The Casebook of NO-BALLs in Cricket — Data Science (PART 2)
In Part 1 of this article series, we started by asking a few questions, and by the end, we had answers to some of the questions, but some questions remained. So In this article, we start from where we left analysing the effect of No-Balls on the winning percentage of the teams over the years, Analysis of No-Balls in various leagues and Player analysis for the teams.
RECAP
Number of No-Balls in International Cricket from 2010 to 2021
Initially, we only suspected that there had been an increase in the number of No-Balls, but the data confirmed our hypothesis. There was a 49% increase in No-Balls from 2019 to 2020 and a 66% increase from 2020 to 2021. These numbers are less shocking when compared to the 147% increase from 2019 to 2021. It was also interesting to find the surprising number of No-Balls per match in 2010, and maybe this number would have been more in 2009, but that data is not available.
Are All International Formats Equally Affected?
We found out that the shortest format of the game, i.e. T20 Internationals, was the least affected with only a 9% increase in the number of No-Balls from 2019(0.79) to 2021(0.86). This number seems more interesting when comparing it with other formats, namely Test Cricket and ODI. There was a 120% and131% increase in No-Balls for ODI and Test Internationals, respectively. This behaviour could be attributed to the shorter format and lower score, which is affected by the no balls and free hits.
Performance of Teams based on Number of No-Balls over the Decade
The number of No-Balls per match for a team is the division of the Number of No-Balls and the Number of games played by the team from 2010 to 2021.
England has the least number closely followed by New Zealand. At the same time, West Indies has the highest value with a significant difference from the second last team, i.e. Pakistan.
The charts of all teams look different, but a common feature could be associated with many teams. Almost all graphs have a year around the middle of the decade, which has an exceptionally high value from its neighbouring years. For the five years from 2013 to 2017, 3 teams had increased value in 2015, and 3 teams had increased value in 2016.
END OF RECAP
Finding the least and most affected teams
Most of the teams were affected by this phenomenon, but some more than others. To compare the groups, we find the “PERCENTAGE_INCREASE_2019_max(2020,2021)”, the percentage increase between the value of 2019 and the maximum of two values of Number of No-Balls per match of the year 2020 and 2021.
The results are pretty shocking, and more surprising is the disparity between the least and the most affected team. The least affected team England saw a meagre increase of only 18% (from 1.18 to 1.40), whereas India, the most affected team, saw a whopping 823% (from 0.48 to 4.40) increase. This unexpected significant difference is also because, in 2019, India’s value was half that of England, and in 2021 India’s value is three times that of England.
Effect of Number of No-Balls Per Match on Winning Percentage
Pearson correlation coefficient can be a starting indicator, and if it shows potential, further research like Causal analysis would be worth a shot. A look at the winning percentage of teams tells about their general performance. Take a closer look at the charts of Sri Lanka and West Indies and comment on what did you find out.
The value of Pearson correlation will always remain between -1 and 1. Values near 0, either positive or negative, indicate that both the variables don’t affect the other.
A negative correlation or inverse correlation is such that values of the two variables generally move in opposite directions from one another like one increases and the other decreases and vice versa.
A positive correlation is a relationship between two variables in which both variables move in tandem — that is, in the same direction. A positive correlation exists when one variable decreases as the other variable decreases or one variable increases while the further increases.
The values of Pearson Correlation are relatively small for England, New Zealand, Pakistan and South Africa and higher for countries Australia, India, Sri Lanka and West Indies. One could deep dive further for more analysis, but there is a chance that it might be a coincidence too.
Comparing IPL, BBL, CPL and PSL
A first look tells that PSL has been consistent over the years, and CPL has been the most erratic. We could analyse each league, but it’s time we explore the players. We are starting with the players in the International matches, namely ODI, T20 and Test Matches. Since we can’t just compare the numbers, we will use a KPI called Number of No-Balls Per Balls bowled or Number of No-Balls Per Match, which are similar to the KPI we have used to compare different teams.
The number of No-Balls Per Balls bowled is defined as the Number of No-Balls bowled divided by the number of Balls bowled by the Bowler/Player.
The number of No-Balls Per Match is defined as the Number of No-Balls bowled divided by the number of Matches played by the Bowler/Player.
Maximum number of No-Balls Per Balls Bowled
Simply comparing the values could be misleading because not everybody has played the same number of matches. So first, we divide all 824 players into four categories ‘Q1’, ‘Q2’, ‘Q3’, ‘Q4 ’, each consisting of 206 players depending on the number of Balls Bowled. I have chosen four categories for simplicity; this could be lower or higher by choice.
Players in the categories ‘Q1’ and ‘Q2’ are not necessarily bowlers; they can be batsmen or part-timers. It’s not fair to comment on them; that’s what I feel.
Players from ‘Q3’ and ‘Q4’ are bowlers. Still, most of them have stopped playing, which indicates that either bowler nowadays doesn’t bowl that much of No-Balls (Numerator) or they play a lot of matches, so the number of balls bowled is higher (Denominator). A yearly analysis could help find that out.
Minimum number of No-Balls Per Balls Bowled
The number of No-Balls can’t be less than 0, so starting with all bowlers who have bowled 0 No-Balls and find the player who bowled the most balls.
As expected, players from Q1 and Q2 are primarily part-timers. Nevertheless, a shout out to them as they didn’t bowl s single No-Ball.
Here is the real deal—the players who bowled significantly but didn’t bowl a single No-Ball. The player at the Podium is Mohammad Hafeez. He hasn’t bowled a single No-Ball from his 10410 balls bowled. He represents Pakistan, a team that has produced excellent pacers like Shoaib Akhtar, Waqar Younis, Wasim Akram and many more over the years. Still, this feat is also something to be proud of. I would also like to mention that this analysis is just a superficial one, and one should do more investigation to conclude something.
This analysis could be a never-ending process using different KPI’s, using shorter or longer timeframe or different leagues, but that’s for some other day. I want to end this series for now (maybe something later) or perhaps move to technical aspects of the visualisations. I want to thank all who have been reading and supporting me.
I hope you all liked the article. Feel free to connect with me on LinkedIn.
