607_Jeyaraman_Ramalingam_Final_Project

Cricket Players Analysis

With the growth in Data Science in the field of Sports, It is essential to study and understand the historical data of Cricket Players. This Analysis helped us to erase some of the myths in the Cricket World. Following are several Insights analyzed and Visualized.

1. Grounds which are Paradise for the batsman over the Years Tests and ODIs. Batsmen were able to score easily in these grounds.
2. Batting Performance Analysis , Further breakdown into below categories. This analysis helps us to understand the difficulty of batting over the years
   • Batting Average
   • Centuries Scored
3. Bowling Performance Analysis , To Understand the variance of bowling performance over years this analysis is performed. Number of Wickets taken is considered as parameter to evaluate the performance of bowlers historically.
4. Dominating Players in International Cricket - Players who ranked higher by ICC over the years. This snapshot provides the list of players who ranked high every year.
5. Sachin The World Conqueror - Sachin Tendulkar dominated the world by scoring runs across countries. This snapshot helps us understand the places he scored most and least.

Data Sources

Below are the data sources used for this Project.

Source Data - Data is primarily sourced from CSV file downloaded from Kaggle. It contains Player’s every Innings played in Test and ODI Cricket games.

Reference Data - Rating Information sourced from Wikipedia and Latitude/Longitude Information from Google. CSV files are loaded to MOngoDB and housed in the MongoDB collections. This data is combined with the Source data for analysis and reporting.

Approach

Traditional approach of Data Science workflow such as Data Ingestion, Data Transformation, Data Analysis and Data Visualization are performed.

Libraries

Below are the libraries referenced.

viridis leaflet dplyr plotly hrbrthemes tidyverse dplyr mongolite data.table ggplot2 ggmap RColorBrewer knitr ggplot2

## Loading required package: viridisLite

## 
## Attaching package: 'dplyr'

## The following objects are masked from 'package:stats':
## 
##     filter, lag

## The following objects are masked from 'package:base':
## 
##     intersect, setdiff, setequal, union

## Loading required package: ggplot2

## 
## Attaching package: 'plotly'

## The following object is masked from 'package:ggplot2':
## 
##     last_plot

## The following object is masked from 'package:stats':
## 
##     filter

## The following object is masked from 'package:graphics':
## 
##     layout

## NOTE: Either Arial Narrow or Roboto Condensed fonts are required to use these themes.

##       Please use hrbrthemes::import_roboto_condensed() to install Roboto Condensed and

##       if Arial Narrow is not on your system, please see http://bit.ly/arialnarrow

## -- Attaching packages ----------------------------------------------------------------------------------- tidyverse 1.2.1 --

## v tibble  2.1.3     v purrr   0.3.2
## v tidyr   0.8.3     v stringr 1.4.0
## v readr   1.3.1     v forcats 0.4.0

## -- Conflicts -------------------------------------------------------------------------------------- tidyverse_conflicts() --
## x plotly::filter() masks dplyr::filter(), stats::filter()
## x dplyr::lag()     masks stats::lag()

## 
## Attaching package: 'data.table'

## The following object is masked from 'package:purrr':
## 
##     transpose

## The following objects are masked from 'package:dplyr':
## 
##     between, first, last

## Google's Terms of Service: https://cloud.google.com/maps-platform/terms/.

## Please cite ggmap if you use it! See citation("ggmap") for details.

## 
## Attaching package: 'ggmap'

## The following object is masked from 'package:plotly':
## 
##     wind

Data Ingestion

#Main Inputs

test_input <- data.table(read.csv("https://raw.githubusercontent.com/jey1987/DATA607/master/Final%20Project/Test%20Player%20Innings%20Stats%20-%20All%20Teams.csv",sep = ",", stringsAsFactors = F))
odi_input <- data.table(read.csv("https://raw.githubusercontent.com/jey1987/DATA607/master/Final%20Project/ODI%20Player%20Innings%20Stats%20-%20All%20Teams.csv",sep = ",", stringsAsFactors = F))
bat_Ratings <- data.table(read.csv("https://raw.githubusercontent.com/jey1987/DATA607/master/Final%20Project/batsmen_ratings_all091217.csv",sep = ",", stringsAsFactors = F))
bowl_Ratings <- data.table(read.csv("https://raw.githubusercontent.com/jey1987/DATA607/master/Final%20Project/bowler_ratings_all091217.csv",sep = ",", stringsAsFactors = F))
location <- data.table(read.csv("https://raw.githubusercontent.com/jey1987/DATA607/master/Final%20Project/latitude.csv",sep = ",", stringsAsFactors = F))

#Reference Data

mongodb_bat <- mongo(collection = "BatsmanRating", db = "Rating", url = "mongodb://localhost",verbose = FALSE, options = ssl_options())
mongodb_bowl <- mongo(collection = "BowlerRating", db = "Rating", url = "mongodb://localhost",verbose = FALSE, options = ssl_options())
mongodb_loc <- mongo(collection = "locationCoord", db = "Country", url = "mongodb://localhost",verbose = FALSE, options = ssl_options())

if(mongodb_bat$count() > 0) mongodb_bat$drop()
if(mongodb_bowl$count() > 0) mongodb_bowl$drop()
if(mongodb_loc$count() > 0) mongodb_loc$drop()

mongodb_bat$insert(bat_Ratings,stop_on_error = TRUE)

## List of 5
##  $ nInserted  : num 189233
##  $ nMatched   : num 0
##  $ nRemoved   : num 0
##  $ nUpserted  : num 0
##  $ writeErrors: list()

mongodb_bowl$insert(bowl_Ratings,stop_on_error = TRUE)

## List of 5
##  $ nInserted  : num 189233
##  $ nMatched   : num 0
##  $ nRemoved   : num 0
##  $ nUpserted  : num 0
##  $ writeErrors: list()

mongodb_loc$insert(location,stop_on_error = TRUE)

## List of 5
##  $ nInserted  : num 245
##  $ nMatched   : num 0
##  $ nRemoved   : num 0
##  $ nUpserted  : num 0
##  $ writeErrors: list()

bat_ratings_ds <- mongodb_bat$find()
bowl_ratings_ds <- mongodb_bowl$find()
location_ds <- mongodb_loc$find()

bat_ratings_ds <- separate(bat_ratings_ds,rate_date,into=c("rate_year","rate_mon","rate_day"),sep="-")
bowl_ratings_ds <- separate(bowl_ratings_ds,rate_date,into=c("rate_year","rate_mon","rate_day"),sep="-")

names(bat_ratings_ds)[1] <- paste("Innings.Player")
names(bowl_ratings_ds)[1] <- paste("Innings.Player")

Data Transformation

#Tidy Source Data

# Rename Columns
names(test_input)[1] <- paste("Innings.Player")
names(odi_input)[1] <- paste("Innings.Player")

test_input_subset <- test_input %>%
  select(Innings.Player,Innings.Runs.Scored.Num,Innings.Date,Ground,Innings.Number,X100.s,X50.s,Innings.Minutes.Batted,Innings.Wickets.Taken,Opposition,Country,Innings.Overs.Bowled) %>%
  mutate(Cricket_type = "Test") %>%
  mutate_if(is.character, str_trim) 

test_input_subset[is.na(test_input_subset)] = 0

test_input_subset$Innings.Runs.Scored.Num <- suppressWarnings(as.numeric(test_input_subset$Innings.Runs.Scored.Num))
test_input_subset$Innings.Minutes.Batted <- suppressWarnings(as.numeric(test_input_subset$Innings.Minutes.Batted))

odi_input_subset <- odi_input %>%
  select(Innings.Player,Innings.Runs.Scored.Num,Innings.Date,Ground,Innings.Number,X100.s,X50.s,Innings.Minutes.Batted,Innings.Wickets.Taken,Opposition,Country,Innings.Overs.Bowled) %>%
  mutate(Cricket_type = "ODI") %>%
  mutate_if(is.character, str_trim) 

odi_input_subset[is.na(odi_input_subset)] = 0

odi_input_subset$Innings.Runs.Scored.Num <- suppressWarnings(as.numeric(odi_input_subset$Innings.Runs.Scored.Num,0))

odi_input_subset$Innings.Minutes.Batted <- suppressWarnings(as.numeric(odi_input_subset$Innings.Minutes.Batted))

names(test_input_subset)[5] <- paste("Centuries")
names(test_input_subset)[6] <- paste("Half Centuries")
names(odi_input_subset)[5] <- paste("Centuries")
names(odi_input_subset)[6] <- paste("Half Centuries")

rm(test_input)
rm(odi_input)

#Calculate Match Count - Test

test_input_subset_mtch_Cnt <- test_input_subset %>%
  select(Innings.Player,Innings.Runs.Scored.Num,Innings.Minutes.Batted) %>%
  filter(Innings.Minutes.Batted != "-" & Innings.Minutes.Batted !=0 ) %>%
  group_by(Innings.Player) %>%
  dplyr::summarise(mtch_cnt = n()) 

#Calculate Total Runs - Test

test_input_subset_total_runs <- test_input_subset %>%
  select(Innings.Player,Innings.Runs.Scored.Num) %>%
  group_by(Innings.Player) %>%
  dplyr::summarise(total_runs = sum(Innings.Runs.Scored.Num,na.rm=TRUE)) 

#Calculate Match Count - ODI

odi_input_subset_mtch_Cnt <- odi_input_subset %>%
  select(Innings.Player,Innings.Runs.Scored.Num,Innings.Minutes.Batted) %>%
  filter(Innings.Minutes.Batted != "-" & Innings.Minutes.Batted !=0 ) %>%
  group_by(Innings.Player) %>%
  dplyr::summarise(mtch_cnt = n()) 

#Calculate Total Runs - ODI

odi_input_subset_total_runs <- odi_input_subset %>%
  select(Innings.Player,Innings.Runs.Scored.Num) %>%
  group_by(Innings.Player) %>%
  dplyr::summarise(total_runs = sum(Innings.Runs.Scored.Num,na.rm=TRUE)) 

test_input_subset <- inner_join(test_input_subset, test_input_subset_mtch_Cnt, by = "Innings.Player")
test_input_subset <- inner_join(test_input_subset, test_input_subset_total_runs, by = "Innings.Player")

odi_input_subset <- inner_join(odi_input_subset, odi_input_subset_mtch_Cnt, by = "Innings.Player")
odi_input_subset <- inner_join(odi_input_subset, odi_input_subset_total_runs, by = "Innings.Player")

rm(test_input_subset_mtch_Cnt)
rm(test_input_subset_total_runs)
rm(odi_input_subset_mtch_Cnt)
rm(odi_input_subset_total_runs)

#Calculate Batting Average

test_input_subset <- test_input_subset %>%
  mutate(batting_average = total_runs/(mtch_cnt) )

odi_input_subset <- odi_input_subset %>%
  mutate(batting_average = total_runs/(mtch_cnt) )

test_odi_subset <- rbind(test_input_subset,odi_input_subset)
test_odi_subset <- separate(test_odi_subset,Innings.Date,into=c("year","mon","day"),sep="/")

rm(test_input_subset)
rm(odi_input_subset)


#Tidy Reference Data

bat_ratings_rank <- bat_ratings_ds %>%
  select(Innings.Player,rank,rate_year,rate_mon) 

bat_ratings_rank <- bat_ratings_rank %>%
  select(Innings.Player,rank,rate_year,rate_mon)  %>%
  filter(rank == 1)

bowl_ratings_rank <- bowl_ratings_ds %>%
  select(Innings.Player,rank,rate_year,rate_mon) 

bowl_ratings_rank <- bowl_ratings_rank %>%
  select(Innings.Player,rank,rate_year,rate_mon)  %>%
  filter(rank == 1)

rm(bat_ratings_ds)
rm(bowl_ratings_ds)

player_ratings <- rbind(bat_ratings_rank,bowl_ratings_rank)
rm(bat_ratings_rank)
rm(bowl_ratings_rank)

player_ratings$corrected_name <- str_extract(player_ratings$Innings.Player, " [:alpha:]{2,}")
test_odi_subset$corrected_name <- str_extract(test_odi_subset$Innings.Player, " [:alpha:]{2,}")
bat_ratings_rank_player <- inner_join(player_ratings,test_odi_subset,by="corrected_name")
test_odi_subset$Opposition_corr <- str_replace_all(test_odi_subset$Opposition,"[v]","")
rm(player_ratings)

Exploratory Analysis

Most Runs Scored in Grounds Test and ODI

#Most Runs in Grounds

ggplot(most_runs_ground_test, aes(x=Ground, y=total_Runs_gr)) +
geom_bar(stat="identity",fill=alpha("Red", 0.7) ) + coord_polar(theta = "x") +  theme_minimal() + theme(axis.line = element_blank(),
axis.ticks = element_blank(),plot.margin = unit(rep(-1,4), "cm") ,
plot.title = element_text(hjust = 0.5, color = "#6632666")) + theme_light() +   xlab("Ground") +
  ylab("Total Runs")

ggplot(most_runs_ground_odi, aes(x=Ground, y=total_Runs_gr)) +
geom_bar(stat="identity",fill=alpha("Green", 0.7) ) + coord_polar(theta = "x") +  theme_minimal() + theme(axis.line = element_blank(),
axis.ticks = element_blank(),plot.margin = unit(rep(-1,4), "cm") ,
plot.title = element_text(hjust = 0.5, color = "#6632666")) + theme_light()+   xlab("Ground") +
  ylab("Total Runs")

Batting Performance of Players

Batting Average

p<-ggplot(bat_perf_bat_avg, aes(x=as.Date(paste0(bat_perf_bat_avg$year,"/01","/01")), y=avg_batting)) +
  geom_segment( aes(x=as.Date(paste0(bat_perf_bat_avg$year,"/01","/01")), xend=as.Date(paste0(bat_perf_bat_avg$year,"/01","/01")), y=0, yend=avg_batting), color="grey") +
  geom_point( color="blue", size=4) +
theme_bw() + theme(axis.line = element_blank(),
axis.ticks = element_blank(),plot.margin = unit(rep(-1,4), "cm") ,
plot.title = element_text(hjust = 0.5, color = "#6632666")) +  xlab("Year") +
  ylab("Batting Average")

ggplotly(p)

Number of Centuries

## NULL

Bowling Performance of Players

p<-ggplot(bowl_perf_wickets, aes(x=as.Date(paste0(bowl_perf_wickets$year,"/01","/01")), y=total_wickets)) +
  geom_area(fill="#FFFFFF", alpha=0.5) +theme(axis.line = element_blank(),
axis.ticks = element_blank(),plot.margin = unit(rep(-1,4), "cm") ,
plot.title = element_text(hjust = 0.5, color = "#6632666")) + 
  geom_line(color="#04F339") + ylim(1000,4000) + xlab("Year")
  ylab("Minutes Batted") + theme_bw()

## NULL

ggplotly(p)

# Player Rank Snapshot

p <- ggplot(bat_ratings_rank_player_domin, aes(x=rate_year, y=Innings.Player.x,size=10,color="#F004F3")) +
  geom_point(alpha=0.7) +
  scale_size(range = c(1.4, 20), name="Player with Most ratings") +
  theme(legend.position="none") + theme_bw()+   xlab("Year") +
  ylab("Player Name")

pp <- ggplotly(p)
pp

Player Runs across World

wmap <- borders("world", colour = "black", fill = "gray")
m<-ggplot() + wmap + 
  geom_point(data = sachin_scores_loc, aes(x = as.numeric(longitude),y = as.numeric(latitude), colour = total_runs,size=10000))   + ggtitle("Sachin's Runs Across World") +
  xlab("Longitude") + ylab("Latitude") + scale_size(range = c(0, 25)) +
  theme(plot.title = element_text(hjust = 0.5)) +theme_minimal()

myPalette <- colorRampPalette(rev(brewer.pal(11, "Spectral")))
sc1 <- scale_color_gradient2( low="blue", mid="green",
                              high="red", space ="Lab" )
m<-m+sc1
m

Challenges

1. The Player Names across the Source and Reference data was not same, Some tidying was required to get the clean names out of both Data Source.

2. Joining Player Rating with the actual Players dataframe

Conclusion

1. Lords Ground in London holds the record for most runs in Test and Sharjah in ODI
2. Batting Average declines during 2000 but increasing in the late 2010s.
3. Number of Centuries has increased during 2000s but declining in the late 2010s.
4. The above two factors concludes that eventhough individual scores are more the consistency was lacking overall contributing to low average with more centuries.
5. Bowlers recorded high performance during 2000s matching with the low batting average by taking more wickets.
6. Snapshot of Ranking depicts that Only two players has retained the ranking in the subsequent year.
7. Looking at the World Map it is obvious that Sachin has scored in all parts of Cricket playing countries leaving no place behind.

References

1. https://www.kaggle.com/cclayford/cricinfo-statsguru-data
2. https://en.wikipedia.org/wiki/ICC_Player_Rankings
3. https://developers.google.com/public-data/docs/canonical/countries_csv
4. https://www.kaggle.com/runbmc/icc-test-player-rankings