THREE INSIGHTS INTO THE BAZBALL ASHES USING CONTROL: A LONG READ

Australia retained the Ashes in England for only the third time in this century, repeating their successful defense in 2019. The last two Ashes have come following a run of four consecutive trips to England in 2005, 2009, 2013 and 2015 at the end of which England held the Ashes. It is a remarkable achievement for a great Australian side against an English side which showcased its customary depth and quality with bat and ball in home conditions. This article offers a review of the 2023 Ashes via three insights offered by ESPNCricinfo’s elegant control measurement.

In his excellent overview of the trade-offs involved in Bazball, S. Rajesh finds that Bazball - sacrificing control in pursuit of quick runs - has worked for England against most opponents. Control records the essential fact about the essential encounter in cricket. It answers the question - was the batter in control of the bowler’s delivery? Note that control classifies deliveries. This is a different approach to the expected Wickets or expected Runs approach you may have seen on Test match broadcasts, which are regressions.

The approach known as Bazball involves batters being less selective about the deliveries they choose to attack, and therefore, playing attacking shots more frequently in the pursuit of quicker runs. Such batting trades more frequent false shots for quicker runs. The Control record can be studied using three measurements:

1. Balls per NIC: This shows how often a false shot is played.
2. Runs per NIC: This shows runs scored per false shot.
3. NIC per wicket: This shows the false shots per wicket.

We should expect that Balls per NIC for England will be smaller than it is for the opposition, Runs per NIC will be larger. We should also expect that NIC per wicket will be larger since, by attacking more, a team forces the opposition to defend runs and attack less thereby making it less likely that the average false shot will result in a dismissal.

A summary of Test series involving England since the restart of Test cricket in July 2020 after the covid pandemic is given in Table 1. The record is broadly divided into the Bazball era, starting on June 1, 2022, and the pre-bazball era from July 2020 to May 2022. Note that figures for England’s Test tour of West Indies towards the end of the 2022-23 season are unavailable.

In its broad outline, the control record suggests that the intuitions of Bazball are well founded. Before the advent of Bazball, in series where England produced more runs per false shot, they also typically played false shots less often than their opponents. In the Bazball era, England have played false shots 17% more often than their opponents and scored 18% more runs per false shot than their opponent. Their preference for the attacking shot has earned them about 8% more false shots per dismissal than their opponents. If England’s average opponent has lost 10 wickets in 95 false shots, England have survived 103.

In every series in the Bazball era, barring the Ashes, England produced 19-30% more runs per false shot than their opponent. In the Ashes, Australia managed to keep this English advantage down to 5.8%. Over the course of the series, England’s batters scored 474 runs from false shots, while Australia’s batters managed 403. Australia’s bowlers also produced false shots 23% more often than England’s bowlers did.

One reason why England failed to regain the Ashes was that Bazball didn’t work for them in this series as it had in the 2022 summer and in New Zealand and Pakistan in 2022-23. In the 2023 Ashes, England traded 6% extra runs per false shot in exchange for 23% more frequent false shots.

Why did this happen? The best answer lies in the quality of the Australian bowling despite the fact that Australia lost their experienced off-spinning ace Nathan Lyon midway through the second Test. A summary of the record of the two bowling attacks is given in Table 2 below. The effect of England’s shot making approach is evident in the fact that Australia’s fast bowlers induced a false shot every 3.9 balls and conceded 441 runs from those 807 false shots (or 3.3 runs per over). England’s fast bowlers beat the bat less often (once every 5.1 balls), and conceded 336 runs from those 841 false shots (or 2.4 runs per over). 

Pat Cummins induced more false shots (256 in 952 balls bowled) more often (1 every 3.7 balls) than any bowler on either side (except Mitchell Starc who induced 247 in 769 balls), and earned a wicket once every 14.2 false shots. Cummins’ misfortune did not extend to the other Australian fast bowlers. Mitchell Starc managed a wicket every 10.7 false shots, and Josh Hazlewood managed one every 8.9 false shots.

For England, Stuart Broad, Chris Woakes and Mark Wood were the stand out wicket taking threats. James Anderson and Ollie Robinson induced false shots once every 6.6 and 6.3 balls respectively, far below their recent record in home Tests of inducing false shots once every 4.3 balls. Mark Wood (a false shot every 3.9 balls) and Chris Woakes (a false shot every 4.5 balls) were arguably England’s two best bowlers. Woakes got a wicket once every 7.9 false shots, better than any other specialist bowler on either side.

On the fast bowling side of things, the ineffectiveness of Anderson was a problem for England. He bowled more balls in the series than any bowler other than Cummins and Broad. Australia’s spinners induced false shots more often (once every 6.1 balls) than Anderson.

Lyon’s absence meant that Australia fielded a spinner in four out of five Tests, and bowled spin for less than 10% of their deliveries in an Ashes series for the first time since 1989. Given that control figures show that Lyon conceded only 11 runs off the 58 balls on which he induced false shorts, his injury also reduced Australia’s ability to control proceedings in the field. Todd Murphy filled in admirably, but could not be reasonably expected to make up for Lyon’s 122 Tests at 496 Test wickets worth of experience.

These figures against pace and spin point to an interesting difficulty of Bazball. It is one thing to throw the bat at the quick bowler. Edges tend to fly. Against the spinner, the batter has to generate all the power. Mishits tend not to go very far. If England intend to employ the Bazball approach in India when they tour in early 2024, they will have to think about this spin-specific fact. It is very unlikely that the pitches in India will be as inhospitable to spin bowling as the pitches on which the 2023 Ashes were played, and the Indian spinners will induce false shots significantly more frequently than one every six balls.

When compared with the historical control record in England, the evidence suggests that the conditions (the pitches and the ball, about which, more in the concluding part of this article) in the 2023 Ashes were also less hospitable to seam bowling than has been the norm in England in recent years. From 2018 to 2021, English fast bowlers induced a false shot once every 4.4 balls in Tests in England

¹

. In the 2023 Ashes, this increased to one every 5.1 balls. 

Whether it is by design or by accident, the Bazball era has coincided with a period of easier batting conditions in England (specifically, with an era of lesser seam movement). It has allowed English batters to come into their own for the first time since 2016 and 2017. But this has also negated England’s greatest strength - their unrivaled seam bowling depth - to a significant extent. Given that bowlers win Test matches since 20 wickets are required to win them, the extent to which Bazball makes England more efficient at competing in Test matches is debatable.

So far, the evidence shows that England’s Bazball approach involves trading more frequent false shots in exchange for quicker runs, and that this has earned England about 8% extra false shots per dismissal during the Bazball era. In the 2023 Ashes, England’s batters lost 10 wickets in 109 false shots, while the Australians lost 10 wickets in 103. Keeping this basic trade-off evident in Bazball in mind, the innings summary (see Table 3 below) points to the direction in which conditions moved as each Test progressed. 

In the first Test at Edgbaston, batting was easier for both teams in the first innings. This was also the case at Lord’s where Australia extended their lead cautiously in the third innings as the conditions became less batting friendly. They scored only 2.7 runs per false shot compared to their 3.8 in the first innings, and batted for 7 balls more than they did in the first, for 8 fewer false shots, and 137 fewer runs.

At Leeds, batting got easier as the match progressed, and England traded more frequent false shots for quicker scoring in the 4th innings. Batting also got easier at Old Trafford, where England rode their luck on what was arguably the finest batting day of the series. Zak Crawley scored a remarkable 189 in 182 balls, which ended with his 53rd false shot. Of the 1770 Test innings for which I have control records, only 8 lasted as long as at least 53 false shots. None of those eight innings lasted fewer than 226 balls. For comparison, Usman Khawaja’s 141 at Edgbaston involved only 27 false shots in 321 balls faced on a similarly batting friendly pitch.

The record at The Oval is most intriguing. The ball was changed after going out of shape after 36.1 overs in the fourth innings of the match. The replacement ball offered significantly greater help to the bowlers leading to speculation that the replacement ball might have been from the 2018-19 batch. 

The control figures suggest that Usman Khawaja’s perception that the replacement ball was doing more than any other in the series was probably well founded. Batting had gotten easier as the match progressed. England had managed 3.7 runs per false shot in the third innings of the match, their third best return in the series after their 393/8 in the 1st innings of the first Test and their 592 in the 1st innings at Old Trafford. The original ball in the 4th innings produced 21 false shots in 217 balls. The changed ball produced 96 false shots in 351 balls - one every 3.7 balls. 

Australia played false shots more frequently in the last 351 balls of the 4th innings at the Oval than in any other innings in the series. Even in the most difficult batting conditions of the series in the 1st innings at Headingley, Australia managed a false shot every 3.8 balls after being put in. The replacement ball helped England significantly.

After a series as exciting as the 2023 Ashes, it is tempting to wonder about what-ifs. What if it hadn’t rained at Old Trafford? What if the replacement ball had not been as lethal as it proved to be? What if Nathan Lyon had not been injured midway through the 2nd Test? What if Chris Woakes and Mark Wood had played all five Tests? What if Pat Cummins had the same luck that every other bowler on both sides had, and managed 24 wickets in 256 false shots instead of just 18? These are all tantalizing what-ifs.

But for me, the most tantalizing what-if offered up is: what if every ball used in the 2023 Ashes had been like that rogue ball used for the last 351 balls in the 4th innings of the fifth Test? Would England’s seam attack have been much harder to play for Australia’s batters? Would Australia’s fast men have been even more effective? Who would have won a seamers series? My money would have been on England to win a seamers series. Bazball unquestionably produces more runs quicker, but England’s success in the Bazball era has been due to the depth, quality and variety of their bowlers.

As it happened, Australia retained the Ashes because their three fast bowlers had enough quality to limit the extent to which Bazball produced quick runs in large enough quantities for England. In other words, Australia’s bowlers held their own in their battle against England’s bowlers. They played the percentages expertly, even when it looked ugly as it assuredly did when Australia bowled with the new ball with sweepers on the square boundary either side of the wicket. Australia were not the better side in the 2023 Ashes, especially after losing Nathan Lyon. But they weren’t worse than England either.

1. Based on the control figures ESPNCricinfo have shared with me over the years, when some out of Anderson, Broad, Archer, Robinson, Wood, Overton, Curran, Woakes, Potts, Tongue and Stone played for England in a Test in England from 2018-21, they induced 3568 false shots in 15796 legal balls, or 4.4 balls per false shot.