Understanding the art of over construction in T20 cricket

On 30 April 2023, with Sikandar Raza needing three runs off one ball, Matheesha Pathirana bowled a slower ball instead of a yorker. Earlier in the over, a similar pace-off ball had gone for two runs through square leg. It went for three off the final delivery, with CSK captain MS Dhoni having brought his fielder inside the circle.

The incident was lost in Pathirana's overall death-bowling brilliance that brought his team back into the game, as well as the glory of the Super Kings' sensational run to their fifth IPL title. But for those who remember it, the event serves as an important reminder to ask one of the most important questions in T20 cricket.

Which ball do you bowl when?

Of course, this is not a question that exists in isolation from broader planning and strategy. Rather than simply deciding in the moment what delivery to bowl to a particular batter, the bowler must plan ahead and consider in what order his primary and secondary options' effectiveness can be maximised. It fits into the broader question of how you construct an over.

While different bowlers will naturally have different approaches, the data points to some trends - particularly among pace bowlers. And the case of Pathirana offers insights into what is currently proving effective among T20 bowlers, and why the Sri Lankan has lost some potency this IPL season.

Matheesha Pathirana - A case study of 2024 and 2025

The middle overs are arguably the best phase to analyse for fast bowlers from an over construction perspective. Among them, the 12-16 over phase holds special importance as it has a distinctly higher percentage of pacers in operation than the 7-11 over phase. It incorporates the high attack rates of the back half of innings without the high-variance, one-dimensional outcomes of the death overs, making it the best of both worlds in a sense.

Pathirana's economy rate in the middle overs in IPL 2024 was just above seven. This year, it has shot up to 10.32. His bowling average, meanwhile, has skyrocketed from 10.22 to 67.66.

There's no doubt that Pathirana's main weapons are the yorker, the bouncer and the slower ball. The Sri Lankan is bowling exactly as many bouncers as he was last year (29%), and his economy rate while taking pace off is still under five. His yorker accuracy, in fact, has increased - he's nailing his attempts 17% more often than he was in 2024.

The answer to what has changed lies in over construction. While Pathirana may have a set of weapons at his disposal, they shouldn't be confused for stock ball substitutes. In 2024, nearly 45% of his deliveries in the middle overs weren't attempted yorkers, slower balls or bouncers, a figure that has dropped to around 30%. His reliance on conventional fast bowling held him in good stead.

In the ongoing IPL, however, Pathirana has shifted from the conventional approach and is using his wicket-taking options more often and at different junctures. The probability of him starting an over with a length delivery has nearly halved, while the probability of him starting an over with a yorker has nearly tripled. To put it in simple words, the slinger is using his variations in the first half of each over, making himself extremely predictable over the course of six (and sometimes even 10) balls.

Pathirana's slower ball and bouncer usage has remained consistent, but there's been a drastic difference in the output of the latter delivery. Bouncers, which were going at an economy rate of 4.15 last year, are now going at 7.5. Again, the answer lies in over construction. Comparing the 22-year-old to some of the premier bowlers in the league is the natural exercise to undertake, with an emphasis on the 12-16 over phase for the same reasons mentioned earlier.

Unlike Arshdeep Singh, Jasprit Bumrah and Harshal Patel, the probability of Pathirana bowling a bouncer increases as the over progresses. He bowls them more than twice as often in the second half of the over, while the others use it as a mid-over weapon. A similar detrimental pattern can be observed with yorkers.

The key inference here is that the yorker is a delivery that closes out the over. The best bowlers of this ball in the league, including Bumrah, Arshdeep and Harshal, use it as such. The probability of them bowling a yorker is highest on the last delivery, and the trend holds true for most successful fast bowlers in the IPL.

Pathirana's economy rate is 8.3 in overs where he starts with a bouncer and 6.9 when he doesn't. It is 7.76 when he starts with a slower ball and 6.66 when he doesn't. It is 8.12 when he starts with a yorker and 6.73 when he doesn't. Most importantly, it is 6.69 when he starts with a length ball and 8.31 when he doesn't.

Despite acknowledging the possibility that Pathirana is likely to resort to his variations in overs when he's under pressure, overs which thereby go for more runs, there's a clear pattern that emerges.

The above image illustrates this clearly. Batters are taking advantage of the second scenario, where they are expecting Pathirana to bowl a yorker first up. A boundary off the delivery will immediately shift the onus onto the bowler, who must now go length or short. Either way, throughout the rest of the over, he will be predictable and under serious pressure.

A shift to the first scenario, where he can use the yorker and bouncer as surprise options whenever necessary, will go a long way in bumping his numbers up. Even if he goes for runs off his initial length ball, he would always have the option of taking pace off and/or bowling any of his three variations.

It is, in a way, unfair on Pathirana to club him with other fast bowlers. Everything about the Sri Lankan is unique - his slingy action and consequent low release points, the unparalleled amounts of swing he gets with the older ball, his express pace, and so on. But while he has more room for error and more flexibility, he still isn't accurate enough and experienced enough to be using his variations as stock balls.

As a general principle, it makes sense for fast bowlers to be using their stock balls at the start of the over, their primary attacking options midway through the over, and their primary defensive options to close out the over. Every bowler might not have the diverse and unique skill set of Pathirana, but they will certainly have a set of deliveries that they employ according to the situation and the conditions.

Variance is another key component of over construction. There is no guarantee that a bowler who bowls all the deliveries in his toolkit will be more successful, but there are a number of cases where it appears to be true. Since 2022, Prasidh Krishna's economy rate is 8.82 in overs where he attempts all three variations - yorker, bouncer, slower ball - and 11.12 when he doesn't. The trend of better economy rates in high-variance overs holds for Arshdeep Singh, Mohammed Siraj, Pat Cummins, Mohammad Shami, Josh Hazlewood, Khaleel Ahmed, Harshal Patel and many others as well.

In modern T20, the game is almost always in the hands of the batter, who must either take successful risks or perish. The bowler's job is only rarely to get the batter out - usually, it is to prevent a run-scoring opportunity and force a mistake. Constructing an over must be structured around this fundamental concept.

Away spin to win - and bounce to pounce?

Spinners might not have as many unique deliveries as pacers, but they still have a lot to work with. While they may not be able to experiment with length to the same degree as their pace counterparts, turn direction, flight, speed, direction of revolutions on the ball and release points are just some of the variables they can play around with. Turn direction is the most natural starting step for over construction, especially in the modern era.

Among the 26 right-handers who have faced more than 30 balls of away spin in IPL 2025, only six average more than 40 while striking at above 150. These are Suryakumar Yadav, Shreyas Iyer, Jos Buttler, Ayush Badoni, Aiden Markram and Shashank Singh. Among the 14 left-handers, only three do - Nicholas Pooran, Rinku Singh and Ryan Rickelton.

So for most spinners, the stock ball turning away from the batter is a huge advantage. This is reflected in their over construction methods.

Eleven spinners have turned the ball both ways against right-handers in IPL 2025 on a regular basis (at least 12 balls of each type). Of these, Sunil Narine (39.5%) and Noor Ahmad (41.7%) are the only ones who use lesser away-spin than in-spin. They can afford to do that - while Narine's economy rate in the season is a miserly 7.53, Noor has already picked up 20 wickets at a shade over eight.

Applying a similar filter of at least 12 balls each, 10 spinners have turned the ball both ways against left-handers in IPL 2025. Rashid Khan (28.7%) and Ravi Bishnoi (22.8%) are the only ones who use lesser away-spin than in-spin. Interestingly, Bishnoi's googlies to left-handers, which were one of his biggest strengths a couple of years ago, are being picked apart at an economy rate of 12.86.

It's clear from the above graphs that the economy rate of away-going balls is a decisive factor in not only spinners' mode of operation but also their degree of success. Interestingly, Chakaravarthy, Noor and Kuldeep, who have the three best away-spin economy rates, are also the three spinners with the highest total impact against left-handers in IPL 2025.

From an over construction perspective, a combination of four away-spinners and two in-spinners has worked well. Noor Ahmad vs left-handers and Suyash Sharma vs right-handers are among those who have achieved a high rate of success while using this approach.

What can finger-spinners do to adapt to this away-spin trend? In under 60 matches of IPL 2025, left-arm spinners have sent down 37 balls that turned away from left-handers. All previous completed IPL seasons since 2022 had less than 30 - 17 in 2022, 29 in 2023 and 24 in 2024. Sai Kishore, Kumar Kartikeya and Abhishek Sharma are among those who have bowled the away-going delivery to southpaws.

Others, like Krunal Pandya, have resorted to the bouncer. Krunal has bowled 11 deliveries shorter than 10m, including two wickets, and a whopping 25.7% of his balls shorter than 7m. For context, the average spinner has bowled only 13.7% shorter than 7m in IPL 2025.

But the bouncer isn't the key to Krunal's variance and over construction - speeds are. In fact, he has the highest standard deviation of his release speeds among all finger-spinners in the league. Others who have had productive campaigns, like Sai Kishore, Varun Chakaravarthy and Riyan Parag, have varied their pace well too.

The future of T20 spin bowling is likely in quick wrist-spin, but occasionally slowing it down has proven incredibly effective in IPL 2025. So far, spinners may not have had to think about over construction much, especially when compared to their quicker counterparts. But as they take on more death-bowling responsibility and come up against flatter tracks and more powerful hitters, they will need to place an emphasis on every single aspect in their control.

Sai Krishna is a freelance cricket data analyst and data scientist who is currently working with international cricketers and coaches