Consistency

Go Back

The analysis of consistency was made from a dataset including data for all IPL matches between ~2011 to 2021. Only data for batsmen who had atleast 5 innings was used. Every player’s performance with a team was calculated for (a) Batting Average and (b) Consistency. The Gini coefficient was used to calculate consistency. Here are the results.

You can hover over each dot to see the player’s name and team.

Players such as Rahul and Warner have high batting averages but average consistencies with their performances. On the contrary, Jadeja and al Hassan’s stints with their respective teams has resulted in a lower average but a very high consistency.

Outliers

  • Consistency

    • RA Jadeja’s performance with CSK the most consistent in IPL. Other notable mentions are SPD Smith with Pune, and STR Binny with RCB.
    • DL Chahar’s performance with CSK was the least consistent.
  • Batting Average

    • KL Rahul with PBKS has had the highest batting average. Other notable mentions are DA Warner with SRH, RD Gaikwad with CSK.
    • R Bhatia with DC holds the lowest batting average. Most players in this range are bowlers who’ve played for a long time with one team.

Rating

Using the batting average and consistency, the overall performance of the player with a team was calculated with the formula Rating = (1 - Gini) * Batting Average. The results were as follows.

Name Team Average AConsistency NumberInnings Rating
V Kohli Royal Challengers Bangalore 81.1 2.923977 16 53.364
MS Dhoni Chennai Super Kings 74.2 3.134796 15 50.530
MS Dhoni Chennai Super Kings 83.2 2.421308 12 48.838
KL Rahul Punjab Kings 69.6 2.673797 13 43.570
DA Warner Sunrisers Hyderabad 57.7 3.968254 12 43.160
KD Karthik Kolkata Knight Riders 55.3 3.968254 16 41.364
DA Warner Sunrisers Hyderabad 60.6 2.994012 17 40.360
MS Dhoni Chennai Super Kings 65.0 2.631579 15 40.300
DA Warner Sunrisers Hyderabad 58.3 2.717391 14 36.846
KL Rahul Punjab Kings 54.9 2.985075 14 36.508

What follows is also an interesting scatter plot that shows the number of innings against the player’s rating. This shows that players such as Maxwell and KL Rahul, despite not having played as many innings over the past decade, have had a much higher rating than players such as Kohli and Raina who have played upwards of 150 innings.

Team Performances

In addition to the performances of individual players, the performances of a hypothetical average player in a team can also be taken. This is calculated by simply finding the mean of consistency, average, and number of innings of all significant batsmen of a team. Find the figures below.

Team MeanAvg MeanConsistency MeanInnings MeanRating
Chennai Super Kings 33.15873 2.236422 27.95238 18.62449
Rising Pune Supergiants 33.20714 2.255518 16.21429 18.56564
Royal Challengers Bangalore 30.67206 2.174747 22.69118 16.89234
Kochi Tuskers Kerala 27.40000 2.441406 13.00000 16.40560
Sunrisers Hyderabad 27.59322 2.238410 21.28814 15.78075
Kolkata Knight Riders 28.53750 2.111821 24.71250 15.40527
Mumbai Indians 27.70714 2.187842 24.82143 15.33676
Punjab Kings 28.00130 2.125956 22.29870 15.09225
Delhi Capitals 26.88987 2.110268 22.12658 14.38447
Rajasthan Royals 26.92059 2.114034 21.29412 14.38256
Pune Warriors 24.28947 2.345969 18.36842 14.29416
Gujarat Lions 27.08571 2.104630 17.04762 14.22395
Deccan Chargers 25.86957 2.127660 20.95652 14.10257

This table quite clearly shows who has performed the best. Teams which have lasted only for a few years clearly have an advantage, with the best large team coming out to be CSK.

Note: The black lines represent the Median values for average and consistency respectively.

Here is the same data represented in a scatter plot. This reveals an interesting trend; despite the consistency of an average player tending to be almost the same for all major teams, there is a large difference in the batting averages. CSK again stands out, with an average of almost 27 runs, a full 2 runs ahead of the next major team (KKR). SRH appears to be the worst performing team with an average of just over 22.

Notes

  • The consistency value in all places refers to the 1/Gini or 1-Gini. It’s always better to have a high consistency.

  • The data (Gini coefficient, average) has been calculated in JavaScript, while the analysis has been done in R.

  • Data obtained from CricSheet. Click to open