What's luck got to do with it: a control review of the World Cup

Travis Head faced Jasprit Bumrah on the fourth ball of the third over in the second innings of the 2023 ODI World Cup final. Bumrah was bowling around the wicket to him. The delivery was aimed at the off stump from wide of the crease and moved away. Head, according to the ball-by-ball commentary on this site, "stays leg side of the ball and almost nicks it off". Head was beaten on the inside or outside edge in this manner on the seventh, ninth, tenth, 18th, 19th, 22nd, 26th, 28th, 29th and 30th balls he faced. This included him inside-edging past leg stump, playing and missing outside off stump, and being beaten on the inside edge as he fell over. Before scoring 25, Head was beaten multiple times in nearly every single one of the ways it is possible for a batter to be beaten by a bowler in cricket. He played a false shot to ten of his first 30 balls.

He went on to score a brilliant 137 off 120 balls, playing only ten false shots in his last 90 balls. When India batted earlier that day, Rohit Sharma played five false shots in his first 30 balls. He played his sixth to his 31st and was dismissed for 47. A few balls later, Shreyas Iyer stayed leg side of the ball and nicked off against Pat Cummins instead of merely playing and missing.

Head's survival and Iyer's dismissal were not by the batter's design. No player's survival of a false shot is by design. It is a matter of that dreaded thing - luck. In cricket, luck is the accumulation of favourable outcomes for a set of actions to an extent that is significantly different from the average expected outcomes for that set of actions. It is only by accounting for luck that distinctions in skills can be located.

In addition to the three traditional measurements in cricket - runs, balls and wickets - ESPNcricinfo's control measurement records whether or not the batter was in control of the delivery. Control makes three measurements possible:

1. The frequency of a false shot (balls per false shot)
2. How often a false shot results in a dismissal (false shots per dismissal)
3. Runs per false shot

These measurements help us locate luck and skill (or efficiency). For instance, if Head had played only four false shots in his first 30 balls, to dismiss him, India would have needed one in four false shots to go their way. As it happened, they needed only one in ten to go their way in those first 30 balls. That is to say, they forced a false shot every third ball on average. In the remaining 90 balls, they created comparatively little jeopardy. In that early part of the innings, Head was lucky. But after his first 25 or so runs, India would have needed to be lucky to dismiss him.

The control measurement is the hinge of the cricketing contest. It helps to measure how much jeopardy the bowlers create, how much risk the batter has assumed, how lucky the bowlers are, and how efficient the batter is. Before we use it to understand the 2023 World Cup and how its patterns were different from those of the 2019 tournament, it is worth reflecting on the patterns of the control measurement.

A survey of the control record since January 1, 2016 shows that a dismissal occurs every 10.1 false shots in Tests, 7.7 false shots in ODIs, and 5.6 false shots in T20Is. A false shot occurs once every 5.6 balls in Tests, once every 4.8 balls in ODIs, and once every 3.3 balls in T20Is. Quicker scoring involves more frequent risks and also bigger risks (the average false shot is more likely to result in a dismissal in a T20 than in a Test). Within each format, the evidence suggests three things.

First, a batter's capacity to avoid false shots (Balls Per False Shot) and to survive them (False Shots Per Dismissal) depends on skill. The table below gives the rates for each by batting position in Tests and ODIs.

Second, the evidence from the control record for Bazball suggests that while Bazball-era England played false shots more frequently than their opponents, they were dismissed off a false shot at about the same rate as their opponents.

In ODIs, over the last two World Cups, Kane Williamson survived 19.2 false shots per dismissal, and averaged 92.7 (counting only dismissals credited to bowlers) for his 834 runs. The second-best survival rate belongs to Andile Phehlukwayo, who survived 17.3 false shots per dismissal. From third to sixth on this list are Carlos Brathwaite, Nicholas Pooran, Shikhar Dhawan and Shakib Al Hasan. While they are all very good players (and none are tailenders), they do not come readily to mind as being renowned for their batting technique. The four players who have scored more runs than Williamson in the last two World Cups have survived 11.0 (Rohit Sharma), 10.9 (David Warner), 11.3 (Virat Kohli), and 10.3 (Quinton de Kock) false shots per dismissal respectively. The false-shot statistics show producing big scores requires more luck than producing quick scores does.

Third, as seen in the table below, when the ball is hitting the stumps, false shots are produced less often than when the ball isn't hitting the stumps. But when the ball is hitting the stumps, false shots result in dismissals significantly more often than they do when the ball is not hitting the stumps. The significance of this pattern is evident in the comparison of the last two World Cups later in this article. The stumps were in play more often in Indian conditions than they were in England. This especially influenced the effectiveness of spinners.

The Balls Per False Shot figure represents the extent to which batters are prepared to take chances, and the extent to which the wicket is assisting bowlers. Up to a point, the False Shots Per Dismissal represents the extent of luck a batter enjoys, at least in the short term. But beyond that point, as we will see, it indicates something about the approach taken by a batter or bowler. The Runs Per False Shot measure can be understood as a summary figure for control that indicates the efficiency of the batter's approach. A bowler's goal is to minimise efficiency, while the batter's goal is to maximise it.

For example, in 2018, Virat Kohli made 1322 Test runs in 13 Tests. This was a year when bowlers were at their most lethal in Tests since the 1950s. Kohli played a false shot every 5.6 balls in 2018, but survived 16.5 false shots per dismissal. Through 2020 and 2021, he made 652 runs in 14 Tests against most of the same bowlers. Over those Tests, he played a false shot every 7.4 balls, but only survived 8.3 false shots per dismissal. His runs-per-false-shot figure for the 13 Tests in 2018 stood at 3.1, compared to 3.3 for the 14 Tests in 2020 and 2021. Kohli was luckier in 2018 than he was in 2020 and 2021, not more skillful.

In the last two World Cups, the number of balls per false shot ranged from 4.0 to 4.9 (in 2019 in England), and 4.7 to 5.5 (in 2023 in India) by ground, with one exception. The MCA ground in Pune was especially batting-friendly, producing 6.5 balls per false shot. The dismissal rates in 2019 in England ranged from 7.4 false shots per dismissal (at The Oval) to 10.1 balls (at Edgbaston). In 2023, they ranged from 6.4 false shots per dismissal in Pune to 8.5 false shots per dismissal in Chennai.

England in the 2019 and 2023 World Cups
England's batting in the last two World Cups is summarised in the table below. In 2023, their batters averaged 14 runs fewer per dismissal and scored about seven runs fewer per 100 balls faced than they did in 2019, though they played false shots at about the same rate. The tournament in India was, as expected, friendlier to spin (and featured more of it) than 2019. While England's batters were less effective against pace in 2023 compared to 2019, it was against spin that they suffered greatly. Though they played false shots less often, they lost a wicket to every fourth false shot to spin. In 2019 England's marauding batters smashed spinners, averaging about 80 and scoring at nearly seven runs per over against them. In 2023, they managed just over five runs per over against spin and averaged 24.5. In 2019, England lost a wicket to spin every 71 balls. In 2019, they lost one every 29 balls.

Compared to 2019, England's returns against pace in 2023 were close enough to suggest that with slightly better luck, they would have equalled their 2019 showing. But against spin, the drop in returns was spectacular, even if we say the 2019 performance owed to above-average fortune. The control record, combined with the pattern of dismissals, suggests that England's preferred method of slog sweeps, reverse sweeps, conventional sweeps and other cross-batted strokes did not work as well in India as they did in England in 2019.

As if England's problems against spin weren't damaging enough, their fast bowlers were not nearly as effective in India in 2023 as they were in England in 2019. Then, they were led by Jofra Archer and the experienced Liam Plunkett. In Indian conditions, the control record shows, the skill of England's batting against spin, and the skill of England's fast men, was not what it had been in England in 2019. A 6-3 win-loss performance in the league stage 2019 was turned on its head to 3-6 in 2023.

The other teams in the last two World Cups
England's example provides some indication that false shots resulted in dismissals more frequently in 2023 than they did in 2019. The stumps are in play more frequently in the subcontinent. In the 2011 World Cup, 26% of all wickets were bowled or lbw. This figure dropped to 14.5% in Australia and New Zealand in 2015, and 17.6% in England in 2019. In 2023, it went up again, to 20.5%.

Now for a look at the other teams. The bowling and batting figures in the section following are presented in terms of the following tuple: Runs Per False Shot, False Shots Per Dismissal, Runs Per Over (unless stated otherwise).

Australia: They made the last four in both tournaments. They were marginally luckier with the bat in 2023 than in 2019. Their spin attack was significantly more effective in 2023 (5.8, 5.4, 5.1) than it was in 2019 (6.0, 11.6, 6.3), while their magnificent pace attack held its own: 3.5, 7.6, 5.6 in 2019, and 3.8, 7.8, 5.9 in 2023, though the conditions changed from England to India.

On the batting side, they scored at nearly seven an over against seam bowling (4.6, 7.0, 6.8) after managing a run a ball in England in 2019 (3.8, 7.8, 5.9). Compared to 2019, when their top four slots were occupied by Usman Khawaja, Steve Smith, David Warner and Aaron Finch, their approach in 2023 favoured power over technique, with Mitchell Marsh and Travis Head earning promotions and Smith and Marnus Labuschagne dropping down the order on occasion.

South Africa: They improved significantly in 2023 over 2019, mostly by becoming more direct and attacking. The pace attack was more penetrative in 2023 (3.3, 8.1, 6.1) compared to 2019 (3.5, 8.7, 5.4), while their spin bowling, led by Keshav Maharaj, was significantly more effective in 2023 (3.7, 7.2, 4.6) than it had been in 2019 (6.1, 7.4, 5.5). The South Africa quick bowlers appear to have been slightly unlucky, but this could be due to their inexperience relative to, say, the Australian and Indian attacks, who were able to use the conditions slightly better.

South Africa's batting in 2023 (4.7, 8.1, 6.8) was, if anything, even more attacking than the Australian line-up. Temba Bavuma's form (and fitness) deserted him during the World Cup. Had he continued his magnificent run of 2022 and 2023 through this tournament, South Africa might well have won.

New Zealand: Unlike Australia, New Zealand's fast bowlers were not able to do as well in 2023 (4.6, 7.3, 6.2) as they did in 2019 (3.5, 8.2, 5.1), partly due to injuries to Tim Southee and Matt Henry. They didn't have a second spinner who could reliably support their spin-bowling mainstay in both tournaments - Mitchell Santner. New Zealand bowled only 57 balls of spin per 300 balls in 2019; in 2023, they bowled 134 balls of spin per 300 balls. With their fast bowling being less effective, and their spin resources stretched, they were not as strong in the field in 2023.

They survived in the tournament thanks to their batters - Rachin Ravindra, Devon Conway, Daryll Mitchell, Glenn Phillips, and (when he was available) Kane Williamson - who produced cutting-edge efficiency against pace (4.6, 8.2, 6.7) and spin (6.7, 8.3, 6.2).

India: They used spin more than almost every other side in the 2019 tournament (118 balls out of 300). In 2023 this rose to 131 balls of spin out of 300. Their spinners were also significantly more effective, improving from the already high bar of 2019 (4.2, 10.7, 5.4) to (3.8, 7.1, 4.4). The Indian fast bowlers' preference for attacking the stumps was more effective in 2023 (3.4, 6.1, 5.3) than it had been in 2019 (2.9, 8.1, 5.2).

On the batting side, India's approach of using Virat Kohli as an accumulator in the top four sandwiched between three power players at Nos. 1, 2 and 4, made their batting effective. They were ruthless against spin (10.2, 10.1, 5.8) and prolific against pace (7.1, 5.9, 6.8). It was India's finest World Cup campaign by their best team yet.

Pakistan: The remarkable thing about Pakistan in the 2023 World Cup was the weakness in their bowling. After losing Naseem Shah to injury, their seam bowling was depleted and conceded more than six runs an over (4.1, 7.1, 6.3). More surprising was the weakness of their spin attack (8.3, 9.5, 6.3). Pakistan's spinners did worse in India than they had in England in 2019 (6.1, 8.7, 5.5).

Their batting in 2023 was effective against seam (5.2, 6.3, 6.5) and competent against spin (6.2, 6.5, 5.7), though they were not explosive like the line-ups of India, Australia or South Africa.

Sri Lanka: Like Pakistan, Sri Lanka were able to exert very little control with pace (5.0, 6.8, 6.8) and spin (8.2, 9.5, 6.1). Beyond that, their batters were held in check by opposition spinners throughout the tournament (6.0, 5.2, 4.9). They fielded a relatively young team and should be better placed in the 2027 edition.

Bangladesh: Their spinners (9.8, 4.3, 5.8) were the disappointment of the tournament. Before the World Cup began, the depth and experience of their spin attack had many observers tipping them as contenders for the last four.

To compound matters, Bangladesh's batters struggled to score against spin as well (5.1, 7.6, 4.4). These two factors made their tournament significantly more difficult than it was expected to be. Perhaps the fact that Bangladesh had not played an ODI in India between 2006 and this World Cup might be considered a mitigating factor.

Afghanistan: Unlike Bangladesh's spinners, Afghanistan's slow bowlers were superb in the 2023 tournament (4.4, 7.8, 4.9). Ultimately, Afghanistan's underpowered batting line-up told, but their spinners made them competitive throughout the tournament. Afghanistan also bowled more spin (201 out of 300 balls) than any other team.

On the batting side, much of the reason for Afghanistan being underpowered was their limited scoring ability against spin (5.6, 6.4, 4.7).

Netherlands: Like Afghanistan, Netherlands' batting struggled to score against spin (3.8, 5.0, 4.4). Their spinners held their own (7.7, 5.2, 5.3) without being spectacular, but their fast bowlers struggled (5.7, 6.7, 6.7). They produced one of the great upsets in World Cup history when they beat South Africa in Dharamsala.

India were the outstanding team of the 2023 World Cup. Australia, led brilliantly by Cummins, were worthy winners. Cummins' figures in the Ashes and at the World Cup suggest he had an ordinary time in those series. His 2023 has been a lot like Kohli's 2020 and 2021. Australia's opponents should worry that Cummins will probably have a year like Kohli's 2023 before he is done. Even if he doesn't, 2023 will be remembered as the year in which he led Australia to the World Test Championship and ODI World Cup titles, in addition to retaining the Ashes in England.