We all are concerned about the batting consistency of our national team players. I was trying to figure out a way assign a numerical measurement of batting consistency.
Here is the consistency rating of our players. Higher rating means high inconsistency and lower rating means better consistency.
Batting Consistency Rating
The rating shows the numbers of innings interval a batsman takes to bat a certain score. The consistency rating table shows the consistency level of our national team batsman. I included rating of Steve Waugh and Sachin Tendulker to give an idea of wolrd-class batting consistency level.
Consistency Rating at 30+
For scoring 30 & up, the 3 most consistent players are:
1. Manjurul Islan Rana (1.20)
2. Rajin Saleh (1.25)
3. Habibul Bashar (1.31)
It shows, on an average, Manjurul Islam Rana scores a 30+ innings after a interval of 1.20 innings. For Sachin Tendulker, the gap of innings is just 0.80.
Consistency Rating at 40+
But the consistency rating changes for higher scores. For scoring a 40+ innings, most consistent players are:
1. Rajin Saleh (1.70)
2. Habibul Bashar (1.79)
3. Nafis Iqbal (2.00)
Interesting thing, Manjurul Islam Rana, the most consistent player at 30+ level, is not good at all at 40+ levels. His rating falls down to 4.50. This means, he is a consistent player in scoring 30’s, but he can’t consistently produce 40+.
Consistency Rating at 50+
For scoring a 50+ innings, most consistent players are:
1. Habibul Bashar (2.05)
2. Nafis Iqbal (3.00)
3. Md. Rafiq (3.25)
To some extent, Habibul Bashar rocks here. Rafiq’s higher rating is kind of inflated by too many not-outs. Next in line for 50+ is Md. Ashraful (4.44). Aftab’s rating is very low all along. It’s because he played too few tests. Also his poor test scoring rates created negative impact on this rating. It will certainly improve as he plays more tests and make big scores.
I hope, this numerical consistency rating will be a handful tool for comparing batting consistency of different players.
Calculation Appendix
Initially I tried with conventional statistical model using standard deviation, kurtosis and skewness. But it was not helpful for various reasons (i.e. too low and too high scores evaluated equally). So, I followed an alternative way to measuring batting consistency level. First I made a chronological list of all the Test Innings played by each batsman. Then I figured out the gap in between the innings to produce different scores. And the final rating is the average of the innings gap for scoring at certain score level. Please see the example below for a simplified version of calculations used in the rating.
The details version of the calculations also makes adjustments for “Not Out” scores and “open-end” intervals. If you are interested, I can provide details of the calculations.