Roar Guru
Cricket averages across a series tell a tale of how batsmen fared, and though Steve Smith and Alastair Cook lead their teams, but that doesn’t really tell the full story of how consistent they were across the series.
So what would happen if we applied the interquartile mean (IQM) to their batting scores? How would the two teams fare?
To save everyone jumping on Google, the interquartile mean is a statistical measure that basically drops the highest and lowest value. It’s similar to how gymnastics and diving scores are calculated.
The reason for doing this in analysing scores is that I’m interested in which batsmen were consistent across the series without one large score at the MCG inflating the series average. It also highlights how will Mitch Marsh and Steve Smith played.
Australia
Player | Average | IQM average |
Steve Smith | 137.4 | 147.33 |
Mitch Marsh | 106.67 | 130 |
Shaun Marsh | 74.17 | 71.25 |
David Warner | 63 | 64.8 |
Tim Paine | 48 | 61 |
Pat Cummins | 41.5 | 59 |
Usman Khawaja | 47.57 | 30.2 |
Cameron Bancroft | 25.57 | 16.17 |
Peter Handscomb | 20.67 | 14 |
Nathan Lyon | 9.25 | 11.5 |
Mitchell Starc | 8.8 | 7.67 |
Josh Hazlewood | 5 | 6 |
From those stats we can see that Steve Smith had a blinder of series. Taking out his second innings in Adelaide (six runs) at the cost of his 239 actually increased his series average.
Mitch Marsh lost his 181 at the WACA to go with his MCG nine, and most other batsmen didn’t move around too much. Usman Khawaja is the big loser for the Australians, with his big 171 runs at the SCG bringing his highest score for series down to 53.
The other interesting thing it highlights is how strong Cummins was with the bat. Even if he is swinging for the rope for quick runs, a normal series average of 41.50 and IQM average of 61 is massive. Had he been batting for England, he would be their highest run scorer based on IQM.
England
Player | Average | IQM average |
Tom Curran | 33.00 | Not dismissed |
Joe Root | 43.00 | 42.14 |
Dawid Malan | 42.56 | 34.14 |
Mark Stoneman | 25.78 | 25.14 |
James Vince | 26.89 | 22.43 |
Chris Woakes | 20.71 | 21.80 |
Jonny Bairstow | 30.22 | 21.57 |
Moeen Ali | 20.22 | 20.29 |
Alastair Cook | 47.00 | 18.57 |
Stuart Broad | 15.11 | 11.43 |
Craig Overton | 20.67 | 9.50 |
James Anderson | 2.67 | 1.50 |
If nothing else the English analysis shows Joe Root’s inability to convert 50s into 100s but still consistently get starts. Dropping his highest and lowest (83 and nine) he still averages 42 runs. Stoneman, Woakes and Ali are in a similar position of not having a high water mark that inflates their averages.
On the other side of the spectrum is that MCG knock by Alastair Cook. His average dropped by 28.43 to 18.57 for the series once that score was removed. If someone were to scrap the statistics of the series, an average of 47 from the English opener is a good result, but 18.57 is not. It shows how numbers can spin whatever tale you want them to.
But what does this exercise show us, since I am essentially doctoring the statistics to prove my point? With one series being such a small dataset, it is prone to high scores inflating a batsman’s average, whereas it doesn’t tell the full story of their form. Using the interquartile mean reduces the sample set even more, but it tells a completely different story.
As a final point, the Australian team averaged 50.34 across their batting innings with an IQM average of 49.57. The England team averaged 27.91 and 23.45 respectively.