In his book The Greatest Since My Time, Trevor Bailey wrote that an allrounder is a player who commands a place in the XI either as a batter or as a bowler. Another common definition accepts any player as an allrounder if their batting average exceeds their bowling average. Among the 399 Test players who have at least 50 wickets in men's Test cricket to their name at the time of writing, 46 satisfy this condition. Take away bowlers who bowl fewer than 20 overs per Test match on average and 31 satisfy it. The newest addition to this list at the time of writing is Kyle Jamieson, who has 372 runs at 19.6 and 72 wickets at 19.5.
Jamieson's example points to the difficulties presented by Bailey's definition for the statistical standard. The next three players on this list of 31, arranged by increasing batting averages, are Vernon Philander (24.0), Alan Davidson (24.6) and R Ashwin (26.9). While these three are not tailenders, they would not command a place in their respective Test teams for their batting alone. The allrounder remains difficult to define.
The "genuine allrounder", remains even more controversial. Is Jacques Kallis a genuine allrounder? Was Richard Hadlee one?
Considering batters and bowlers by position, the average runs, wickets and balls bowled per match in each position are in the table below. Readers should note that a team has used an eighth bowler in only 244 out of 2466 Tests, while a tenth has only been used in 18 of those matches. By contrast, a team has used a sixth bowler in 2012 out of 2466 Tests. The figures suggest that a Test XI includes up to eight players who can be considered capable of batting (either as specialists or as capable lower-order bats), and up to six who can be considered capable of bowling (part-time or otherwise)
This enables a couple of tentative definitions. A "genuine allrounder" is one who averages at least 49.3 runs per match with bat and takes at least two wickets per Test with the ball - i.e. their contribution with the bat is equivalent to that of a player batting in the top seven, and with the ball, equivalent to that of a bowler in the top five bowling positions.
The 54,287 Test caps won by players at the time of writing have been plotted in the chart below. They are arranged according to the two measures considered here - wickets per match and runs per match. As an example, Dale Steyn's Test caps are highlighted in the chart (outlined circles with no fills).
Each point on the chart represents one Test cap. A Test cap is a selection to a Test XI earned by a player. Over his career, Sachin Tendulkar earned 200 Test caps.
The caps in blue represent the genuine allrounders. There are 943 such caps, or 1.7% of all Test caps in the history of the format.
The caps in red represent players whose contributions are equivalent to those of a top-eight batter and a top-six bowler, but not those of a top-seven batter and top-five bowler. There are 3345 such caps, or 6.2% of all caps.
The caps in grey represent specialist batters or specialist bowlers (like Steyn).
There is a relatively small number of caps in the bottom left corner of the chart; players who don't contribute with either bat or ball tend to be dropped. The vast majority of caps won by players with zero wickets per Test and fewer than 40 runs per Test have been won by wicketkeepers.
This is illustrated in the chart below, which shows all Test caps earned by players who have played at least 20 Tests; numbers from their first 20 caps are excluded.
Any player who played less than 20 Tests overall does not feature in the chart above. As readers will note, the bottom left-hand corner is emptied out, and the chart only features the specialist bowlers (the grey cluster between two and six wickets per Test and fewer than 37 runs per Test), specialist batters (fewer than one wicket per Test and at least 37 runs per Test), and the two categories of allrounders.
The ruthless meritocracy of Test cricket is illustrated in these two charts. Only 479 out of the 25,261 Test caps won by players who have played at least 20 Tests (or 1.9%) have been won by genuine allrounders. By his 20th Test, Dale Steyn's record had stabilised and his caps since then all hover around his eventual career record mark. Eagle-eyed readers will see some caps around the 130-140 runs per Test mark in the chart. These belong to Don Bradman.
Five categories of players can be readily identified based on their average contribution per match (not per dismissal) at the end of each match:
1. Genuine allrounders, who contribute on average at least 49.3 runs per Test and 2.00 or more wickets per Test
2. Bowling allrounders, who contribute on average between 37.0 and 49.3 runs per Test and at least 2.00 wickets per Test
3. Batting allrounders, who contribute on average at least 49.3 runs per Test and between 1.03 and 2.00 wickets per Test.
4. Specialist batters, who contribute at least 49.3 runs per Test and less than 1.03 wickets per Test.
5. Specialist bowlers, who contribute at least 2.00 wickets per Test and fewer than 37.0 runs per Test.
We now have a systematic account of player contributions. Here, let's introduce an admittedly arbitrary, though hopefully reasonable, threshold. Let's say that to qualify as a successful Test player, a candidate must earn at least 20 Test caps. If the thresholds above are applied to all such players, then there have been 13 genuine allrounders in the history of Test cricket, 17 bowling allrounders, and 15 batting allrounders.
The figures in the tables above contain some categorisations that might seem surprising to cricket fans. There were certainly phases in the careers of a few players who are not in the first table during which they would qualify as genuine allrounders.
Imran Khan's categorisation as a bowling allrounder seems noteworthy. As his Test career developed, his batting advanced, and towards the end, his bowling declined. His last 15 Tests (out of a total of 88) brought him 947 runs (out of a career total of 3807) and only 28 Test wickets (out of a career total of 362). Taken as a whole, his career, the record suggests, is more similar to that of a bowling allrounder (perhaps the foremost in this category) than it is to a genuine allrounder. It goes without saying (as Imran's chart below shows) that he was among the very greatest bowlers to ever play Test cricket.
The batting allrounders list shows Kallis as being more similar to Frank Worrell, Ted Dexter and Wally Hammond, than he is to Garry Sobers or Keith Miller.
There are a few notable omissions in these lists. Four of these players are listed below.
The method of categorising Test players demonstrated in this essay involves giving greater emphasis to contributions than to ability. In Ravindra Jadeja's case, his recent Test matches have featured several eye-catching batting performances, and much like Imran's, the trajectory of his career indicates that he is transitioning from being a top specialist bowler to a top bowling allrounder.
Few would doubt that Shaun Pollock had the ability to be a genuine allrounder, let alone a bowling allrounder. But his record shows that his batting was a scarcely utilised luxury.
Andrew Flintoff's career represents both the stress and the triumph of the all-round contribution. His was a career with a pronounced peak following a long apprenticeship. It is extremely difficult to sustain all-round efforts with bat and ball of the kind Flintoff produced for about three years, over the length of a career.
What the record really shows is that allrounders in Test cricket are born as much out of necessity as from ability. Teams use players in roles and positions that are most advantageous to the side, even if this means keeping some of a player's capabilities on the shelf. England, for example, use Ben Stokes as a shock weapon when their specialist bowlers have been used up. Stokes is an attacking bowler who hits the pitch hard, is capable of being seriously quick from time to time, and is prepared to concede runs in the pursuit of wickets. England's circumstances have turned him into a genuine all-round contributor.
Some of these categorisations may come as a surprise to readers. But while there have unquestionably been notable Test allrounders who dominated Test cricket at their peak, there have also been others who were consistent all-round contributors throughout their careers. Admittedly, this was as much due to their ability as to the circumstances which provided opportunities for them.
Tony Greig for instance, was a freak player, who, quite apart from being a terrific middle-order bat, could also bowl seam and spin (much like the great Sobers). This ability, and England's need to field an allrounder once Ray Illingworth retired, gave Greig the opportunities to match his capabilities. His extraordinary record is the result. Would Jadeja have had a Test record similar to Shakib Al Hasan's if he played in a Test team that could accommodate him in the middle order? Very likely, yes. It just so happens that Jadeja has played in a very strong Indian Test XI, where his spin bowling has been far more valuable than his batting.
In an alternative universe, Pollock, Jadeja, Kallis and Hadlee would all play for teams that would desperately require every last bit of their considerable all-round talents. In this universe, the list of the most prolific Test allrounders would look very different.