Test Cricket’s Greatest All-Rounder – Lies, Lies and Damn Statistics

Last summer Australian fans were privileged to see one of the modern greats of world cricket in Jacques Kallis, and his appearance on these shores led to considerable discussion among the Channel 9 and ABC commentary teams about his place in the pantheon of great all-rounders in the history of the game.

The consensus of the punditry was that he is very good, just not as good as a certain West Indian by the name of Sir Garfield Sobers, who incidentally has also been recently rated as the next best batsman after Bradman by some pretty good judges in Mark Nicholas, Martin Crowe and Ian Chappell.

A lot of the discussion was based on the personal experience of playing with or against a particular player or great performances witnessed. This got me to thinking about how one can be so definitive on a subject such as the greatest ever given that Test cricket has been played for 136 years and has had over 2,000 contests. Clearly no-one has witnessed them all, indeed for many it is nigh on impossible to find an account or sepia tinged photograph, let alone a sound or visual recording.

One of the great things about cricket is its consistent range of statistical measures over history, and I believe that it is possible to use these numbers in a way that allows comparison of players across eras in a consistent way, owing less to personal preference, memory and most importantly, to having actually witnessed players like Sobers, Miller, Benaud and the like in the flesh.

So this is a piece on my quest to find an approach that would allow me to contribute to the debate on the greatest all-rounder of all time, and perhaps spark a discussion that might not be as easily settled as one of the protagonists piping up with “well you never saw him play and I can tell you …”. It might be true, but then again millions of others including me have never seen Bradman play and yet he is universally regarded as the greatest without debate. The reason for this is the enormous statistical gap between him and the next best, however I sense the all-rounder question might not be such an open and closed case.

My first step was to establish a criteria for what actually defines an all-rounder. Traditionally, it has been viewed as a player who could genuinely be selected as either a batsman or bowler, however my subjective opinion is that in the history of the game there would be very few on this list. Perhaps this actually defines their greatness, however it is subjective.

I believe there is a simple criteria that at least qualifies a player to be part of the discussion, one which might leave out a couple of surprising players like Alan Davidson, and include some not so obvious players, but which at least forms a basis of comparison and is easy to get data for. Accordingly the criteria I have come up for entry to this list is that firstly a player must have played at least 20 Tests, and that they must have scored at least one Test century AND taken five wickets in an innings at least once. As far as I can gather, surprisingly there are only 97 players in the history of the game that qualify for discussion.

Players were ranked from 1 to 97 across each of a range of factors designed to capture their consistency of excellence, combined with their ability to make a match-winning impact. These factors were then weighted to an overall scale of 100 multiplier points, with the person achieving the lowest score being the top ranked player. The criteria are as follows:

Consistency of Excellence Factors and Multipliers

  1. Batting average = rank X15 multiplier
  2. Bowling average = rank X15 multiplier
  3. Runs per Test = rank X12.5 multiplier
  4. Wickets per Test = rank X12.5 multiplier

Impact Factors and Multipliers

  1. Hundreds per Test = rank X10 multiplier
  2. 5WI per Test = rank X10 multiplier
  3. Bowling S/R = rank X10 multiplier (batting not included as balls faced not always recorded)
  4. Best bowling figures = rank X5 multiplier
  5. Highest score = rank X5 multiplier
  6. Catches per Test = rank X5 multiplier

The rating factors are designed to put all players on an equal footing assuming similar longevity in the game and capture both consistency of performance together with match-winning impact. It differs from the traditional method of taking the difference between the batting and bowling averages, in that it places a higher emphasis on impact factors that have a significant impact on the outcome such as centuries and 5WI hauls.

The analysis throws out a top 10 which contains a couple of surprises but which contains all of the usual suspects. It is based purely on Test match performances and is designed most importantly to highlight those the greatest ability to significantly influence the outcome of a Test with EITHER bat or ball.

So, drumroll please ……. without further ado here they are in reverse order:

10) Mushtaq Mohammad (Pakistan) – a surprising start to my top 10, Mushtaq was one of 5 very talented cricketing brothers that included the great Hanif. Potentially he is the youngest player to have scored a Test century and was Wisden’s cricketer of the year in 1963. In 57 Tests across a 20 year period he scored 3,643 runs at a respectable average of 39.17 with 10 centuries. He is a surprising inclusion given that he only took 79 Test wickets with his leg breaks, however he did so at a good average of 29.22 with 3 five wicket hauls. Considering that he took 936 first class wickets at an excellent average of 24.34 with a strike rate of just over 50 he was probably very much under-utilised as a Test bowler.

9) Chris Cairns (New Zealand) – another surprise entry but a quite deserving one. In 62 Tests over 15 years he scored 3,320 runs at 33.53 at a fast clip and captured 218 wickets at a good average of 29.40. What really gets Cairns into this list is his match winning impact based primarily around his bowling with 13 five wicket hauls, and most surprisingly the best strike rate of any player in the top 10 (53.6)

8) Jack Gregory (Australia) – from a very different era he appeared in 24 Tests between 1920 and 1929. In an era of uncovered pitches his 1,146 runs at 36.96 with 2 centuries is a very meritorious effort combined with 85 wickets at 31.15 with 4 five wicket hauls, albeit figures far inferior to his first class record 504 wickets at an average of 20.99.

7) Shakib al Hasan (Bangladesh) – I hear many of you exclaim “WHAT ??? How does a Bangladeshi make this esteemed list ?”. Understandable yes, but not warranted. Although he has only played 28 Tests to date he plays for a very weak team and in that context his performances are superb – 1,835 runs at 35.98 with 2 centuries and 102 wickets at 32.56 with a staggering 9 five wicket hauls.

6) Tony Greig (England) – the late and much-loved Greigy was a very underrated cricketer who played against some ferocious opponents like Lillee, the Chappells, Roberts, Holding, Richards, Imran, Javed, Kapil Dev, Gavaskar, and the list goes on. Given the quality of the opposition, the excellence of his performances came as a surprise to me when compiling this list. Greigy is a great example of the tricks perhaps memory can play, as when I conducted the exercise off the top of my head he was not in my top 10. In 58 Tests across one of the most competitive eras worldwide, Greigy successfully and combatively captained his adopted country, scoring 3,599 runs at a very proficient 40.43 with 8 centuries, and taking 141 wickets at a respectable 32.20 with 6 five wicket hauls.

Okay so that was the warm up – the real action (and debate) starts now and you could certainly make a case for any of these players being the number one.

5) Jacques Kallis (South Africa) – when the curtain is drawn on his career, Kallis will be regarded as one of the greats across an enormous number of Tests. His total of 13,128 runs at 56.10 with 44 centuries is that of a truly great batsman and the fact that he is near to 300 Test wickets and also one of the great slippers is testimony to his undeniable quality. Where Kallis falls down this list is his match winning ability with the ball having only captured 5 five wicket hauls in his career with an average of 32.43.

4) Keith Miller (Australia) – an exceptional and very fondly remembered player across 55 Tests, I would suggest he was the first truly great all-rounder. Miller was by all accounts a thrilling player with both bat and ball with a terrific back story and endlessly attacking approach. He is probably the first bowling all-rounder on this list and boasts a truly world-class specialist average of 22.97 for his 170 wickets with 7 five wicket hauls and a strike rate of 61.5. With the bat he was a very attacking and hard-hitting player who scored 2,958 runs at 36.97 with a symmetrical 7 centuries to go with his 7 five wicket hauls, a statistic that suggests he performed equally well in either discipline.

3) Imran Khan (Pakistan) – I have always thought Imran to be a vastly underrated player in these discussions of greatness, especially when combined with his inspirational captaincy. Imran was another bowling biased all-rounder however his batting improved dramatically over his career to the point where, in a career spanning a remarkable 21 years, he managed to accrue 3,807 runs over 88 Tests at a more than respectable average of 37.69 with 6 centuries, the majority of which came in the latter part of his career. With the ball Imran is by far the best credentialed player in this top 10, capturing 362 wickets at 22.81 with the world-class strike rate of 53.7 and 23 five wicket hauls.

2) Sir Garfield Sobers (West Indies) – I know this is controversial so I better get to the explanation quick. Across 93 Tests in a 20 year career, Sobers was a world-class bat capable of destroying an attack, hell with 8,032 runs at an average of 57.78 and 26 centuries he may even be the greatest batsman after Bradman. However, like Kallis it is the bowling where he loses out. Yes he could bowl whatever you wanted … left arm fast-medium, leg spin, off spin, slow medium …. BUT although he took 235 wickets his strike rate of 91.9 is NOT that of a true front line bowler in his era or any other, and his average of 34.03 is the worst of any in this top 10 list. When combined with his modest 6 five wicket hauls, Sobers in my opinion would rarely be a match winner with the ball.

And so we come to our surprising, but I think nonetheless deserving winner ….

1) Sir Ian Botham (England) – somewhat appropriately it is another knight of the realm who slips past Sir Garfield into the top position.Certainly Botham was no Sobers with the bat but he was more than useful with 14 centuries in his 5,200 runs at 33.54 and his destructive batting could be match winning, no better exemplified than by his extraordinary hundred against Australia at Headingley in 1981 and a top score of 208. Botham was truly a player who could win a Test with either bat or ball (or even in he slips) across his 15 year career, and his although his 383 wickets came at a surprisingly high average of 28.40, his strike rate of 56.9 and 27 five wicket hauls mark him as a match winner of the highest order.

And so there we have it, at least in my opinion.

Cricket is a great sport that provides so many indelible memories. But what happens when you have not have the fortune to experience the same ones as others with whom you are having a debate. Are you unqualified to participate ? Are the powerful memories of the past, so long-held, more valuable than those more current ? I am not sure I really know, but the other beauty of cricket is the similar statistical records that can be compared and that may shed light on the past. Then again there are lies, lies and damn statistics …..

Until next time, that is stumps.