Which bowler had the highest dot-ball percentage over their career?

During the first few months of 2026, I wrote a series of in-depth analytical articles covering a couple of very important topics. And, during the three months after, to get some much-needed relief for me and the readers, I concentrated on some fact, fun, and fantasy. All these articles were very well-received by the readers - especially the last one, imagining the possibility of no World Wars and the impact that would have had on the cricketers' careers. But now it is time to move back to my forte, which is pure analysis.

Recently, I realised that there was a treasure trove of data on wides, fours, and sixes, individually attributable to bowlers, available to me starting from the Pakistan-Australia series in the UAE in 2014. And for almost all the Tests since then, barring three. The batter-specific data on fours and sixes has been available from way back, but the bowler-related data has been available only for the past dozen years - for 486 Tests.

Once I integrated the data, I had to only decide on the types of analyses I needed to do and the tables I had to create. The tables are featured below. Readers should not forget that the analysis is limited to matches between 2014 and 2026. As such, the tables will cover the complete careers of only the recent bowlers, like Jasprit Bumrah, Josh Hazlewood, Kagiso Rabada, and will only cover the latter part of the careers of those who started their careers before 2014, like James Anderson, R Ashwin and Mitchell Starc.

Player level
In the period covered, Vernon Philander has the highest dot-ball percentage - just over 81%. That is nearly five dot balls every over. Colin de Grandhomme is very close behind. As is Mohammad Abbas. Anderson follows next. The selection criterion is 500 overs. A feature of the selection is that almost all the bowlers are the medium-pace variety. Ravindra Jadeja and Ashwin are the best spinners in this regard, clocking over 75%.

What about those legendary bowlers who could put the ball on the spot for days on end, like Trevor Goddard and Bapu Nadkarni? Goddard had an economy rate of 1.65 and Nadkarni was slightly behind, with 1.68. The one problem I have is that Goddard bowled a lot in South Africa, with its eight-ball overs policy, and some of his maiden overs were the eight-ball ones. And Nadkarni did the same in Australia. So I had to do a match-level analysis to get at some relevant facts relating to both.

Goddard bowled 11,736 balls and conceded 3226 runs. He bowled 706 maidens, out of which 181 were eight-ball maidens and the rest six-ball ones. That gives 4598 dot balls in the form of maidens. He conceded 3226 runs in the other 7138 balls. For want of any other ball-by-ball data, I have to assume that he would have needed around 1843 balls to concede these runs (3226/1.75 = 1843) in the form of mostly singles, and some twos, threes and boundaries. That leaves us with 5295 dot balls. The total number of dot balls comes to 9893. This gives him a dot-ball percentage of 84.3 (9893/11736). This is the best among all bowlers considered. It could even be higher, or perhaps a touch lower.

Nadkarni bowled 9165 balls and conceded 2559 runs. He bowled 665 maidens, out of which 18 were eight-ball ones. That gives 4026 dot balls in the form of maidens. He conceded 2559 runs in the other 5139 balls. For want of any other ball-by-ball data, I have to assume that he would have needed around 1462 balls to concede these runs (2559/1.75 = 1462). That leaves us with 3677 dot balls. The total number of dot balls comes to 7703. This gives him a dot-ball percentage of 84.0 (7703/9165). This is the second best among all bowlers considered.

At the other end, the table with low dot-ball percentages is dominated by spinners. Adil Rashid bowled scoreless balls only around 64% of the time - that means more than two balls every over were scored off. Dhananjay de Silva comes close behind, followed by Roston Chase. Shoaib Bashir is in fourth position. Then comes the only pace bowler in this group - Lahiru Kumara. The cut-off for this table is also 500 overs.

Now, we come to the bowlers who conceded boundaries most frequently. Josh Tongue leads the table, conceding a boundary every 11 balls - almost one every two overs. He is followed by Ebadot Hossain and Naseem Shah. They are followed by a string of pace bowlers. It can be seen that most of these bowlers are of recent vintage. Maybe this depicts the change in attitude of the current batters. These are all pace bowlers. Kuldeep Yadav leads the spinners conceding a boundary every 12.7 balls.

And finally, the players who were very good at not conceding frequent boundaries. Most of the featured bowlers are spinners. They are led by Kraigg Brathwaite, who conceded a boundary every 28 balls. He is nearly an over ahead of the next bowlers, Rahkeem Cornwall and de Silva. They are followed by a number of spinners who conceded a boundary every 20-plus balls.

Innings level
In the Delhi Test in 2015, South Africa scored 143 runs in 143.1 overs as Hashim Amla and AB de Villiers tried to save the match by simply occupying the crease. This bowling marathon included nearly 800 dot balls - over 92.5%. Meanwhile, in Mount Maunganui in 2025, New Zealand bowled 80.3 overs to West Indies, including 431 dot balls - just over 89%. In Adelaide in 2022, Australia bowled just over 40 overs against West Indies and had no fewer than 212 dot balls - 86.5%.

A peak into the past
In the history of Test cricket, there have been only two innings of length exceeding 50 overs that have ended with an runs-per-over value below 1.0 - the Delhi innings referred to above and the innings in which New Zealand scored 69 in 90 overs against Pakistan in Dacca in 1955. There was no play on the first three days. On the fourth day, New Zealand were bowled out for 70 (inside 40 overs) and Pakistan declared on the fifth day with a lead of 125 with 90 overs left in play. New Zealand then played the strangest innings in the history of the game. Their 90 overs contained 55 maidens (330 balls). That left New Zealand with 35 overs (210 balls) to score 69 runs. Let me assume an average rate of 1.75 runs per ball in the scoring deliveries. The balls required would be 39 (69/1.75). That means that the total dot balls would be 501. The dot-ball percentage is 92.8. This is still below the dot-ball percentage in the Delhi Test.

Now, we move on to the other end. In Brisbane in 2015, the New Zealand bowlers bowled 121 dots in 42 overs in Australia's second innings - only 48%. This is one of only two occasions, during the period covered, in which a set of bowlers did not even reach the 50% mark. Australia scored at nearly six runs per over. In Wellington in 2024, New Zealand bowled just over 82 overs and had a dot-ball percentage just below 50. England scored at over five runs per over. In the first Test in Multan the same year, England went on a rampage scoring at around 5.5, this time across 150 overs. There were only 459 dot balls in this innings - just over 51%.

Innings-level bowling
In that Delhi Test which recurs often in these narratives, the bowler who bowled the tightest spell was Umesh Yadav, whose 21 overs had a mere six scoring balls - a whopping dot-ball percentage of 95.2. The circumstances of the match explain such a spell. In the same match, Jadeja bowled 276 balls, out of which 262 were not scored off - leading to a dot-ball percentage of 94.9. Ishant Sharma and Ashwin, with their parsimonious spells, appear later in the table. In third place is New Zealand's Ajaz Patel, who had a dot-ball percentage of 93.2 in Mount Maunganui against West Indies in 2025.

A peak into the past
Let me now look at Nadkarni's famous spell of 32-27-5-0 against England in 1964 in Madras. There were only five scoring overs and five runs were scored in these. So, necessarily, these have to be five singles. This means that it is certain that Nadkarni bowled exactly 187 dot balls. A rare instance of an exact computation of the dot balls from an old analysis. That gives us a dot-ball percentage of 96.2, which is by far the best ever. I cannot think of any cricket spell more parsimonious than this. Maybe the only spell that stands comparison is Phil Simmons' 1992 ODI classic of 10.0-8-3-4 against Pakistan in Sydney.

This table looks at the number of fours conceded by a bowler in his innings spell. Yasir Shah conceded as many as 30 against England in 2016 at Old Trafford. He ran into Joe Root at his commanding best. Chris Morris conceded 27 fours in Cape Town, also against England in 2016, this time against the famous Stokes-Bairstow partnership. Zahid Mahmood and Tim Southee conceded 26 fours each in their respective innings.

Now, onwards to that relative rarity in Test cricket - sixes conceded by the bowler. Dane Piedt conceded no fewer than nine sixes in the second innings in the Visakhapatnam Test against India in 2019. Piedt bowled only 102 balls (and conceded 102 runs). Taijul Islam conceded nine sixes against South Africa in Chattogram in 2024, this time off 198 balls. Finally, Ben White of Ireland bowled 203 balls and conceded nine sixes

Finally, a look at the frequency with which a bowler conceded boundaries in an innings. The same Umesh Yadav who was so difficult to score off in Delhi, conceded a boundary every 1.8 balls against Australia in Sydney in 2015. In three overs of mayhem, Australia smashed ten fours, looking for a declaration. Hasan Mahmud of Bangladesh had a tough time against India in Kanpur in 2024, conceding a four every 3.3 balls. Two bowlers, Zimbabwe's Trevor Gwandu and Sam Curran of England, conceded a boundary every 3.6 balls.

Bazball-era analysis
This is a special coverage of the Bazball Tests - from Lord's in 2022 to Trent Bridge in 2026 - a sequence of 49 Tests. I will do two comparisons. the first between England - pre-Bazball and during Bazball. The other comparison: England during Bazball vs other Test teams during the Bazball period.

The first one is the straightforward dot-ball percentage. England's was 74.8 in the eight-year period between 2014 and 2022. That meant that they scored at 2.12 runs per ball off the scoring balls (49,806/23,525). Their overall scoring rate was 3.20.

During the Bazball era, England had a dot-ball percentage of 65.1. That meant that they scored at the same 2.12 runs per ball off the scoring balls (27,821/13,098). However, their overall scoring rate was 4.45. It is indeed amazing to note that there is a difference of 1.25 runs per over but the runs per scoring ball was exactly the same. England achieved the increase in scoring rate simply by cutting down on dot balls by nearly 10 percentage points.

Now, for the comparison during the Bazball era. The teams other than England have a dot-ball percent of 71.6. That meant that they scored at 2.03 runs per ball off the scoring balls (140,368/69,121). Their overall scoring rate was 3.46. This difference of nearly one run per over was caused by a slight increase in dot-ball percentage and a slight decrease in the scoring rate off the scoring balls.

England needed nearly 16 balls to score a boundary during the pre-Bazball era. In the Bazball era, they needed only 11 balls per boundary. That is indeed the difference between the two eras. Now, let me extend the additional analysis to the scoring deliveries. Pre-Bazball England scored a boundary off every 3.96 balls (23525/5928). Surprisingly, Bazball-era England needed almost the same number - 3.88 (13098/3371) to score a boundary. This is a validation of my earlier statement that the real difference was in reducing the number of dot balls.

During Bazball, the other teams are nearly as close in this measure to England's pre-Bazball numbers, and needed around 15 balls per boundary. And that is the difference during the past four years. And, expectedly, they scored boundaries less frequently than England - 4.25 balls per boundary (69,121/16,238).

I have added a third table, which is not connected to the data gathered during this exercise. This is on the percentage of maidens bowled in the three domains. One-fifth of the overs faced by England during the pre-Bazball era were maidens. During the Bazball era, only half of this number, one-tenth of the overs faced by England, were maidens. The bowlers were not allowed to settle down. The non-English teams were slightly better - around one sixth of their overs were maidens.

Potpourri
After a few months' gap, I am happy to resume the Potpourri section. This month again sees a table of contrasts: batter innings in a Test that were more than 250 runs apart.

Don Bradman has had two instances of pairs of innings with 250 runs of separation. Only three batters in Test history have played innings separated by 300 or more runs - Hanif Mohammad, Bob Simpson, and Azhar Ali.

Talking Cricket Group
Any reader who wishes to join my general-purpose cricket-ideas-exchange group of this name can email me a request for inclusion, providing their name, place of residence, and what they do.