The uniqueness of the paper lies in the introduction of new variables and refinement to the existing variables that help to be more accurate. The factor analysis approach can also be extended to the other formats of the game. The rapid growth in popularity of some of the sports activities over others can be attributed to large-scale commercialization.
Therefore, a proper ranking system that can reflect a true picture of a player's current form becomes imperative. The classification method introduced here attempts to address some of the issues in the ICC classification system as described above. It is believed to capture some of the important variables affecting batting and bowling performance that have been ignored in both previous papers and the widely accepted ICC ranking system.
The uniqueness of the paper lies in the introduction of new variables and the refinement of existing variables, which help to more accurately measure performance and its impact on ranking.
Variables for ranking batsmen
Comparative Performance – Runs scored by a batsman are often used as the primary variable affecting the performance score of a batsman. Hence, to gauge the impact on performance, here the runs scored by a batsman on a match-by-match basis is divided by the average of runs scored by all the batsmen during the period. Location (ie home/away/neutral) impact – The location of the match (ie home/away/neutral) is generally believed to have an impact on the performance of a batsman.
However, as the statement is debatable and likely to change based on the era in which the matches are played, the runs scored by a batsman are divided by the average of runs scored by all batsmen in that respective location (eg home/away). /neutral) during the period under consideration. This gives a higher weightage to runs scored at a venue (eg home/away/neutral) that is unfavorable to the batsmen. If home location tends to be less favorable for batsmen over time, the model will automatically take care of this and give a higher weight to runs scored at home compared to runs scored at the away venue.
Therefore, a proxy is used here in relation to the runs scored by a batsman divided by the total run rate of the match. Modified run rate is calculated by dividing the runs scored by the batsmen by the number of overs or 50 if the team is all out. As a result, to capture this effect, the runs scored by a batsman are divided by the average of runs scored by all batsmen in the respective innings during the period considered.
In a game where more batters score centuries, the runs scored by a single batter are of lesser value. Therefore, the ratio of a batsman's runs to the runs scored by the batsmen of the opposing team is taken into account here. Therefore, the ratio of a batsman's runs to the runs scored by the batsmen of the same team is also taken into account here.
Variables for ranking bowlers
Economic Bowling - The higher the number of wickets taken by a bowler, the better. However, a bowler's performance should not be judged by the number of wickets taken alone. To exclude the possibility of a bad player who takes more wickets due to poor shot selection by the batsmen or sheer luck being awarded a higher score than a player who has taken less wicket, regardless of the best bowling skills, we must also consider the bowler's economy rate.
A good bowler who bowls consistently will have a lower economy rate compared to someone who is used to being hit around the park on a regular basis. However, we will rely on the economy rate of the bowler because the nature of the dismissals will prove to be very difficult to follow. In the example given here, Kumar's economy rate of 5.33 compared to Aaron's economy rate of 7.42 shows the consistency of bowling good balls.
In both cases, a wicket taken in the 1st innings can be considered more valuable as compared to a wicket taken in the 2nd innings. When the number of wickets is divided by the average wickets of the innings, the resulting ratio will be able to give a weight to the wickets taken in that innings. Innings average wickets are calculated as the average number of wickets taken by bowlers in the 1st or 2nd innings over all matches played during that period.
To capture this on a match-by-match basis and give a better score to a bowler who has a lower economy rate in high-scoring matches, the number of wickets taken plus one, multiplied by the total run-rate and divided by the economy rate of the bowler is considered here. To avoid categorical variables and at the same time to capture the impact, wickets plus one are multiplied by modified net run rate and divided by economy rate to give greater weight to bowlers who have contributed to increasing the margin of victory or reducing the margin of loss. Modified run rate is calculated by dividing the runs scored by the batsmen divided by the number of overs or 50 if the team is all out.
3 Results and discussion
In the event of a tie, Net Run Rate is one of the first criteria taken into account to advance to the next stage of the tournament. This way, after each match, player rankings can be updated based on their performance. The data is not available because the player is ranked under the top 100 bowlers in the 2015 ICC rankings or has retired in the same year.
The claim that ranking variables need to be changed because ICC rankings are subjective and do not reflect actual performance is warranted based on the results obtained from the ranking system used in this article. Among batsmen, for example, Virat Kohli (India), who had an average performance in 2015, remained second in the ICC ranking as of December 31, 2015. However, ICC rankings removed them from the list due to their retirement, which, ideally, should have happened after the end of the calendar year.
JP Faulkner (Australia) and Nasir Hossain (Bangladesh) who had a poor year batting are ranked 29th and 39th respectively in the ICC rankings above some of the better players. This could be partly due to him not being part of the ODI team for much of the year. While ST Finn (England), MM Ali (England), DL Vettori (New Zealand), PJ Cummins (Australia), Mohammed Shami (India) are not at all in the ICC rankings despite having a much better performance compared to Narine .
Similarly, SMSM Senanayake (Sri Lanka) also had a below average performance and yet he is ranked 14th in the ICC rankings. These variables have been identified by us but still ignored in the current model due to several reasons. The next section, which presents the limitations of this model, also discusses the possibilities for further improvement of the same by elaborating more on the variables that have been ignored in the current study.
4 Limitations and scope for future research
The ranking system presented here accurately reflects a player's true performance over a given period of time upon which his ranking for the period is determined. But having said that, there is an opportunity to make this system even more robust by embedding a few more variables that will help capture every possible aspect that affects a player's performance. Building (Breaking) Partnerships / Powerplay Strike Rate (Economy Rate) / Powerplay Rates Scored (Wickets Taken) – Using these three variables as well as all others may be most appropriate.
However, at this stage the only rankings available are the ICC rankings and as this attempt is to address the loopholes therein it wouldn't have made sense to use the same ones for our analysis. These rankings will then be used in achieving rankings in subsequent stages, taking into account additional variables. Opposition based match pressure - Pressure in matches between the top ranked teams (e.g. India v Australia, Australia v South Africa), or between traditional rivals (e.g. India v Pakistan, Australia v England, Australia and New Zealand). ) is much higher than the pressure in other competitions.
Also, in low scoring matches, a bowler who defends a low target is under higher pressure and should therefore be rewarded more for a wicket taken compared to a bowler who takes a wicket in the first innings. Runs scored under physical duress/Ground size/Tournament type (world series/tri-series/bilateral etc.)/Tournament stage (league/knockout)/Batting slot – The use of these variables and all others may prove to be the most suitable method. However, as a batsman not out at the end of the innings is of very little value, it is not considered part of the analysis.
The inclusion of the aforementioned variables will make the model more robust in a way that captures all aspects that affect a player's performance. The model presented in this paper is able to provide a better indication of a player's recent form which is more relevant in the current context. With the growing popularity of cricket, more and more cricket tournaments involving cricketers around the world are becoming common in most countries.
Using this model is highly recommended and especially in the popular era. Data not available as the player is among the top 100 batsmen in the ICC rankings for 2015 or retired in the same year.