Neil Wagner and Kyle Jamieson celebrate after dismissing Najmul Hossain Shanto

Neil Wagner and Kyle Jamieson are both in the top ten among seamers who have taken the most wickets coming on first-change or later in Tests since 2018

© Getty Images

The age of bowling

Stunning strike rates, incredible depth in attacks, battered batters: a deep dive into an unparalleled era of bowler domination in Test cricket

S Rajesh (with inputs from Shiva Jayaraman)  |  

What is common to Pat Cummins, R Ashwin, Kagiso Rabada, Tim Southee, Jasprit Bumrah, James Anderson, Stuart Broad and Mohammed Shami? They are all terrific bowlers, of course, but here is something a little more specific: they have all taken 100-plus Test wickets at sub-25 averages in the four-year period since the start of 2018. Read on to find out how rare this is in Test history.

One of the (several) fascinating aspects of cricket - Test cricket, especially - is the enduring tussle for supremacy between bat and ball. Over the last few decades there has tended to be a cyclical pattern to which skill holds the upper hand. Through much of the 1990s bowlers held sway. There was quality and depth in bowling talent in most teams: Curtly Ambrose, Courtney Walsh, Wasim Akram, Waqar Younis, Allan Donald, Shaun Pollock, Glenn McGrath, Shane Warne, Muttiah Muralitharan, Anil Kumble, Darren Gough and Heath Streak were all in their pomp, and batting against them wasn't much fun.

Then came the decade of the 2000s, when the tables turned and batters got an opportunity to make up for lost time. The pitches generally became flatter the world over, many of the bowlers mentioned above either retired or were past their prime, and their replacements weren't as experienced or as good. The batters weren't complaining, of course: between 2003 and 2010 the bowling average soared to 35.20 runs per wicket, compared to 30.71 between 1995 and 2000, an increase of nearly 15%.

Parity was restored somewhat through the first half of the 2010s - bowlers averaged 33.25 between 2011 and 2015 - but even that will not have prepared anyone for what we have witnessed more recently, in the last four years. Pitches and conditions all around the world have become spicier - more swing in England, more bounce, pace and seam in Australia and South Africa, more turn and uneven bounce in Asia - while high-class bowling talents have emerged in all the top teams. Simultaneously, there is also an argument that, despite the presence of a few top-class batters, defensive techniques have generally declined due to an excess of 20-over cricket. The result is a stunning period of bowling domination, the likes of which we haven't seen in a long, long time. Sample these stats:

- The bowling average since 2018 is the best in any four-year period in the last 60 years

- The bowling strike rate has never been better in any four consecutive calendar years in which at least 50 Tests were played

- As many as 19 bowlers have taken 50-plus wickets at averages under 25 in this period

- Change seamers have averaged just 27.85 during this time, almost as low as new-ball bowlers

- Only five batters have managed 1500-plus runs over these four years at a 50-plus average

The four-year blocks comparison
In 2018 the bowling average for the year was 27.37, the lowest since 1960. Since the start of 2021, it has been 28.57, while in the two intervening years - 2019 and 2020 - it barely exceeded 30.

Put that all together and it means that since the start of 2018, bowlers have collectively averaged 28.87 in 159 Tests. That figure is almost 20% better than that for the 2007-10 block, only about a decade before, when batters were at their dominant best: in that four-year period, the bowling average was 35.67 from 162 matches, the highest it has ever been in four successive calendar years.

To find the previous four-year period (calendar years) when bowlers had a better average, you would have to go back all the way to 1956-59, when they picked up a wicket every 26.85 runs. The top seven wicket-takers in that period - Richie Benaud, Jim Laker, Tony Lock, Fred Trueman, Alan Davidson, Brian Statham and Fazal Mahmood - all conceded fewer than 23 runs per wicket.

The low average also coincides with the fact that scoring rates are generally faster now than they were in the past. That means the bowling strike rate is terrific too. In fact, the strike rate of 57.5 balls per wicket since the beginning of 2018 is the best ever in a four-year period in which at least 50 Tests have been played. In the 1956-59 period, when the average was 26.85, bowlers only struck every 74 balls; the economy rate of 2.17 was much lower than the current 3.01.

The 150-Test blocks comparison
While the bowling average was superb in that late 1950s period, only 64 Tests were played in those four years, compared to 159 since 2018. So why not compare the averages over 150-Test blocks? That will provide the answer to the question: "When was the last time bowlers were as dominant over so many matches?" (To briefly explain the 150-Test rolling blocks, the first one is from Test No. 1 to No. 150, the second from Test No. 2 to No. 151, the third from Test No. 3 to No. 152, and so on. The last one is from Test No. 2300 to 2449, the most recent Test. There are a total of 2300 overlapping blocks.)

As you might expect, the answer once again goes back to the 1950s: in the 150-Test period between July 1952 and January 1961, the bowling average was 28.82. Thereafter, the average was over 29 for each set of 150 matches until the spell was finally broken in matches spanning December 26, 2017 to December 4, 2021: in the 150 Tests in this period, the average dropped to 28.96. In fact, even if the cut-off is lifted to 30, there is still a massive gap of nearly 40 years, between 1977 and 2016.

Of the 2300 blocks of 150 Tests, in only 71 is the bowling average better than in the best one from the 2018 era - between February 8, 2018 and January 3, 2022 - when the average was 28.78. At the other end of the spectrum (the blocks where the bowling averages were the poorest), the 19 blocks that make up the bottom are all those that started in 2007. That shows how far the pendulum has swung in less than a decade.

For all the wealth of bowling talent that was available in the 1990s, there wasn't a single 150-Test block with a sub-30 average in that decade. That is not a rap on the quality of bowlers, of course - factors like quality of batting, pitches, and overall conditions come into play as well. What it does tell you, though, is that the current era of bowler domination is a pretty unusual one.

And the strike-rate numbers are even starker. Of the 2300 blocks of 150 Tests, there have only been 23 where the bowling strike rates have been below 58; 15 of those 23 have been since July 2017.

Before July 2017, the last time the strike rate dipped below 58 was way, way back, between March 1882 and August 1924. From start date to start date, that is a gap of more than 135 years.

Where are the weak bowling attacks?
As mentioned earlier, conditions and bowling attacks have become challenging the world over. That is seen in the bowling averages of all teams in the last four years. Of the nine sides that have played at least 20 Tests during this period, seven have bowling averages of under 30; Sri Lanka and Bangladesh are the only teams that don't make the cut, and even they average under 36.

Girish TS / © ESPNcricinfo Ltd

A comparison with the team-wise numbers from 2007-10 illustrates the stark difference between the two periods. Between 2007 and 2010, only one team averaged below 30, and just barely: South Africa's bowlers conceded 29.66 runs per wicket. Just three teams averaged below 35, while two topped 40.

India's bowlers averaged 36.6 runs per wicket in the 2007-10 period; since 2018 that average has dropped to 23.64, a stunning improvement of 35.4%. Similarly, West Indies have improved from 43.81 to 29.59, and Pakistan from nearly 39 to under 29. Barring three teams, all others have improved their averages by more than 15%, with four teams doing so by over 25%.

The eight bowlers mentioned at the beginning of this article have been at the forefront of this revival, and their numbers again highlight how unusual these times are for Test cricket: there has never previously been a four-year period where as many as eight bowlers have taken 100-plus wickets at sub-25 averages. There were seven each in 2017 and 2016, and excluding years that overlap with the current four-year span, the previous best was six, in 1981, 1993, 1995 and 1997.

Since 2018, 11 bowlers have taken 100-plus wickets, which means a whopping 73% of these bowlers average under 25. Contrast that with the 2007-10 period, where too 11 bowlers took 100-plus wickets; only one of them, though, averaged under 25 - Dale Steyn. His average was an astonishing 21.29, more than six runs clear of the next best in the group; that one stat should end any debates on how good he was. In the current period, there are five bowlers in the sub-23 category, and the average of the tenth-ranked bowler is better than that of the second-best of the 2007-10 four-year period.

Girish TS / © ESPNcricinfo Ltd

And it isn't only these 11 bowlers who have stood out. With a 50-wicket cut-off, the list of bowlers averaging under 25 increases to 19. Topping that list is Kyle Jamieson (60 wickets at 17.3), with Jason Holder, Hasan Ali and Ishant Sharma all averaging under 22. In a non-overlapping four-year period, the next-best showing on this parameter is 11 bowlers, between 1997 and 2000.

Hunting in packs
If 19 bowlers are averaging under 25, and another seven between 25 and 30, then it's clear that a lot of teams have a lot of bowlers taking wickets cheaply. Only two of those 26 are spinners - the Indian duo of Ashwin and Ravindra Jadeja - which shows that fast bowlers are doing much of the damage. More on the pace-spin split in the next section, but a feature of these last four years has been the quality of the first- and second-change fast bowlers. Their ability to sustain the pressure means there is little respite for batters and little opportunity to relax and cash in once they have seen off the new-ball pair. That relentlessness makes batting far more demanding, leading to errors and wickets.

Neil Wagner, Cummins, Jamieson, Shami, Holder, Mark Wood - these are just some of the names who have made life difficult for batters in this period despite not usually taking the new ball. Since the beginning of 2018, seamers bowling first change or later in an innings have taken 1287 wickets at 27.85, compared to 2057 wickets by the new-ball fast bowlers at 25.69. Those numbers underline the effectiveness of the change seamers.

Two other parameters are worth looking at: the contribution of the change fast bowlers to the total number of wickets falling to pace, and to the total number of five-fors by fast bowlers. The first number illustrates their overall effectiveness; the second, their ability to stand up and take the lead role in instances where the new-ball pair haven't taken a lot of wickets.

On both counts, the current four-year period stands out. In terms of wicket percentage, the current value of 38.49 (1287 out of 3344) was last bettered in the period between 1904 and 1907 - that's about 115 years ago; only 18 Tests were played in that four-year span. Through most of the last 40 years, the percentage has stayed in the late 20s or early 30s.

Similarly, the five-for percentage of 33.86 was last bettered in the four-year span between 1929 and 1932, when only 43 Tests were played, compared to 159 since 2018. Through most of the last 40 years, that percentage value stayed in the late teens or early 20s.

These stats illustrate how unusual and special this era has been in terms of fast-bowling depth. The 1990s had some great fast bowlers, but most teams had a pair of them - the percentage of fast-bowler wickets by change bowlers in the 1995-98 span, for example, was 31.92, and the five-for percentage was 22.79.

Girish TS / © ESPNcricinfo Ltd

Fast bowlers rule
As we saw, in the last four years, seven out of eight bowlers who have taken 100-plus and averaged under 25 are fast bowlers. (The odd man out is Ashwin.) When the wickets cut-off is reduced to 50, as many as 17 of the 19 are fast bowlers.

Clearly, these are fun times for all bowlers, but certainly more so for quick bowlers. Since the start of 2018, they have averaged 26.52 and taken 66% of the total wickets to fall to bowlers, while spinners have averaged 33.32.

Girish TS / © ESPNcricinfo Ltd

Spinners have obviously done well in Asia, achieving sub-30 averages in Tests in India, Bangladesh and Sri Lanka, but in all other countries their average has exceeded 37, going all the way up to 64.08 in 16 Tests in New Zealand in this period. Fast bowlers, on the other hand, have averaged under 32 in six of the eight countries that have hosted at least ten Tests.

The 150-Test moving averages graph illustrates just how good these last few years have been for fast bowling. In the period between January 24, 2018 and December 26, 2021, seamers averaged 26.62 in 150 Tests; the last time there was a 150-Test sequence where the seamers did better (marginally) was between January 1952 and August 1960. And of a total of 2300 such 150-Test blocks, there are only 36 where the average was lower.

Spinners haven't done too badly in this period either, but their numbers pale before those of the fast bowlers. Their best 150-Test sequence since the start of 2018 came between February 2018 and January 2022, when they averaged 32.58. There are 317 blocks of 150 matches where spinners have averaged better than that, but only 11 of those have come after 1965. That still means the last few years have been among the better periods for spin in the last half-century, but spinners' performances don't quite compare with the chart-breaking numbers that the fast bowlers have compiled in this period.

No countries for batters
There used to be a time, not too long ago, when batters - even overseas ones - could look forward to scoring some runs in India. In that golden period for batters between 2007 and 2010, bowlers conceded 41.65 runs per wicket in India, and even the home team could only manage a bowling average of 36.43. Cut to the last four years and the stats have changed dramatically: the overall average in India has dropped to 27.54, while India's bowling average at home has halved to 18.79.

The story is similar in the West Indies, where the averages have dropped from 36.63 to 24.65, possibly helped by the use of the Dukes ball in the last few years. In Australia, Adelaide Oval used to be a batter's paradise in years gone by, but with day-night Tests becoming the norm there, it is bowlers who now look forward to bags full of wickets. Thirty-six all out, anyone?

In the last four years New Zealand is the only country in which the bowling average has exceeded 31; by comparison, between 2007 and 2010, there was no country where the bowling average was under 31.

The frequency of 400-plus team totals is another data point that illustrates how much tougher it has become for batters. In South Africa there have been only five such totals in 21 Tests since the start of 2018; between 2007 and 2010, there were 16 scores of 400 or more in the 20 Tests they hosted. For England and West Indies too, the ratios of matches per 400-plus total have more than doubled, as it has for India: from more than one 400-plus total per Test, it has now come down to one every two Tests.

Battered batters
Before we conclude, let us spare a thought for the tribe that has been at the receiving end of this phenomenon. Among the ten batters who have scored 2000-plus runs during this period, only two - Kane Williamson and Marnus Labuschagne - have 50-plus averages. Reduce the cut-off to 1500 runs and 25 batters qualify, but the percentage of those topping a 50 average remains at 20%, with only Babar Azam, Rohit Sharma, and Steven Smith (just about) adding themselves to the two mentioned earlier.

Girish TS / © ESPNcricinfo Ltd

For comparison, in each four-year block from 1999 to 2013 (that is, from 1999-2002, 2000-03, all the way to 2013-16), there were at least nine batters who made 2000-plus runs at 50-plus averages. In each of those 15 blocks, at least 21 batters made over 2000 runs, and at least 35% of those players averaged over 50.

The last time only two batters breached the 50-average mark with 2000-plus runs in a four-year block was in the 1992-95 period, when Steve Waugh (63.66) and Brian Lara (62.47) made the cut, and Michael Slater (49.92) narrowly missed it.

Analyse more batting metrics and they all point in the same direction. A century has been scored by a batter in the top six positions every 16 innings since the beginning of 2018, the poorest four-year rate since 1993-96. Teams have scored a 400-plus total every 2.15 Tests in this era, the worst rate since 1969-72. Comparing the numbers with the 2007-10 period - when batting numbers were at a high and teams were posting a 400-plus total virtually every match - tells you just how much Test cricket has changed in just over a decade. The bowlers are certainly not complaining.

Girish TS / © ESPNcricinfo Ltd

All numbers in the article are updated to January 31, 2022

S Rajesh is stats editor of ESPNcricinfo. Shiva Jayaraman is a senior stats analyst at ESPNcricinfo