A criticism sometimes made of Don Bradman’s career record — generally by fans of a more recent player attempting to claim that their favourite was better — is that he only played Tests in two countries. His famous average of 99.94 would, they claim, have been brought down to earth if he had played in other countries — specifically in India, which they think of as the ultimate challenge for Australian batsmen, and one in which very few succeed. Michael Jones attempts to assess, by comparison with other batsmen, how Bradman might have fared in India.
It is a well-rehearsed argument: modern batsmen (usually intended to mean one modern batsman in particular) must be better than their predecessors, because they have to play in more different conditions — along with others such as video analysis putting their techniques under greater scrutiny from opposing bowlers and coaches, supposedly more punishing modern scheduling etc. Advocates of the greats of the past counter by pointing out that batsmen in the 21st century benefit from bigger and more technologically advanced bats; better protective equipment giving them the freedom to take risks, knowing that their chances of injury are reduced; and the covering of pitches, meaning that they will never have to experience the difficulty of batting on a wet wicket which players prior to the 1960s knew all too well. I do not intend to go into all these arguments here: they are for another day. I will only look at the question of playing in different countries, specifically India.
Bradman only missed one full series during his career: the 1935-36 tour to South Africa, when doctors advised him not to rest as he was still recovering from the bout of appendicitis he suffered at the end of the 1934 England series. South Africa were a weak side at the time, and Arunabha Sengupta has previous examined how other Australian batsmen, even those who fared poorly against England, boosted their averages with big scores there.
Incidentally, we can add advances in medicine to the list of differences between eras — when Brett Lee was diagnosed with appendicitis in 2011, all that was required was a minor operation, and he was playing again within months; Bradman’s case was so severe that newspapers editors instructed their cricket reporters to prepare obituaries, and his wife was actually told he had died, a report only corrected by a phone call to the hospital.
As to India, there was nothing for him to miss: Australia did not play their first Test series there until some years after Bradman’s retirement. If we are to estimate how many he might have averaged if he had played there, we must look elsewhere for comparisons.
Before doing so, let us consider a hypothetical scenario — suppose Bradman had played 10 Tests in India, a reasonable number for a full career: two 5-match series or three shorter ones.
The worst possible scenario would be that he made a pair in every match. In this case, his career aggregate of 6,996 would be divided by 90 completed innings instead of 70, and the famous figure of 99.94 would be reduced to 77.73. Remember that, given a reasonable minimum number of innings, the highest career batting average of anyone other than Bradman is 61.87 (this is not the place for arguments over whether that figure for Adam Voges is misrepresentative of his true status as a batsman: enough of those have been had elsewhere).
In other words, even if Bradman had made 20 ducks in India (or anywhere else in the world), his average would still have been more than 15 runs higher than any other batsman in history. For it to be reduced to the level of the next best would have required him to have been dismissed a further 43 times without scoring another run. That is the weight of statistical evidence which anyone would have to overcome in order to argue that Bradman was not the greatest batsman ever.
Let us consider the general perception that Australian batsmen do poorly in India. It is easy to see how it has arisen: during the period 1995 to 2005, when they were all but invincible everywhere else, India remained their seemingly unconquerable final frontier. In 2000-01 they won the first Test and enforced the follow-on in the second, but still lost the match and the series. Since finally breaking their bogey with a historic series victory in 2004-05, it has returned to its previous status as a fortress they cannot breach: their four series since then have yielded 3 draws, 10 defeats and a solitary victory at Pune this year.
Fans wishing to emphasise how difficult Australian batsmen find Indian conditions invariably point to the record of Ricky Ponting. While Ponting — otherwise the greatest Australian batsman of his generation — was certainly poor in India (average 26.48), he is only one player. How did others fare? There are 20 Australians with over 5,000 Test runs, with Steven Smith the latest to join the groupafter reaching the milestone during the recent series — a reasonable sample size. Let us first see how they fared against India overall. All 20 played at least one series, with Greg Chappell’s 3 matches the fewest.
|Batsman||Overall||Against India||Rest of career||Diff||Ratio|
The figure which sticks out like a sore thumb is Bradman’s own: which his career average was phenomenal, his single series against India was freakish even by his own standards. He started the series with 185 at Brisbane, on a pitch where the next highest score on his own side was 58 and the opposition only made 156 in 22 innings (plus extras) between them. A rare failure followed in Sydney, bowled by Vijay Hazare for only 13, but he came back with 132 and 127* at Melbourne (the only time in his Test career he made a century in each innings: he usually scored enough to ensure he didn’t have to bat a second time), 201 at Adelaide, and had reached 57 in the second Melbourne match when a torn rib muscle forced him to retire hurt.
The other batsmen are a mixture. David Boon and Steven Smith have fared exceptionally well against India while Greg Chappell — albeit only in a single series — and Doug Walters also averaged significantly higher against that opposition than they did overall. Ponting has a marginally higher average against India, his poor record away being balanced out by an outstanding one at home. Adam Gilchrist and Michael Slater came off worst in their contests with India, while David Warner and both Waugh twins also have averages significantly lower than their career figures.
Taking them all together, the ‘average’ leading Australian batsman scored 5.91 more runs per dismissal against India than they did against all other opposition, but this figure is skewed by Bradman’s remarkable difference of 83.58. Take him away, and the difference is only 1.83. Put simply, India have, over the course of their history, been average opposition for Australian batsmen: they score almost exactly the same against Indian attacks as they do against anyone else.
Now let us examine home and away performances separately.
|Batsman||Overall||Home against India||Rest of career||Diff||Ratio|
Bradman’s figures remain exactly the same, since his only series against India was at home; so do Greg Chappell’s. Smith’s single home series against India to date brought him almost Bradmanesque figures. Walters also enjoyed himself in one series, while Neil Harvey’s high average is based on only 2 innings; Ponting, Michael Clarke and Boon all enjoyed more sustained success.
Of the five batsmen who fared worst against India overall, only Warner has done relatively well at home: both Waughs, Slater and Gilchrist failed to take advantage of more familiar conditions. Overall, the top Australian batsmen have scored at 15.67runs per dismissal higher at home against India than they have in aggregate; even if Bradman’s extreme result is excluded, there is still an improvement in average of 12.10. This should come as little surprise: Indian bowlers have typically struggled in Australian conditions, and with a few notable exceptions the home batsmen have taken advantage of it.
|Batsman||Overall||Away against India||Rest of career||Diff||Ratio||Diff H-A||Ratio H/A|
Bradman and Greg Chappell are the only two on the list who never played in India. Australia’s only series there during Chappell’s career, in 1979-80, fell shortly after World Series Cricket had come to an agreement with the Australian board for the WSC players to be allowed back into the national team, but it was felt they would have had insufficient match practice before the tour and thus were not selected — although even out of practice they would probably still have been better than the severely weakened team which was sent.
However, there are still figures for 18 players to consider. Ponting’s is the most obvious contrast with his career record, and his average in India only reached as high as its eventual 26.48 thanks to significant improvement in his final two tours there: it was even worse before that (Wisden 2007 noted that at the time of publication, Ponting averaged 108.10 in 7 Tests at home against India, but only 12.28 in 8 Tests away).
Most of the batsmen on the list actually did fairly well: Boon leads with an average of 65 from one series, but Smith, Harvey, Matthew Hayden and Allan Border all maintained averages above 50 over longer periods, and 12 out of the 18 averaged above 40 — not bad figures for what are widely regarded as the most difficult conditions an Australian batsman can face.
A curious feature of the table is that only one batsman, Mark Taylor, has an average in the 30s: while most did well, those who did badly did very badly indeed. Besides Ponting, four others average below 30 — Justin Langer, Gilchrist, Slater and Warner, although the last of them is still active so he has a chance to improve on his record.
The last four columns are the ones which provide our first method of estimating how Bradman might have fared. First is the difference between the player’s average in India and his career average; for most of the batsmen this is negative, but only for the aforementioned five does it reach double figures. Boon’s high positive figure is a result of comparing a single series to a whole career.
The 18 batsmen between them averaged 6.34 fewer in India than they did everywhere else; if we suppose that Bradman would have done as much worse in India as the average Australian batsman did, that would still leave him with an average of 93.60. One can argue that it should be the ratio between averages, rather than the difference, which is important here — since a difference of 10 runs is clearly more significant for a player who averaged 40 overall than for one who averaged 100. Most are at least above 0.8, but Slater and Langer averaged a third less than elsewhere, while Ponting and Warner were only half the batsmen in India that they were in the rest of the world.
The geometric mean of the 18 figures (the average based on multiplication of the values rather than the usual addition, since we are dealing with ratios) is 0.84: the average Australian batsman is 16 per cent worse in India than elsewhere. If we suppose that Bradman would have done as much worse as that average, he ends up with a figure of 84.03 — below his usual standards, but still a record which most batsmen could only dream of. Even in the worst-case scenario that Bradman found Indian conditions just as difficult as Ponting or Warner, that would still give him an average of around 75 (based on difference) or 50 (ratio); if he prospered as Boon did it would rise to 120 (difference) or 150 (ratio).
The last two columns give, respectively, the difference and ratio between the batsman’s home and away averages against India. Unsurprisingly, most have a positive difference (and thus a ratio greater than 1): common wisdom is that playing at home is usually an advantage.
In ratio terms Ponting is the extreme here, with a home average more than thrice his away one, while Smith has an excellent record against India away but phenomenal one at home.
The same applies, albeit less extremely, to Walters, Hayden and Clarke: they were good in India, but even better against them at home. The Waugh twins — Mark in particular — and Ian Chappell stand out as anomalies in that they did much better away than at home.
In aggregate, the 17 averaged 16.56 more at home than away, or 1.26 times as much. Remember that Bradman’s average at home against India was 178.75: if he had done as much worse away as the other batsmen did, that would still have given him a figure of 162.19 (based on difference) or 129.48 (ratio). The worst-case scenario, of faring as badly as Ponting, would result in an average of around 119 or 55 by each method.
The careers of the batsmen considered above span a period from before India’s inaugural Test to the present day, so any calculation comparing Bradman to an average of the others assumes that the opposition he faced would be “average India” — that is, an Indian team as good as it has been on average over the course of its Test history. India, like any other team, have not been completely consistent over that history: there have been some times when they were markedly better than others — and, like most teams, one of the times when they were at their poorest was in their early years of Tests, which happened to coincide with the peak of Bradman’s career. What if we examine Indian teams during that era?
Although Bradman’s Test career lasted twenty years, and India played their inaugural Test only a few years into it, they only hosted one Test series during the period, against England in 1933-34. England won the first and third Tests comfortably, while India held on to draw the second after setting a fourth-innings target of 82, but leaving the visitors only half an hour in which to score them — a challenge which might be considered achievable in the T20 era, but certainly not in the 1930s.
Let us examine this series, and attempt to estimate how Bradman might have fared if Australia had toured at that time instead. Indian conditions are, if recent history is anything to judge by, just as difficult for English batsmen as they are for Australians, so witnessing how the former fared should give us some clue as to how the latter might have done in their place.
What was India’s bowling like then? Amar Singh led the attack, with 14 wickets in the series at an average of 27.28. Mohammad Nissar took 7 at 32.71 in the first 2 matches before missing the third, in which his replacement Nazir Ali took 4 at 20.75. The rest contributed little: Lala Amarnath 4 wickets in 3 matches, CK Nayudu 3. Rustomji Jamshedji returned figures of 3 for 137 in the only Test of his career.
An important point to note is that the squad who travelled to India were very far from being a full-strength England side. Just as with New Zealand and West Indies four years earlier, MCC — who were responsible for organising touring teams at the time — judged that the teams new to Test cricket were so weak it wasn’t worth asking the best players to make the trip.
Who were the best batsmen available at the time? Jack Hobbs had played his last Test in 1930, while Len Hutton, Denis Compton and Bill Edrich were still a few years off making their debuts. Frank Woolley was close to the end of his career, and scored 4 and 0 when he played his final Test in the following summer’s Ashes. Herbert Sutcliffe was still playing at international level, though, and Wally Hammond was in his prime; Patsy Hendren, Eddie Paynter and Maurice Leyland were also active, while Les Ames was the best keeper-batsman. None of those went to India: the only top-order batsman who would have been likely to retain his place in a full strength team was the captain, Douglas Jardine, returning to the country of his birth. It was a second-string team, yet they still won comfortably — that illustrates how weak India were at the time, and may portend what might have happened if Bradman had toured then.
What of the batsmen who did go on the tour? Jardine’s overall average of 48 suggests he was a good, if not great, Test batsman. Digging a little deeper, however, tells a different story. He averaged 96 against India, and was never dismissed for less than 50 against them — in 2 innings at home the previous year and 4 on the 1933-34 tour, he crossed fifty 5 times and scored an unbeaten 35.
Against New Zealand it was 59, although in 2 of his 5 innings against them he came to the wicket shortly before a declaration and shortly before the end of a drawn match; scores of 7* and 0* do not tell us much.
Against West Indies his average was 63, although here we must give him some credit: they, too, were a team new to Test cricket, but they still had an excellent pair of opening bowlers in Manny Martindale and Learie Constantine, and when Jardine had made his only Test century the previous summer he proved he was willing to drink his own medicine, in withstanding unflinchingly the same Bodyline tactics which he had employed in Australia a few months earlier.
However, Australia were unquestionably the best team of the era, and for all Jardine had antagonised them as captain, he averaged a shade under 32 against them with the bat, repeatedly undone by Bert Ironmonger and Tim Wall. To summarise: Jardine averaged more than twice as many on that India tour as he did against Australia.
Trivia-lovers will know that Bryan Valentine was the first batsman to score a Test century in India. Some may even remember that he played in the last timeless Test at Durban five years later, and was at the crease when the match was finally called off with England 654 for 5 in the fourth innings. Few will have heard of him for much else, and there is a reason for that: he was not very good. His final Test batting average of 64.85 was merely a result of only playing two series, both against weak teams: 59.67 against India and 68.75 against South Africa (the attack England faced in the latter series — Bob Newson, Chud Langton, Norman Gordon and Eric Dalton — was not up to much either). He never played a Test at home, and never one against Australia, because those were the circumstances in which England fielded a full-strength team — and that certainly didn’t include Valentine. In a 23-year career at First-Class level, he averaged 30.15: he scored all but twice as many runs per dismissal against India, in India, as he did against county attacks.
Cyril Walters did play against Australia — in the 5 Tests of the home series in 1934 — and fared better than Jardine: 4 half-centuries in the series and an average of 50.12. However, that jumped to 71 in India — 1.4 times as much.
James Langridge was another who did not play Australia. His average of 37 in India was unspectacular, but still better than he managed against anyone else: he had scores of 70 and 46 in the series, whereas in 1 Test against South Africa and 2 against West Indies he failed to reach 30. His First-Class batting average was 35.20: he improved on that in India, although not to the same degree which Valentine did.
Arthur Mitchell and Charlie Barnett were the only top-order batsmen who struggled in India. They were Mitchell’s first Tests, and he made only 114 in 5 innings; he did better against South Africa in 1935, scoring twin fifties, but never played a Test against any other opposition.
Barnett was the first, but certainly not the last, genuinely Test-class visiting batsman to fail in India: he had scored 52 on debut against West Indies and went on to make 2 Ashes centuries, but on the India tour he failed to pass 33 and was dismissed in single figures thrice. Fred Bakewell played in only one match, making 85 and 4: too small a sample to draw any meaningful conclusions from.
We may note in passing that Harry Elliott, the team’s wicketkeeper in an era when they were not expected to be able to bat, made 37* and 14 against India; while again this is too small a sample to form the basis of any serious analysis, his scores of 1, 3 and 6 in his other 3 Test innings suggest that he, too, found India rather easier than other opposition.
Hedley Verity, the only player other than Jardine who had a significant international career, made his maiden Test fifty, although he did later make one against Australia too: his batting ability was not entirely to be scoffed at.
How can we distil these figures into a measure of the difficulty of facing that particular Indian team in their home conditions? For the three batsmen who did play Tests against Australia, let us compare their performances in that series to those against the strongest team of their generation:
|Batsman||Against Australia||In India||Diff||Ratio|
Combined, they scored 13.79 runs per dismissal more in India than they did against Australia, or 1.14 times as many. Of course Bradman never played against Australia, so let us use for comparison his average against the best opposition he did face — England. That figure was 89.78 (his overall average was raised to the familiar 99.94 by scoring even more heavily against South Africa and India); if we assume that Bradman’s performance there would have improved by as much as the average English batsman, his result would have been almost the same whether calculated using the difference or ratio: 103.57 and 102.37 respectively. The worst case scenario, that he struggled in India as badly as Barnett did, would have seen his average descend to the high 60s (difference) or low 40s (ratio); if Bradman had prospered to the extent that Jardine did, the figure would have sailed into the 130s or even over 200.
For the remaining three, it is difficult to obtain meaningful results from comparisons with the rest of their Test careers, since one would merely be comparing one weak team with another. Let us instead use their First-Class averages as the yardstick, assuming the average strength of county opposition faced by each would be roughly constant.
|Batsman||FC career||In India||Diff||Ratio|
The three between them averaged 5.72 more in India than they did in their First-Class careers overall, 1.09 times as many runs per dismissal — Mitchell’s failure was more than balanced out by the successes of the other two.
Bradman averaged ‘only’ 95.14 at First-Class level, almost 5 runs fewer per dismissal than he did in Tests — perhaps, at least in part, because he had to face Clarrie Grimmett and Bill O’Reilly in domestic matches, which he never did in internationals. Although no bowler ever completely dominated Bradman, the leg-spin duo managed to contain him slightly more successfully than most others did.
Nevertheless, if Bradman had improved on his overall First-Class average to the extent that Valentine, Langridge and Mitchell did between them, he would have averaged 100.86 in India (difference) or 103.38 (ratio). Even if he had found Indian conditions as difficult as Mitchell did, his average would have fallen to about 60 at worst; if he had found them as much to his liking as Valentine he could have come close to 190.
The next team to play Tests in India was West Indies in 1948-49, who won the 5-match series 1-0; it came only a few months after Bradman played his last Test, so it should give some indication of how strong an attack he would have faced if he had played there towards the end of his career.
The playing days of Nissar and Amar Singh were long gone: the attack was led by Vinoo Mankad, who took 17 wickets but paid the rather high price of 43.76 for each, and Dattu Phadkar, whose 14 came at the more respectable average of 29.35. The debutant Ghulam Ahmed took 8 at 45.12, while two Banerjees, Shute and Mantu, took 5 wickets each in what, for both of them, turned out to be their only Test.
Unlike pre-War England, post-War West Indies selected from the same pool of players regardless of the identity of the opposition: all their top order on the India tour also played Tests against England and Australia, some of them also New Zealand and Pakistan, so they all have a reasonable sample size for the rest of their Test career, against which we can compare their record in India.
One batsman took a particular liking to the home attack, and set a record which still stands: while there is no doubt that Everton Weekes was a world-class player — a Test batting average of 58.61 establishes that — his scores in the India series were freakish. Having made 141 in his previous innings, at home against England earlier in the year, he reeled off 128, 194, 162 and 101 in his first 4 innings in the subcontinent, still the only batsman to score 5 consecutive Test centuries — and only a run out for 90 in Madras stopped him making it 6; finishing the series with 56 and 48 in Bombay brought his average down slightly, but it still stood at 111.28 for the series.
Jeff Stollmeyer was a key figure in the history of West Indian cricket off the field, but he fell short of greatness as a batsman, averaging just under 40 in the rest of his Test career — a figure boosted to 42.33 by his success in India. At Madras he made 160, his highest Test score, and added 239 for the first wicket with Allan Rae; either side of that were scores of 66 and 85 in the 2 matches at Bombay, and he finished the series with at 68.40.
Rae had made his debut in the first match of the series, scored 104 in the second, then 109 in the partnership with Stollmeyer and a near-miss with 97 in the second Bombay match; he only scored 2 other Test centuries and his average of 53.43 in India was more than 10 runs higher than the rest of his career.
Another of the three Ws, Clyde Walcott, also played in India (Frank Worrell was not on the tour). Although the difference was not as marked as it was for Weekes and Stollmeyer, he too did better there than against other opposition: 2 centuries, 2 fifties and an average of 64.57.
Robert Christiani had a modest overall Test record, but at Delhi he made his only century (although he had scored 99 on debut against England earlier that year). Gerry Gomez and John Goddard, both also with greater achievements to their names off the field, were allrounders whose bowling was somewhat better than their batting; Gomez, like Christiani, scored his only Test century in the India series, while Goddard failed to reach 50 but had his average boosted to 47.50 by 2 not out scores. All three averaged significantly higher in India than elsewhere. George Headley only played one innings in the series; let us disregard him in analysing the averages.
Between them, the seven averaged 21.89 runs more in India, or 1.55 times as many per dismissal as in the rest of their careers. Applying this difference and ratio to Bradman, we get estimates of 121.83 or 154.54 if he had toured India at that time.
|Batsman||Rest of Test career||In India||Diff||Ratio|
Australia did not play their first Test series in India until 1956-57 — more than eight years after Bradman’s last Test.Of the five bowlers who took wickets for the home team in the series, Mankad was the only one who had made his Test debut while Bradman was still playing, so it is of limited help in determining how he might have fared against their attacks of the 1930s and 1940s.
Let us recap the estimates of his likely average had he played in India: 93.60, 84.03, 162.19, 129.48,103.57,102.37,100.86,103.38,121.83,154.54. There is significant variation between the figures, which is not entirely unexpected when attempting to calculate the hypothetical outcome of a situation which never occurred in reality; however, most of them are in the range 100-130, and it is noticeable that the only two methods of estimation which yield a result lower than his career average are those based on the entirety of India’s Test history: every method based on the strength of Indian teams in his own era gives an estimate of over 100. The average of all the estimates is 115.59.
How would a hypothetical India tour have affected his career average? To estimate this, let us assume that the series would have been of 5 Tests, since that was standard for Australia in the era: they played 5 either at home or away against England (the 1938 series was reduced to 4 due to 1 being washed out), 5 in South Africa in 1935-36 and 5 at home to India in 1947-48.
Bradman would surely not have been required to bat in 10 innings: he never did in a series, because his team so often won by an innings or 10 wickets. In the home series against India he only went to the crease 6 times, and finished not out in 2 of them. Suppose the hypothetical series followed the same pattern, and he would have been dismissed only 4 times. Calculating series records based on those averages (rounded in order to give an integer number of runs in the series), and adding them to his actual career record, yields the following results:
|Estimate based on comparison with||India series||Career|
|All Aus batsmen in India (d)||5||374||93.50||57||7370||99.59|
|All Aus batsmen in India (r)||5||336||84.00||57||7332||99.08|
|Own record home vs India (d)||5||649||162.25||57||7645||103.31|
|Own record home vs India (r)||5||518||129.50||57||7514||101.54|
|Eng in India 1933-34/Aus (d)||5||414||103.50||57||7410||100.14|
|Eng in India 1933-34/Aus (r)||5||409||102.25||57||7405||100.07|
|Eng in India 1933-34/FC (d)||5||403||100.75||57||7399||99.99|
|Eng in India 1933-34/FC (r)||5||414||103.50||57||7410||100.14|
|WI in India 1948-49 (d)||5||487||121.75||57||7483||101.12|
|WI in India 1948-49 (r)||5||618||154.50||57||7614||102.89|
|Average of above||5||462||115.50||57||7458||100.78|
At worst, his career average would drop by a fraction of a run (while remaining streets ahead of every other batsman in history); in almost all other cases it would rise sufficiently for the final innings duck to be rendered irrelevant in any quest to keep it in three figures.
The claim that Bradman’s greatness is somehow diminished by not having played in India, then, fails to hold water: not all Australian batsmen have done as badly there as Ponting did — in fact the majority have done far better — and India in the 1930s and 1940s were a poor team, yet to see the rise of the successive generations of spinners who have bamboozled many a visiting batsman in subsequent decades.
Even mediocre batsmen prospered there in Bradman’s era, and one great rewrote the record books; it is safe to assume that if the man himself had played there, his reputation — and average — would have been enhanced rather than diminished.