Problem Statement - Indian Premier League (IPL) is the largest domestic cricket tournament in the world. A better pool of team would boost the revenue of any tournament. Analyse the players for making better decisions at the auctions.
A good pool of balanced team will directly impact the revenue and retain the trust of stakeholders. Major sources of revenue are, but not limited to,
- Sponsorship - ~60% with official partner costing 300$ million
- Broadcasting Rights - ~ 2.5$ billions
- Team Sponsors - ~ 400 crore INR
- Ticket Sales - ~ 200 crore INR
- Merchandise - ~2 billion
There are two data sets containing ball-by-ball details and match wise details from IPL 2008 to IPL 2017
Major Highlights:
- Cleansed and preprocessed the dataset using pandas library.
- Using feature engineering created different parameters like strike rate, economy, wickets taken, matches played, team-wise analysis, win and loss trends and their relation with the decision made after winning the toss.
- Using K-means clusters similar players into different classes. This would give us more economic options at the time of auctions.
- For the most effective communication, a Tableau dashboard is created. In this dashboard, players are further sieved through different filters for better player selection.
Conclusion and Future Work - The dashboard and classification helps us understand each player in detail. As future work, I would like to incorporate recent(past 6 months performance in other tournaments) performance of the player which will help us understand the form of the player. A recommendation system which would give suggestion based on the type of player needed in the team.
For most effective communication, a tableau dashboard is created and you can find it here: https://public.tableau.com/profile/kshitij6814#!/vizhome/IPL_Analysis_15776535832250/Batting