Sports Analytics: Business Intelligence using Big Data

This report analyzes the importance of big data and analytics in the sports industry, using IPL matches data from 2007 to 2017. It covers the data collection, storage, application, and processing, as well as the business continuity and recommendations. The study area includes artificial intelligence, NoSQL, and big data technologies. The report concludes that sports analytics is a game-changer for predicting the outcomes of players, games, and businesses.

Business Intelligence using Big Data
“Sports Analytics “

Table of Contents
Objectives……………………………………………………………………………………..1
Executive Summary…………………………………………………………………………..1
Introduction……………………………………………………………………………………2
Data Collection………………………………………………………………………………..2
IPL Batsman……………………………………………………………………………….3
IPL Bowlers………………………………………………………………………………..3
IPL Matches………………………………………………………………………………..4
Head to Head………………………………………………………………………………4
Data Storage…………………………………………………………………………………...4
Data Application and Processing…………………………………………………………….5
IPL Batsman Analysis………………………………………………………………………6
Strike Rate.…………………………………………………………………………………..6
Runs as per Location.………………………………………………………………………7
Analysis of Batsman Performance……………………………………………………….7
Business Continuity..…………………………………………………………………………..8
Recommendation. ……….……………………………………………………………………10
Conclusion………………………………………………………………………………………12
References……………………………………………………………………………………...13

Objectives
The main objective of this study is access and analyzed the difference in precipitation and
the necessity of sports analysis. In this study, we analyze and predicting the importance of
big data in sports industry.
For analysis purpose we will using IPL matches data from 2007 to 2017.
 We analyzing the Individual batsman scores
 Individual bowlers in IPL
 Analysis any IPL match from 2007 to 2017.
 All matches of an IPL team from starting to ending.
 Highest Scorer in IPL
 Highest wicket taker in IPL
Executive Summary
As per the study, Australia is the leader in the field of sports, thats why every people is
involved in sports activities.
This reports provides a more formal understanding of the high performance and sports
workforce with the purpose of informing and policy development of an government bodies
and education providers to aware about sports activities that can enhance the future
leaders in sports.
The data presented in this report is based on IPL matches and it provides the performance
about the players so that we can improve the weakness point.
1

Introduction
We will analysis this records using R-Shiny library. It will predict the data according to
the batsman and bowlers, simultaneously visualize the data in dashboards so that we can
easily analyze the data.
Here we will analyze data as per individual batsman’s and bowlers performance. We will
add all features to the column using rbind function. Count the data as any IPL player could
have played more than one team.
We have all data records for particular team wise with batsman and bowlers performance
in each team. Further will form this data to data frames to analyze the list of records.
Data Collection
For the data analysis we will take some most informative features variables from the
dataset, which is given below :
IPL Batsman : This will having all the records of batsman who is playing in the IPL match.
Also will adding the performance if one player is playing in more then one team in different
IPL session year wise.
 Runs and Bowling’s
 Four & Sixes
 Out of the Batsman
 Runs with strike rate
 Average in an inning
 Average strike rate
 Runs against team wise
2

 Runs at venue wise
 Predicting runs of Batsman for future innings
IPL Bowlers : This will used to analyze individual bowlers in IPL matches. The same as
most bowlers who played more than one game in IPL.
 Economy rate
 Means of the runs by a bowlers
 Moving Average
 Average Wickets
 Wickets plot
 Wickets against another teams
 Wicket counts at the venues
 Bowlers wickets prediction
3

IPL Matches : It will be using for analysis for individual IPL matches at venues.
 Partnerships
 Bowlers vs Batsman
 Wicket type
 Wicket runs
 Wicket in a Match
 Graph for match
Head to Head : This will analysis the data for 2 IPL team. Ex. All matches between
Rajasthan Royal vs Kings Eleven Punjab.
 Team Batsman batting partnership in all matches
 Batsman vs Bowlers in all matches
 Bowlers wickets in all matches
 Runs for bowlers in all matches
 Matches loss or win
Data Storage
Data is collected in different phases and we stored data into cloud storage or in central
warehouse, where we can fetch the data as per requirements.
In data model we are collecting data from difference sources as mentioned and storing in
database management system.
We created three layer as per the data – Structured data, Unstructured data and
Semistructured data. Based on the analysis we can visualize the data in the form of graph,
charts, plot and dashboards.
4

Data Application & Processing
Now in this process we will create an template for end to end analysis of IPL T20 matches.
So before going to analyze the data it requires some dependencies library to run the
analysis model.
 Yorkr package – library(yorkr)
 dplyr package – library(dplyr
In the first step we will we will take care about the dataset. The dataset is requires before
for any analysis for IPL batsman’s, bowler’s and matches.
There is 5 dataset directories :
 database
 IPL matches list
5

 Matches list between two teams
 Matches played against with another team
 Batting and Bowling details in each match
IPL Batsman analysis : As per the data analysis of IPL batsman according to runs vs
bowls, we can see that the graph is continuously increasing with bowls.
Strike Rate : From the graph we can easily see that the batsman strike rate is
continuously increasing and creasing. It is not stable.
6

Runs as per location : As per the location wise we can see the batsman is scoring
different runs. In some IPL matches the score is high and in some its less.
Analysis of Batsman Performance : Different match having better performance and in
some matches having low. From the dataset we can say that the batsman Ashwin having
better performance in first over but letter on its decreasing.
7

So from the analysis of the dataset we get a point is that in big data we can also use
sports analysis to predict the future prospect based on the data and results.
Business continuity
Applied data analysis is the term that is used to refer to the use of quantitative methods to
and technology to gather, store and analyze the multi structures data in order to make
business decisions.
Rapid increasing in data analysis is important for sports management. All the sports
events having data analysis, they are analyzing each and every activities by the players.
Data analytics is not limited in one sport, but it is applied for all kind of sports like Crickets,
Football, Baseball, Athletics, Cycling and many more to improve sports analysis.
Why use Big Data in sports, because it not easy to collect all types of data from sports
activities. Data is generating in sports in unstructured format which is not easy to collect to
by traditional software or systems, thats why big data comes into the pictures to collect
these types of data.
The below applications of Big Data in Sports –
 Games data analysis
8

 Player actions
 Sports broadcasting
 Fitness management
The streaming data comes into big data module kafka and collected all streaming data to
spark streaming system to store into database.
In the first step data are collected in kafka queue which is data warehouse for
organizational analysis purpose.
Kafka process the data to spark engine for filtering the data based on key and values
paired.
Http processed data on user portal portal in to form of dashboard to visualize the user
streaming data to insights form.
This is cricket match streaming process how big data application used in sports as well as
for business purpose. As per the analysis based on previous records, businessman can
interact with the batsman’s or players to take auction and sell themselves.
9

Recommendation
Day by day increasing role of analytics in sports is good for business scope. Peoples can
be get training and make some strategy planning on the sports. Data analytics is the game
changer for sports to analysis the players data and records.
Many sports coaches believe that performance analytics is the most important point to
analyze the weakness of the players. Based on the analysis we find out the most weak
point in performance so that we can improve It.
Analytics for Winning : It is an strategic planning to win the game from opposition. But
before we analyzed the historical data for each activities. Based on the analysis we target
to player or opponent on key point.
High powered big data analytics algorithms and CCTV camera’s replacing human and
statistician helping them to find out the performance of the player and outcomes the game.
Study area
Sports is having huge amount of data which can not analyzed by basic software or
systems. It requires big data processing techniques and algorithms which can give the
result in seconds. Big data algorithm capturing each activities of the players and game and
generating amount of data.
Today’s world is focused on Artificial Intelligence which provides real time analytics and
data visualization quickly. The data scripting data processed in bulk amount and
algorithms find out the outcomes from the data.
Artificial Intelligence : It is the main source of data analytics. AI used deep machine
learning algorithms to find out patterns, sentiment analysis, taxonomy, language
10

processing and many more patterns from the scripting data. Based on the analysis team
coaches can take action and improved the players performance.
Big Data : Big data is the main challenges in today life. Data is collecting from different
sources and processed into an system. Now the challenge is to find out the outcomes from
the data. How to predict the outcomes and if it comes then on which basis we need to
predict the results.
In big data Hadoop system works fine which includes 3V’s to process the data.
Volume : As per current study about 3.5 zeta-bytes data are created each day and the
volume is continuously increasing. The data were generated in past 20 year, now its
generated in a day.
Velocity : Some applications using speed is more important then volume of the data. Data
is generating through different channels in different velocity with speed.
Variety : We posted blogs, images, quotes and messages over the social media. These
data is generating in variety, it can be text files, audio file, video, parquet and blackbox.
NoSQL : The sports analytics is developed using many big data technologies and NoSQL
is one of them to store and processed data in unstructured format. The apache hadoop
framework, mapreduce framework and hive data warehouse is responsible to store,
processed and find the outcomes in sports.
The JSON file format data is captured and stored through NoSQL database for real life
applications. It stores the data in text files format, which having key and value.
All unstructured data is captured and analyze by NoSQL database. We can integrate
NoSQL to hadoop and Big data platform to process real time data.
Mostly Airlines using big data analytics in sports activities so that employees can become
happy. It is an study that good environment gives better results always.
11

Conclusion
Sports are the products human cultures. It provides the distinctive specialty from the sports
(athletics, crickets, football, baseball, racing, cycling and many more) to find out the
relationship with their environments.
The issue is generating day by day due to increasing the social media data and sports
cultures. The analysis of sport and its models are capable to detect weakness point to
survive the sportsman.
Sports analysis is in big boom right now to get the actual prediction and analysis of the
players, game, ground and business. It is not easy to get the real time data analysis on the
sports. It includes many artificial intelligence and machine learning algorithms to find out
the insights.
Business is also an dependencies on sports analysis to sell and buy the products and
services. Based on the real scenarios analysis is predicting the outcomes from the data.
A variety of organizations are attached in collecting data about players or sportsman that
can end up virtually anywhere, through different types of information : bank, markets,
media, government authorities, individuals and employees.
12

References
Haines, M., and M.H. (2013). The role of performance analysis within the coaching
processes. pp.10.1018/24748668.2013.11868645 .
A Lynn, A.L. (2010). Effective Sports Coaching: A Practical Guide, Wiltshire: The Crowood
Press Limited.pp.201-208.
Carling, C.C., Reilly, T.R.and Williams, (2009). Performance assessment for field sports.
Riewald, S., S.R. (2011). Video Analysis in Sports. pp. 14.
Dogramac, S., S.D. , Watsford, M. and Aron, J. (2011). The Reliability and Validity of
subjective National Analysis. pp. 852
O’Donoghue, P., P.O. (2010). Research Methods for Sports Performance Analysis. pp.
1022.
Hughes, M., M.H., T.C. and James, N. (2004). National Analysis of Sports.(2nd ed.) Oxon.
Routledge.
Jaques, T.D., Pavia, and G.R. (1974). An Analysis of movement patterns of players in
Australian Rules league football match. pp.10.
Smith, R. (1980). An Analysis of the running patterns of field umpires in Australian
Football. Sports coach. pp.16.
Reilly, T., and Thomas, V. (1990). A motion analysis of work rate in different positional
roles in professional matches play. pp.87.
Pyne, D.(1995). Fitness testing of AFL Umpires. Australian Institute of Sports, London /
New York. pp. 343.
13