Cricket's Greatest Test Batting Averages Explained

What Batting Average Measures

A batting average in Test cricket is calculated as runs scored divided by innings in which the batsman was dismissed. Not outs (innings where the batsman remained not out) are excluded from the denominator — meaning a batsman who scores 50* repeatedly in winning chases has those runs counted without the dismissal. This has always been cricket statistics' most significant limitation: lower-order batsmen and captains who bat long in winning causes are systematically advantaged. The average was historically meaningful as the primary measure of a batsman's worth: how many runs can this player be expected to score before being dismissed? The higher the number, the better the run-for-wicket economy.

Bradman's 99.94: The Unreachable Summit

Don Bradman's Test batting average of 99.94 across 80 innings (52 Tests, 1928–1948) is the most famous number in cricket and arguably the most remarkable individual statistical achievement in any major sport. The next closest career Test average for a player with 20+ innings is Graeme Pollock (60.97 in 23 Tests) and Herbert Sutcliffe (60.73 in 54 Tests). Bradman is not 10% better than the next best; he is 60% better. An average of 99.94 means Bradman was, on average, dismissed once for approximately every 100 runs he scored across his entire Test career. The closest any modern player has come: Steve Smith's 62.58 average after 100 Tests, which is extraordinary. It is also 37 runs below Bradman.

Modern Greats: Smith, Kohli, Williamson

The 'Fab Four' debate — Steve Smith, Virat Kohli, Kane Williamson, and Joe Root — dominated cricket analysis from approximately 2018 onwards. All four have career Test averages of 50+ across 100+ Tests. Steve Smith led the group at approximately 62; Kohli at 48+ (affected by a period of inconsistency in England 2021); Root at 50+ consistently across 150 Tests; Williamson at 54+. The significant contextual point: all four play in the post-DRS era, where LBW and caught-behind decisions are more accurately called, reducing the 'gift wickets' that marginally inflated averages in earlier eras. Comparing their averages to 1970s or 1980s players is genuinely difficult because of this systematic difference.

The Not Out Distortion

The not-out problem becomes significant for specific players. MS Dhoni's Test batting average of 38.09 across 350 innings includes 33 not outs — if those not outs are included in the denominator, his 'adjusted average' drops to approximately 35. Conversely, Curtly Ambrose and Malcolm Marshall were dismissed infrequently (as tail-enders bat briefly) meaning their averages of 11–16 are mathematically affected by not outs in ways that make their batting contribution appear better than their skill level. Modern cricket analytics has moved toward 'runs per dismissal' as a supplement to averages; mean average is increasingly supplemented with median score (the typical innings) which better represents consistency.

Context Adjustments and ERA+ Equivalents

Cricket statistical analysis has developed 'era-adjusted' averages that attempt to normalise for the conditions each player faced. A batsman averaging 50 in the 1950s — on uncovered pitches, without helmets, against limited bowling attack depth — is not directly comparable to a batsman averaging 50 in the 2020s on modern flat pitches with protective equipment. Several statistical models suggest the 'effective adjustment' required is between 5–10 runs per era (pre-1960 cricket being harder for batsmen). Under these models, Bradman's 99.94 reduces to approximately 88–90 adjusted — still the most extraordinary number in cricket history but contextualised against the era's specific difficulty. Steve Smith's 62+ adjusted for era (pitches flatter than the 1990s) is approximately 58–60. The adjustments change nothing fundamental: the hierarchy of greatness remains the same.