It has become one of the most familiar sights in Test cricket. The new ball moves, two batters nick behind, another is trapped on the crease, and before the morning has properly settled the scoreboard reads something like 35 for 3. Within minutes, the comparisons begin. Where are the Dravids who could leave for an hour? Where are the Pontings who would dominate once they survived the first spell? Where are the batters who treated a hundred as the start of an innings rather than its destination?
There is plenty of evidence to feed that nostalgia. Scores of 300 now regularly win Tests, top-order averages have dropped sharply and even established batters can go entire series without producing the kind of marathon innings that once seemed routine. But before declaring that an entire generation has forgotten how to bat, there is another question worth asking: what if the examination itself has become considerably harder?
The numbers show that modern Test batting has undeniably become less productive. What they do not show nearly as clearly is that modern batters have become proportionately less skilful.
Across a rolling sample of 100 Tests between 2018 and 2020, the overall batting average fell to 27.65, the lowest for any such period in roughly six decades. Remove the tail and the decline still looks enormous. Batters from Nos. 1 to 7 averaged 41.41 around the 2008-10 peak, but only 32.87 in a comparable 2017-19 window, a fall of more than 20%.
Where those runs disappeared is even more revealing. Openers averaged 37.17 during the 2000s, 35.16 in the 2010s and 32.87 in the 2020s. No.3 batters have fallen even further, from 43.41 in the 2000s to 39.81 in the following decade and 34.96 in the 2020s.
Further down the order, the decline becomes smaller. No.4 has dropped from 44.10 to 38.93 between the 2000s and 2020s, while No.5 has moved from 40.95 to 38.32. The closer a batter is to the new ball, the more severe the statistical damage has been. That is an important clue. A universal deterioration in batting ability should theoretically hurt the entire order more evenly. Instead, the players repeatedly exposed to fresh pace, lateral movement and difficult early conditions have suffered most.
The easiest explanation is that modern batters simply do not possess the patience of their predecessors. T20 cricket has supposedly taught them to attack everything and removed the ability to occupy the crease. Yet the strike-rate numbers make that argument difficult to sustain.
Top-order Test batters scored at 50.14 runs per 100 balls in the 2000s. That figure was 49.81 during the 2010s and stands at 51.21 in the 2020s. Despite the enormous drop in averages, scoring tempo has barely changed.
If batters were regularly getting themselves out because they were trying to play Test cricket like T20, the obvious expectation would be rapidly rising strike rates alongside falling averages. Instead, today’s top orders are scoring at almost exactly the same speed as those of 15 or 20 years ago. They are simply surviving for fewer deliveries. The 2024 Test year captured that contradiction particularly well. Around 1,820 wickets fell at roughly 29 runs apiece, while the overall run rate touched approximately 3.65. Test cricket was simultaneously producing exceptionally fast scoring and exceptionally frequent dismissals.
The World Test Championship era has intensified the shift towards decisive cricket. Across the six years immediately before the WTC, Test cricket produced around 32.06 runs per wicket, with a dismissal every 59.5 balls. In the WTC period, those numbers fell to approximately 30.31 runs per wicket and one dismissal every 54.7 deliveries.
The 2023-25 cycle went considerably further. Batters averaged only around 29.11 per wicket, while a dismissal arrived every 47.8 balls. The average Test lasted approximately 1,638 deliveries, compared with roughly 1,850 during the previous two championship cycles.
Of the 69 Tests played in the 2023-25 WTC, 50 had produced a result by the end of Day 4. Only four matches were drawn, and each of those was affected by rain. The points system cannot be ignored either. Victory is worth 12 WTC points while a draw earns four, giving stronger home teams a clear reason to pursue decisive surfaces. The batting decline predates the championship, so the WTC cannot be blamed for starting it, but it has arrived alongside an increasingly result-driven version of Test cricket.
Pace bowling has been central to the change. During the major batting downturn around 2018-20, fast bowlers were averaging close to 26, among their strongest sustained returns in decades. There has also been a change in how the ball moves. Tracking data from England showed conventional swing declining during the late 2010s while seam movement increased. That is particularly dangerous because seam deviation occurs after pitching, leaving a batter considerably less time to react.
The rise of wobble-seam bowling has made that uncertainty even greater, while modern attacks are increasingly difficult to escape. In Australia during the 2010s, opposition new-ball pacers averaged around 32.15 while the rest of the fast bowlers averaged 37.88. Since 2020, those numbers have been almost identical at approximately 30.27 and 30.72. Surviving the opening pair once carried the promise of easier runs. Against many modern attacks, the third seamer is barely less dangerous than the first.
The same decline is visible in Australia’s domestic system. Sheffield Shield batters averaged 32.25 in 2014-15. That number had fallen to 25.87 in 2023-24 and 26.01 in 2024-25, a reduction of roughly 19% from the earlier mark. Yet strike rates barely moved. The competition’s batting strike rate was 50.71 in 2014-15 and 49.40 in 2024-25. Once again, there is little evidence of a dramatic shift towards recklessness; batters are simply being dismissed far more often.
By 2025-26 the situation had become pronounced enough for Cricket Australia to rethink its pitch philosophy after nearly one-third of regular-season Shield matches ended inside three days. If even the domestic development system has tilted heavily enough towards bowlers to require intervention, batting averages cannot be read purely as a measure of declining player quality.
There is still evidence that certain red-ball skills have weakened. Rahul Dravid had warned years ago about players arriving for A tours after six or seven months without red-ball cricket, something far less common in previous generations.
Franchise leagues, compressed international calendars and reduced first-class preparation give modern players less uninterrupted time to develop techniques specifically designed around batting for five hours rather than 50 balls. India’s numbers against spin are particularly striking. Indian batters averaged about 47.27 against spin at home between 2015 and 2019. Since 2020, that figure has fallen to roughly 31.39. South Africa’s overall batting average similarly dropped from approximately 39.38 during 2004-13 to 28.83 across the following decade.
Some teams have genuinely lost exceptional generations, while defensive technique and long-form preparation cannot simply be dismissed as irrelevant.
There is one final distortion in the debate: the generation being used as the standard played during an extraordinary batting boom.
During the 2000s, 21 regular top-seven batters averaged 50 or more. In the 1990s, there were only five. The age of Tendulkar, Ponting, Dravid, Kallis and Sangakkara did not merely contain exceptional individuals; it existed in an unusually productive period for Test batting generally. That is why an average cannot measure batting ability independently of its environment. Modern players are clearly producing fewer runs, and some of that decline almost certainly reflects reduced red-ball preparation and weaker survival skills. But the scale and location of the statistical fall suggest that conditions and bowling have moved just as significantly.
Today’s Test batter is getting less time against the red ball, facing deeper attacks and operating in matches in which wickets are falling more frequently than they have for decades. That does not make an average of 35 automatically equal to 45 from another era, but it does make the comparison far less straightforward than the scorecard suggests.
