
kbl(booktabs = F, align = "c", linesep = "", midrule = "") %>%
kable_classic("striped") %>%
kable_styling(latex_options = c("striped", "HOLD_position"), bootstrap_options = c("condensed", "bordered"),
full_width = F, font_size = 10) %>% #, position = "float_left") %>%
add_header_above(c("Table 1A: Career Model" = 3), bold = T, font_size = 12, monospace = T, background = "#0047AB", color = "white") %>%
row_spec(0, bold = T, monospace = T, background = "cyan") %>%
row_spec(1:4, bold = T, monospace = T) %>%
# column_spec(c(1,3), border_left = T, border_right = T) %>%
footnote(general = paste("Train Set Accuracy = ", round(mean(trainTotalPreds == trainSetTotal$Jersey_Retired), digits = 4) * 100, "%"), general_title = "")
Table1A|
Table 1A: Career Model
|
||
|---|---|---|
| Predictions | True Outcome | n |
| 0 | 0 | 621 |
| 1 | 0 | 6 |
| 0 | 1 | 22 |
| 1 | 1 | 31 |
| Train Set Accuracy = 95.88 % | ||
Table1B <- as_tibble(table(Predictions = trainTeamPreds, `True Outcome` = trainSetTeam$Jersey_Retired, deparse.level = 2)) %>%
kbl(booktabs = F, align = "c", linesep = "", midrule = "") %>%
kable_classic("striped") %>%
kable_styling(latex_options = c("striped", "HOLD_position"), bootstrap_options = c("condensed", "bordered"),
full_width = F, font_size = 10) %>% #, position = "float_right") %>%
add_header_above(c("Table 2A: Team Model" = 3), bold = T, font_size = 12, monospace = T, background = "#0047AB", color = "white") %>%
row_spec(0, bold = T, monospace = T, background = "cyan") %>%
row_spec(1:4, bold = T, monospace = T) %>%
# column_spec(c(1,3), border_left = T, border_right = T) %>%
footnote(general = paste("Train Set Accuracy = ", round(mean(trainTeamPreds == trainSetTeam$Jersey_Retired), digits = 4) * 100, "%"), general_title = "")
Table1B|
Table 2A: Team Model
|
||
|---|---|---|
| Predictions | True Outcome | n |
| 0 | 0 | 954 |
| 1 | 0 | 9 |
| 0 | 1 | 32 |
| 1 | 1 | 22 |
| Train Set Accuracy = 95.97 % | ||
Performance on Validation Sets
valSaveTotalPreds <- predict(trainTotalBestGLM$BestModel, newdata = valSetTotal, type = 'response')
valTotalPreds <- ifelse(valSaveTotalPreds > 0.5, 1, 0)
valSaveTeamPreds <- predict(trainTeamBestGLM$BestModel, newdata = valSetTeam, type = 'response')
valTeamPreds <- ifelse(valSaveTeamPreds > 0.5, 1, 0)Table2A <- as_tibble(table(Predictions = valTotalPreds, `True Outcome` = valSetTotal$Jersey_Retired, deparse.level = 2)) %>%
kbl(booktabs = F, align = "c", linesep = "", midrule = "") %>%
kable_classic("striped") %>%
kable_styling(latex_options = c("striped", "HOLD_position"), bootstrap_options = c("condensed", "bordered"),
full_width = F, font_size = 10) %>% #, position = "float_left") %>%
add_header_above(c("Table 2A: Career Model" = 3), bold = T, font_size = 12, monospace = T, background = "#0047AB", color = "white") %>%
row_spec(0, bold = T, monospace = T, background = "cyan") %>%
row_spec(1:4, bold = T, monospace = T) %>%
# column_spec(c(1,3), border_left = T, border_right = T) %>%
footnote(general = paste("Validation Set Accuracy = ", round(mean(valTotalPreds == valSetTotal$Jersey_Retired), digits = 4) * 100, "%"), general_title = "")
Table2A|
Table 2A: Career Model
|
||
|---|---|---|
| Predictions | True Outcome | n |
| 0 | 0 | 264 |
| 1 | 0 | 5 |
| 0 | 1 | 13 |
| 1 | 1 | 10 |
| Validation Set Accuracy = 93.84 % | ||
Table2B <- as_tibble(table(Predictions = valTeamPreds, `True Outcome` = valSetTeam$Jersey_Retired, deparse.level = 2)) %>%
kbl(booktabs = F, align = "c", linesep = "", midrule = "") %>%
kable_classic("striped") %>%
kable_styling(latex_options = c("striped", "HOLD_position"), bootstrap_options = c("condensed", "bordered"),
full_width = F, font_size = 10) %>% #, position = "float_right") %>%
add_header_above(c("Table 2B: Team Model" = 3), bold = T, font_size = 12, monospace = T, background = "#0047AB", color = "white") %>%
row_spec(0, bold = T, monospace = T, background = "cyan") %>%
row_spec(1:4, bold = T, monospace = T) %>%
# column_spec(c(1,3), border_left = T, border_right = T) %>%
footnote(general = paste("Validation Set Accuracy = ", round(mean(valTeamPreds == valSetTeam$Jersey_Retired), digits = 4) * 100, "%"), general_title = "")
Table2B|
Table 2B: Team Model
|
||
|---|---|---|
| Predictions | True Outcome | n |
| 0 | 0 | 376 |
| 1 | 0 | 7 |
| 0 | 1 | 14 |
| 1 | 1 | 9 |
| Validation Set Accuracy = 94.83 % | ||
Combined (Training + Validation) & Known Test Set
Known Test Set contains all recently retired players who outcomes are known.
RecentPlayers <- read_csv("CurrentPlayers.csv")KnownCareerTestSet <- testSetTotal %>%
filter(Player %nin% RecentPlayers$Player)
UnknownCareerTestSet <- testSetTotal %>%
filter(Player %in% RecentPlayers$Player)
KnownTestCareerXy <- KnownCareerTestSet %>%
select(all_of(CareerVarsXy))
UnknownTestCareerXy <- UnknownCareerTestSet %>%
select(all_of(CareerVarsXy))KnownTeamTestSet <- testSetTeam %>%
filter(Player %nin% RecentPlayers$Player)
UnknownTeamTestSet <- testSetTeam %>%
filter(Player %in% RecentPlayers$Player)
KnownTestTeamXy <- KnownTeamTestSet %>%
select(all_of(TeamVarsXy))
UnknownTestTeamXy <- UnknownTeamTestSet %>%
select(all_of(TeamVarsXy))CombinedCareerXy <- nonTestTotal %>%
select(all_of(CareerVarsXy))Best Subet (Combined Set)
set.seed(11)
BestOverallCareerGLM <- bestglm(Xy = as.data.frame(CombinedCareerXy), family = binomial, IC = "BIC", method = "forward")CombinedTeamXy <- nonTestTeam %>%
select(all_of(TeamVarsXy))set.seed(11)
BestOverallTeamGLM <- bestglm(Xy = as.data.frame(CombinedTeamXy), family = binomial, IC = "BIC", method = "forward")Performance on Combined Sets
FinalCareerPreds <- ifelse(BestOverallCareerGLM$BestModel$fitted.values > 0.5, 1, 0)
FullTeamPreds <- ifelse(BestOverallTeamGLM$BestModel$fitted.values > 0.5, 1, 0)Table3A <- as_tibble(table(Predictions = FinalCareerPreds, `True Outcome` = CombinedCareerXy$Jersey_Retired, deparse.level = 2)) %>%
kbl(booktabs = F, align = "c", linesep = "", midrule = "") %>%
kable_classic("striped") %>%
kable_styling(latex_options = c("striped", "HOLD_position"), bootstrap_options = c("condensed", "bordered"),
full_width = F, font_size = 10) %>% #, position = "float_left") %>%
add_header_above(c("Table 3A: Career Model" = 3), bold = T, font_size = 12, monospace = T, background = "#0047AB", color = "white") %>%
row_spec(0, bold = T, monospace = T, background = "cyan") %>%
row_spec(1:4, bold = T, monospace = T) %>%
# column_spec(1:3, border_left = T, border_right = T) %>%
footnote(general = paste("Combined Train Set Accuracy = ", round(mean(FinalCareerPreds == CombinedCareerXy$Jersey_Retired), digits = 4) * 100, "%"), general_title = "")
Table3A|
Table 3A: Career Model
|
||
|---|---|---|
| Predictions | True Outcome | n |
| 0 | 0 | 885 |
| 1 | 0 | 11 |
| 0 | 1 | 36 |
| 1 | 1 | 40 |
| Combined Train Set Accuracy = 95.16 % | ||
Table3B <- as_tibble(table(Predictions = FullTeamPreds, `True Outcome` = CombinedTeamXy$Jersey_Retired, deparse.level = 2)) %>%
kbl(booktabs = F, align = "c", linesep = "", midrule = "") %>%
kable_classic("striped") %>%
kable_styling(latex_options = c("striped", "HOLD_position"), bootstrap_options = c("condensed", "bordered"),
full_width = F, font_size = 10) %>% #, position = "float_right") %>%
add_header_above(c("Table 3B: Team Model" = 3), bold = T, font_size = 12, monospace = T, background = "#0047AB", color = "white") %>%
row_spec(0, bold = T, monospace = T, background = "cyan") %>%
row_spec(1:4, bold = T, monospace = T) %>%
# column_spec(1:3, border_left = T, border_right = T) %>%
footnote(general = paste("Combined Train Set Accuracy = ", round(mean(FullTeamPreds == CombinedTeamXy$Jersey_Retired), digits = 4) * 100, "%"), general_title = "")
Table3B|
Table 3B: Team Model
|
||
|---|---|---|
| Predictions | True Outcome | n |
| 0 | 0 | 1334 |
| 1 | 0 | 12 |
| 0 | 1 | 46 |
| 1 | 1 | 31 |
| Combined Train Set Accuracy = 95.92 % | ||
Performance on Known Test Sets
KnownSetSaveCareerPreds <- predict(BestOverallCareerGLM$BestModel, newdata = KnownTestCareerXy, type = 'response')
KnownSetCareerPreds <- ifelse(KnownSetSaveCareerPreds > 0.5, 1, 0)
KnownTestTeamSavePreds <- predict(BestOverallTeamGLM$BestModel, newdata = KnownTestTeamXy, type = 'response')
KnownTestTeamPreds <- ifelse(KnownTestTeamSavePreds > 0.5, 1, 0)Table4A <- as_tibble(table(Predictions = KnownSetCareerPreds, `True Outcome` = KnownTestCareerXy$Jersey_Retired, deparse.level = 2)) %>%
kbl(booktabs = F, align = "c", linesep = "", midrule = "") %>%
kable_classic("striped") %>%
kable_styling(latex_options = c("striped", "HOLD_position"), bootstrap_options = c("condensed", "bordered"),
full_width = F, font_size = 10) %>% #, position = "float_left") %>%
add_header_above(c("Table 4A: Career Model" = 3), bold = T, font_size = 12, monospace = T, background = "#0047AB", color = "white") %>%
row_spec(0, bold = T, monospace = T, background = "cyan") %>%
row_spec(1:4, bold = T, monospace = T) %>%
# column_spec(1:3, border_left = T, border_right = T) %>%
footnote(general = paste("Known Test Set Accuracy = ", round(mean(KnownSetCareerPreds == KnownTestCareerXy$Jersey_Retired), digits = 4) * 100, "%"), general_title = "")
Table4A|
Table 4A: Career Model
|
||
|---|---|---|
| Predictions | True Outcome | n |
| 0 | 0 | 89 |
| 1 | 0 | 1 |
| 0 | 1 | 4 |
| 1 | 1 | 2 |
| Known Test Set Accuracy = 94.79 % | ||
Table4B <- as_tibble(table(Predictions = KnownTestTeamPreds, `True Outcome` = KnownTestTeamXy$Jersey_Retired, deparse.level = 2)) %>%
kbl(booktabs = F, align = "c", linesep = "", midrule = "") %>%
kable_classic("striped") %>%
kable_styling(latex_options = c("striped", "HOLD_position"), bootstrap_options = c("condensed", "bordered"),
full_width = F, font_size = 10) %>% #, position = "float_right") %>%
add_header_above(c("Table 4B: Team Model" = 3), bold = T, font_size = 12, monospace = T, background = "#0047AB", color = "white") %>%
row_spec(0, bold = T, monospace = T, background = "cyan") %>%
row_spec(1:4, bold = T, monospace = T) %>%
# column_spec(1:3, border_left = T, border_right = T) %>%
footnote(general = paste("Known Test Set Accuracy = ", round(mean(KnownTestTeamPreds == KnownTestTeamXy$Jersey_Retired), digits = 4) * 100, "%"), general_title = "")
Table4B|
Table 4B: Team Model
|
||
|---|---|---|
| Predictions | True Outcome | n |
| 0 | 0 | 140 |
| 1 | 0 | 1 |
| 0 | 1 | 4 |
| 1 | 1 | 2 |
| Known Test Set Accuracy = 96.6 % | ||
(Outcomes) Unknown Test Set
Note, the test set contains players who are either currently in the NBA or recently retired.
UnknownTestCareerSavePreds <- predict(BestOverallCareerGLM$BestModel, newdata = UnknownTestCareerXy, type = 'response')
UnknownCareerTestSet$Predictions <- ifelse(UnknownTestCareerSavePreds > 0.5, 1, 0)
# BestOverallCareerGLM$BestModel$coefficients
CareerUnknownPredsDF <- UnknownCareerTestSet %>%
select(Player, AllStar_Selections, WS, Predictions) %>%
mutate(Predictions = recode_factor(Predictions, `0` = "NO", `1` = "YES"))UnknownTestTeamSavePreds <- predict(BestOverallTeamGLM$BestModel, newdata = UnknownTestTeamXy, type = 'response')
UnknownTeamTestSet$Predictions <- ifelse(UnknownTestTeamSavePreds > 0.5, 1, 0)
# BestOverallTeamGLM$BestModel$coefficients
TeamUnknownPredsDF <- UnknownTeamTestSet %>%
select(Player, Team, nSeasons, WS, Predictions) %>%
mutate(Predictions = recode_factor(Predictions, `0` = "NO", `1` = "YES"))Decision Boundary Plots
CareerUnknownPlot <- ggplot(data = CareerUnknownPredsDF, aes(y = WS, x = AllStar_Selections)) +
geom_point(aes(color = as_factor(Predictions)), size = 3) +
ggtitle("Combined Career Model: Unknown Test Set") +
geom_abline(slope = -0.249 / 0.814, intercept = 6.176 / 0.814) +
geom_text(label="Decision Boundary", x=11.8, y=3.9, vjust=-1, angle = -15, color = "darkgreen", size = 4.3) +
ylab("Average Win Shares") + xlab("Number of Times an All-Star") +
labs(color = "Jersey Retired") +
theme_grey() +
theme(plot.title = element_text(size = 15, hjust = 0.5, color = "darkgreen", face = "bold"),
axis.title = element_text(size = 14, face = "italic"),
axis.text = element_text(size = 13),
legend.background = element_rect(colour = "black"),
# legend.key.height = unit(1, "cm"),
legend.title = element_text(size=11, face="bold", hjust = 0.5),
legend.text = element_text(size=10),
legend.position = c(0.18, 0.84))
CareerUnknownPlot
TeamUnknownPlot <- ggplot(data = TeamUnknownPredsDF, aes(y = WS, x = nSeasons)) +
geom_point(aes(color = as_factor(Predictions)), size = 3) +
ggtitle("Combined Team Model: Unknown Test Set") +
geom_abline(slope = -0.457 / 0.672, intercept = 9.219 / 0.672) +
geom_text(label="Decision Boundary", x=12.5, y=5.1, vjust=-1, angle = -22, color = "purple", size = 4.3) +
ylab("Average Win Shares") + xlab("Number of Seasons with Team") +
labs(color = "Jersey Retired") +
theme_grey() +
theme(plot.title = element_text(size = 15, hjust = 0.5, color = "purple", face = "bold"),
axis.title = element_text(size = 14, face = "italic"),
axis.text = element_text(size = 13),
legend.background = element_rect(colour = "black"),
# legend.key.height = unit(1, "cm"),
legend.title = element_text(size=11, face="bold", hjust = 0.5),
legend.text = element_text(size=10),
legend.position = c(0.86, 0.84))
TeamUnknownPlot
# cowplot::plot_grid(CareerUnknownPlot, TeamUnknownPlot)Final Predictions (Jersey Retired = Yes)
FutureYesCareer <- CareerUnknownPredsDF %>%
filter(Predictions == "YES") %>%
rename("All-Star" = AllStar_Selections) %>%
arrange(desc(-6.176 + 0.814 * WS + 0.249 * `All-Star`))
FutureYesTeam <- TeamUnknownPredsDF %>%
filter(Predictions == "YES") %>%
rename("Seasons" = nSeasons) %>%
arrange(desc(-9.219 + 0.457 * Seasons + 0.672 * WS))Table5A <- FutureYesCareer[1:14, 1:3] %>%
kbl(booktabs = F, align = "l", linesep = "", midrule = "") %>%
kable_classic("striped") %>%
kable_styling(latex_options = c("striped", "HOLD_position"), bootstrap_options = c("condensed", "bordered"),
full_width = F, font_size = 10) %>% #, position = "float_left") %>%
add_header_above(c("Career Model: Unknown Test Set" = 3), bold = T, monospace = T, background = "#0047AB", color = "white", font_size = 12) %>%
row_spec(0, bold = T, monospace = T, background = "cyan") %>%
row_spec(1:14, bold = T, monospace = T) #%>%
# footnote(general = c("Remaining players not listed: ", toString(FutureYesCareer$Player[15:17]), toString(FutureYesCareer$Player[18:20]), toString(FutureYesCareer$Player[21:23]), toString(FutureYesCareer$Player[24:25]), toString(FutureYesCareer$Player[26:27]), toString(FutureYesCareer$Player[28:29]), toString(FutureYesCareer$Player[30:31])), general_title = "")
Table5A|
Career Model: Unknown Test Set
|
||
|---|---|---|
| Player | All-Star | WS |
| LeBron James | 17 | 14.7 |
| Chris Paul | 11 | 12.9 |
| Kevin Durant | 11 | 12.0 |
| Dirk Nowitzki | 14 | 10.6 |
| James Harden | 9 | 11.4 |
| Stephen Curry | 7 | 10.5 |
| Dwyane Wade | 13 | 8.4 |
| Anthony Davis | 8 | 9.7 |
| Dwight Howard | 8 | 9.4 |
| Russell Westbrook | 9 | 8.9 |
| Damian Lillard | 6 | 9.1 |
| Blake Griffin | 6 | 8.9 |
| Kawhi Leonard | 5 | 9.2 |
| LaMarcus Aldridge | 7 | 7.9 |
Table5B <- FutureYesTeam[1:4] %>%
kbl(booktabs = F, align = "l", linesep = "", midrule = "") %>%
kable_classic("striped") %>%
kable_styling(latex_options = c("striped", "HOLD_position"), bootstrap_options = c("condensed", "bordered"),
full_width = F, font_size = 10) %>%
row_spec(1:14, bold = T, monospace = T) %>%
add_header_above(c("Team Model: Unknown Test Set" = 4), bold = T, monospace = T, background = "#0047AB", color = "white", font_size = 12) %>%
row_spec(0, bold = T, monospace = T, background = "cyan")
Table5B|
Team Model: Unknown Test Set
|
|||
|---|---|---|---|
| Player | Team | Seasons | WS |
| Dirk Nowitzki | DAL | 19 | 10.6 |
| LeBron James | CLE | 10 | 14.0 |
| LeBron James | MIA | 4 | 16.3 |
| Kevin Durant | OKC | 9 | 12.0 |
| Tony Parker | SAS | 16 | 6.8 |
| Dwyane Wade | MIA | 13 | 8.8 |
| James Harden | HOU | 5 | 14.1 |
| Chris Paul | LAC | 6 | 13.0 |
| Chris Paul | NOP | 6 | 12.8 |
| Dwight Howard | ORL | 8 | 10.9 |
| Stephen Curry | GSW | 8 | 10.5 |
| Russell Westbrook | OKC | 9 | 8.9 |
| LaMarcus Aldridge | POR | 9 | 7.7 |
| Marc Gasol | MEM | 9 | 7.7 |
Final Predictions (Jersey Retired = No)
The tables below only show the players that were closest to the decision boundary. In other words, these players just missed the cutoff point.
FutureNoCareer <- CareerUnknownPredsDF %>%
filter(Predictions == "NO", AllStar_Selections > 1) %>%
rename("All-Star" = AllStar_Selections) %>%
arrange(desc(-6.176 + 0.814 * WS + 0.249 * `All-Star`))
FutureNoTeam <- TeamUnknownPredsDF %>%
filter(Predictions == "NO") %>%
rename("Seasons" = nSeasons) %>%
arrange(desc(-9.219 + 0.457 * Seasons + 0.672 * WS))Table6A <- FutureNoCareer[1:3] %>%
kbl(booktabs = F, align = "l", linesep = "") %>%
kable_classic("striped") %>%
kable_styling(latex_options = c("striped", "HOLD_position"), bootstrap_options = c("condensed", "bordered"),
full_width = F, font_size = 10) %>% #, position = "float_left") %>%
row_spec(0, bold = T, monospace = T, background = "#D73B3E", color = "white") %>%
row_spec(1:13, bold = T, monospace = T) %>%
add_header_above(c("Career Model" = 3), bold = T, monospace = T, background = "#800000", color = "white", font_size = 12) #%>%
# column_spec(c(1,3), border_left = T, border_right = T)
Table6A|
Career Model
|
||
|---|---|---|
| Player | All-Star | WS |
| Draymond Green | 3 | 6.6 |
| John Wall | 5 | 5.8 |
| Isaiah Thomas | 2 | 6.5 |
| DeMar DeRozan | 4 | 5.6 |
| Kemba Walker | 4 | 5.4 |
| Joakim Noah | 2 | 6.0 |
| DeMarcus Cousins | 4 | 4.6 |
| Luol Deng | 2 | 5.2 |
| Rajon Rondo | 4 | 4.4 |
| Bradley Beal | 3 | 4.4 |
| Derrick Rose | 3 | 4.3 |
| Khris Middleton | 2 | 3.6 |
| Victor Oladipo | 2 | 3.4 |
Table6B <- FutureNoTeam[c(1:12,52), 1:4] %>%
kbl(booktabs = F, align = "l", linesep = "", midrule = "") %>%
kable_classic("striped") %>%
kable_styling(latex_options = c("striped", "HOLD_position"), bootstrap_options = c("condensed", "bordered"),
full_width = F, font_size = 10) %>%
row_spec(0, bold = T, monospace = T, background = "#D73B3E", color = "white") %>%
row_spec(1:13, bold = T, monospace = T) %>%
add_header_above(c("Team Model" = 4), bold = T, monospace = T, background = "#800000", color = "white", font_size = 12)
Table6B|
Team Model
|
|||
|---|---|---|---|
| Player | Team | Seasons | WS |
| Blake Griffin | LAC | 7 | 8.9 |
| DeAndre Jordan | LAC | 9 | 7.4 |
| Kawhi Leonard | SAS | 6 | 9.2 |
| Pau Gasol | LAL | 7 | 8.5 |
| Al Horford | ATL | 9 | 7.1 |
| Andre Iguodala | PHI | 8 | 7.7 |
| Udonis Haslem | MIA | 14 | 3.6 |
| Mike Conley | MEM | 10 | 6.3 |
| Anthony Davis | NOP | 5 | 9.7 |
| Luol Deng | CHI | 10 | 5.9 |
| Kyle Lowry | TOR | 5 | 9.2 |
| Joakim Noah | CHI | 9 | 6.4 |
| Giannis Antetokounmpo | MIL | 4 | 6.7 |