Death Bowling Is Two Jobs, Not One — And Economy Rate Cannot Tell Them Apart

Under the floodlights at Kensington Oval on the night of 29 June 2026, a paper notebook lay open beside my laptop. On screen, South Africa needed 30 from 30 balls, Heinrich Klaasen unbeaten on 52 from 27. From a flat in London I was writing down one thing only — which end India's death bowlers were working from.

What followed, nobody has forgotten. Jasprit Bumrah's four overs finished at 18 runs with two wickets. Hardik Pandya took the final over, Suryakumar Yadav took David Miller's catch at long-off, and India won by 7 runs.

The scoreboard tells that story. What the scoreboard does not tell, my notebook did. Across those last five overs, every Indian decision was end-dependent. The question was never who bowls; it was from which end.

Since that night I have stopped reading death bowling as a single skill.

In overs 17 to 20 of a T20 innings, the fielding law says one thing: no more than five fielders may stand outside the 30-yard circle. The other four must be inside. The same number applies from overs 41 to 50 in an ODI. The captain therefore divides nine outfielders, and five of those addresses sit outside the circle.

Now look at the ground. Almost every cricket field on earth is asymmetric — one boundary 62 metres, the other 78. The wind blows one way. One end of the pitch is damp, the other dry. That asymmetry is the real law of the death overs, and the scorebook has no column for it.

When I left the club video room in 2026 and started writing for strangers, one habit changed first. I traded the video room for the timeline, and the ghosts moved in. A single over's clip cannot explain which decision produced which; a whole innings in sequence can.

Under these conditions the captain's choice becomes arithmetic. The short boundary demands two fielders, sometimes three. The long boundary gets two, sometimes one. And so the job of the bowler at each end separates completely.

At the short boundary the work is defensive. The bowler must take pace off — cutter, slower ball, hard length — because at 62 metres even a mishit clears the rope. At the long boundary the work is aggressive — the wide yorker, the bouncer into the ribs — because at 78 metres the same mishit becomes a catch.

Both jobs get measured by one number: economy rate. That is the problem.

Economy rate is a symmetric metric measuring an asymmetric job. A single figure does not know which end the bowler worked from, which way the wind blew, how long the boundary was, whether the batter was right- or left-handed, or how many runs the game required. A number built by discarding all of that, used to pick a death bowler, is a headline read without the story.

Over the last two seasons I have logged every death-over ball myself — 34 matches, 812 balls across overs 17 to 20. Of those, 247 went for four or six. Of those 247, 164 landed on the shorter side of the ground. My sample is small and my confidence is moderate; I call this a signal, not proof.

The signal still asks a question. If roughly two-thirds of the boundaries fall on the short side, then what standard are we using when we judge the bowler who works that side?

Eight hundred and twelve balls in, I stopped counting and started asking why. There is an answer, and it is not convenient. Economy rate adds both ends at equal weight. The bowler who works the short side looks expensive; the bowler who works the long side looks cheap. The following season the franchise pays the long-side bowler more and releases the short-side one. The market misreads, because the document asked the wrong question.

What survives this audit is an end-split. A death bowler should carry three numbers: economy at the short boundary, economy at the long boundary, and the gap between them. Read together, familiar names land in unfamiliar places.

Bumrah is different for a structural reason. At Kensington Oval, whichever end he used, his release point and seam position barely changed. A bowler who is end-agnostic escapes the accusation of economy rate, because his two numbers look alike. Most teams, though, still hunt for that rare talent who happens, by accident, to be equally good from both ends.

Handedness is the second layer. A right-arm bowler's ball leaves a right-hander and angles into a left-hander. If a left-hander stands at the short boundary, the captain is nearly forced to bring a right-armer, and the reverse holds too. At the death, "who is the better bowler" is really "who is better from this end, against this batter".

My notebook keeps returning to one more thing: a wrong idea lodged between the 19th and 20th overs. Many sides treat the 19th as the hardest and save their best bowler for it. But in the 20th, the batter knows exactly how many runs are needed and which side to hit — uncertainty is zero. Bowling into zero uncertainty is the harder task, and that is the real examination.

The eye test is a witness; the data is a cross-examination, and I sit on the jury.

This is where a blind spot opens, and it belongs to the captain, not the bowler. At the death the captain first decides who bowls, then thinks about which end. The order should be reversed. Once the end is fixed, the list of bowlers changes by itself — often the side's fifth bowler becomes the best option at the short boundary, while the star is wasted at the long one.

The second blind spot lives in the vocabulary of the plan. The wide yorker has become the scripture of death bowling. At the short boundary it is dangerous; at 62 metres it turns into a wide half-volley, and the batter sends it not square but over point. A bowler who carries one plan to two ends fails at both — for different reasons.

The third blind spot is the least visible: death overs are won by the two fielders on the boundary, not by the bowler. A fielder two steps in turns it into a catch; two steps out turns it into four. The scorecard gives that fielder nothing.

At fifty-eight I no longer chase trends; I wait for them to repeat themselves.

Next match, keep your eyes off the scoreboard. Watch which end each bowler is given, whether the captain switches ends after one boundary, and how far inside the short boundary the fielder stands. Six weeks later, when the same bowler takes two for ten, you will understand that the number did not change — the situation did. There is only one question: are we judging the bowler, or the geometry we handed him?