Using Machine Learning
Cricket is a joy-giving sport. Cricket is followed by millions of people around the world, especially in countries where it is played en masse. People are passionate about their teams and players have a huge fan following all over the world.
Given that more than just sporting moments a widely followed cricket game has much more in stake than just a few days or the few hours during which it is played.
We can find that fans frequently discuss cricket matches and possible teams before every match be it a Test Match or a One day international or a T20.
The IPL has thrown open the doors and introduced many good players via their good performances. These players are unproven at the international or national level but may make it count on the international stage.
The IPL has also introduced a lot of good selection problems, namely many more players to choose from.
When a national team is announced it involves only players who have proven themselves at the international level and does not involve players who are in the limelight via franchise cricket.
Also, when the team management announces the team for a match before the match commences, sometimes the selections shock all the fans the world over.
Fans even if they are not players do understand why a specific player should have been included or should have been excluded from the team.
Here is where we must rely on the scientific method to provide us the much-needed result, a selection policy without human errors, bias, or negligence.
Whereas a captain may just be able to judge a player using a few selection criteria he has in his mind, there is no way the captain of a team or the team management will know of the many more factors which can influence the outcome of match vis-à-vis the set of players who may be part of the final squad.
Here a good machine learning algorithm or an AI algorithm/model can use tons of data (Big data) and can be built using many measured or partially measured selection criteria.
These may be the specific performance of a player against an opposition player using historical data, quantitatively measured performance of a player in a venue, or weight/fitness parameters of a player.
A dietary pattern followed by a player, performance during the second innings, first innings, performance during a time of the day, many more technical parameters related to batting and bowling and match performance parameters such as how many times has a player won a match for his or her team, etc., All these parameters can be quantified and a final score can be arrived at.
This balanced scorecard approach when balanced with the objective that maximizes the team’s chance against an opponent at a venue will remove the vagaries and inconsistent selection policies that are sometimes observed by the fans.
This will go in a long way to improve the credibility of International Cricket which may, unfortunately, be marred by partisan or incompetent decision-making by the management.