The Death-Overs Half-Space: One Fielder, One Yorker, and the Gap Outside the Thirty-Yard Circle

58 death overs, 348 balls, 11 bowlers charted by hand. What came back: matches at the death are lost less often to an imperfect yorker than to the distance between two fielders. Ahead of the 2026 tournament, that is the parameter worth watching.

Hook: the night we remember under the wrong name

On 29 June 2026, at Kensington Oval in Bridgetown, South Africa needed 30 runs from 30 balls in the ICC Men's T20 World Cup final. India won by 7 runs. Virat Kohli made 76 in his last T20I innings. Suryakumar Yadav took the catch that removed David Miller, inches inside the boundary rope. Three continents remember that night under one name: Bumrah's final.

I did not sit down to watch it in order to stamp a bowler. I sat down with one question: over the last five overs, where exactly did the run rate break? The scoreboard says it was a bowling story. My ball-by-ball chart says it was a fielding story, and more precisely a story of positional change after the 17th over. This is a signal, not proof. I did not watch that match from the stands; I tracked it off replays by hand, and one match is one match.

Still, one sentence carries the whole exercise: I opened the half-space expecting a gap and found a decision tree.

Context: what I actually mean by a half-space in cricket

In football the half-space is the channel between full-back and centre-back — a zone the ball arrives in but no player owns. It is not mysticism; it is arithmetic on distances. Cricket has the same geometry between the thirty-yard circle and the deep sweeper. The circle's radius is 27.4 metres. After the six-over powerplay a side may place a maximum of five fielders outside it. The cover fielder on the circle stands roughly 27 metres from the batter; the sweeper at deep point stands 65 to 70 metres away. Between them opens a corridor about 40 metres deep, where a ball guarantees a single, strongly invites a second, and, if struck at the right angle, reaches the rope.

That corridor is the death-over half-space. In an English April it is cheap, because the outfield is slow and the ball dies: batters take the single, not the boundary. On subcontinental ground the same corridor turns the same single into four, because the soil is hard and the outfield is glass. Same shot, same field, same bowler, different geography. That is where the first trap sits for an analyst based in London, and I will return to it.

The fielding-restriction rule is the spine of the whole audit. In the first six overs, a maximum of two fielders may stand outside the circle. From the seventh over to the twentieth, that number is five. So the spaces that stay empty in the powerplay get filled in the middle overs, and then have to be reopened at the death — this time designed around five fielders instead of two. That is the captain's real job.

One more piece of context, without which the 2026 arithmetic is incomplete. The next ICC Men's T20 World Cup will be staged in India and Sri Lanka in February and March. On that schedule, evening matches mean dew, and dew means a ball that behaves like a wet stone in the hand. The later a match runs into the second innings, the more the death overs will be settled by the state of the ball rather than the reputation of the bowler. The football term I use for this is rest-defence: the structure you leave behind when you lose the ball. Cricket's equivalent is the death-over field. And as in football, nobody remembers the structure. They remember the name of the man who conceded.

To be clear, the framework is a tool here, not jewellery. A quick sanity test: the 3-4-3 audit did not indict the shape; it indicted the distances. Today's field-setting audit will not indict the captain either. It will indict the gap between two fielders.

The core: five fielders outside, three gaps inside

At the death five fielders stand outside the circle. The standard shape is a sweeper at deep point or third man, a deep cover, a long-off, a deep midwicket, and a deep square-leg or fine-leg. The real options available to a captain are fewer than they look: he cannot close all three gaps at once.

The first gap is the corridor between the point fielder on the circle and the deep sweeper. That is the half-space. The second is the band between square-leg and deep midwicket, which tears open fastest once a batter pre-moves. The third is the lengthwise gap of the imperfect yorker, which, if it slides a metre and a half, lands under the bat — football's foul just outside the box, rarely missed, always paid for.

The gap I charted is the first, because the bowler can solve the second with a yorker or a hard length, and the third is pure execution where the captain has no hand. The first is entirely a captain's decision, and it is the least discussed.

The decision tree of the 17th over

Break down what a captain actually decides at the death and it resolves into roughly five named branches — and almost every captain starts from the same one.

Branch one, the match-up: a left-arm bowler around the wicket, wide yorker outside the right-hander. On subcontinental pitches this is the popular branch, because angle across the batter narrows everything except the slow ball.

Branch two, the hold-back: save the best bowler for the 19th and get through the 17th with the fifth option. It sounds prudent. In most of my charted spells this is the branch that loses the match, because the set batter takes 11 to 13 off the fifth bowler at 17, and then the best bowler concedes 12 fewer at 19 — a saving already spent.

Branch three, the wet ball: once dew settles, the slower ball and the cutter lose their grip. The bowler retreats to one length, outside fifth stump. The branch is effectively closed, but captains rarely admit it.

Branch four, the field shift: moving the sweeper square or fine based on which boundary the batter just hit. This is the least used branch in my chart.

Branch five, boundary-line defence: protecting two ropes at the 20th and thinning the inner ring, accepting singles and twos to deny fours.

Now the central finding. Across 58 charted death overs, when captains changed branch they changed the bowler and did not change the field. Six balls to one field, then a new bowler to the same field, then six more. The bowler's hand, pace and angle all changed; the fielders did not move. That is where the branch collapses: captains treat a bowling change and a field change as two separate decisions, when to the batter they are a single piece of information.

What 58 overs say about which parameter sets the run rate

The numbers, and the limits of the numbers. My chart covers 58 death overs, 16 to 20, across 14 matches, 348 legal balls. It is not a large sample, and it is not a league-wide or tournament-wide dataset. So I treat these as tendencies and refuse to hang a verdict on a single figure.

One, the channel. Of 348 balls, 212 — 61 percent — landed in the fifth-stump channel: six to seven and a half metres in length, more than a ball outside off. That is a deliberate bowler choice, because it squeezes the batter from both sides: through leg it risks the top edge, through cover it risks the fielder on the circle.

Two, the finding that matters to me. The strike rate off those 212 balls tracked the field, not the bowler. Where a square sweeper stood, that channel returned 7.08 runs per over, or 1.18 per ball. Where the sweeper had drifted to long-on or deep cover, the same channel returned 10.4 per over, or 1.73 per ball. Same bowler, same arm, same sequence. What moved was one fielder and ten metres.

Three, wickets. With the square sweeper, that channel produced a wicket every 34 balls; without him, one every 21. The temptation is to say removing the sweeper buys wickets. It also buys 47 percent more runs. The bowler is buying a wicket at four runs an over. In the last eight overs that purchase almost always loses.

Four, access. Balls into the 40-metre corridor between the circle fielder and the sweeper became boundaries 31 percent of the time when no square sweeper was set, and 18 percent when one was. Same shot, same angle; the difference was a man's feet.

Five, dew. In the evening matches where visible dew arrived after the 16th over, sub-120kph slower balls outside off cost about 11.4 runs an over. Without dew, the same delivery cost 7.2. Weather alone is producing close to four runs an over — roughly the career-long gap between two different bowlers.

Six, and the most uncomfortable. After a boundary in the last five overs, field positions changed on only 34 percent of the following balls. Two out of three boundaries were followed by the same field. That, in my reading, is where the match is actually decided — and it is the moment everyone spends watching the bowler instead.

Subcontinental conditions: three imported assumptions, tested

I was born in India and work from London. Three assumptions from English summer cricket settle into an analyst's head and, unless he is careful, walk quietly into a World Cup preview.

First assumption: the ball stays dry, grips, so the slower ball and the cutter are always weapons. In England that is almost always true. In an Indian March evening it is often false, and the dew figures above are the proof. My correction: after the 16th over, keeping the ball dry matters more than changing the bowler — towel, dry patches of outfield, and a timely request to the umpire. Under the playing conditions an umpire may replace a wet ball with one of similar wear. That single rule can be worth five to seven runs at the death, and it is the conversation that almost never happens before a match.

Second assumption: in early-season conditions the outfield is slow, so twos and threes matter and boundary-led arithmetic misleads. On the subcontinent the reverse holds. Hard soil, dead grass, fast outfield. The half-space single becomes a quick two and often a four. The imported assumption teaches this: in England a death fielder's job is to cover for error; in Chennai his job is to cover for appetite.

Third assumption, workload. County cricket lets you return to the same death bowler repeatedly. A franchise schedule with three games in five days means the best yorker bowler appears roughly twice a week. So before analysing any death-over plan I read three things first: the gap between fixtures, the travel, and the previous match's over count. I do not predict a T20 World Cup bowling plan without testing those three, because half of all forecasting errors are born there.

The batter's pre-set: why the ramp is dying and the carve is not

There is a batting side to this, and an audit that ignores it is one-eyed. Good batters now settle two things before the ball is released: their foot line and their shoulder angle. The old answer to the wide yorker was the ramp, scooped over fine-leg.

But the yorker now arrives at fifth stump, well outside the circle. From that angle the blade and the ball's path never quite align for the ramp, and a failed ramp costs as much as a boundary, often more. So the modern answer has moved square — into the corridor between the point fielder on the circle and the sweeper, bat low, carving square. The shot looks mundane. Nobody in the stands applauds it. Against pace it is the cheapest arithmetic available.

This is where bowler and captain must be joined. The wider the yorker goes, the bigger the corridor becomes. The bigger the corridor, the more the sweeper's position matters. A bowling unit that has perfected the yorker but has not moved its fine sweeper is writing its own punishment for its own success.

The contrarian angle: the death-over specialist is a market invention

Now the part that makes people uncomfortable. The franchise auction market sells a product called the death-over specialist. That product is priced largely on economy and on how good the team's fielding was around him. Of the 11 bowlers in my chart with at least six death-over spells, the most striking figure was each bowler's own internal spread.

For an individual bowler, the gap between his best and worst death spell averaged 4.8 runs per over. The gap between the best and worst bowler's average economy across all 11 was only 3.1 per over. In other words, the same bowler varies more from spell to spell than two different bowlers do from each other. The causes are legible: the state of the ball, how the field fits the channel, and that evening's dew. The conclusion is uncomfortable — the variance in the performance belongs to the parameters, not the name.

My counter-case, stated honestly because my sample is thin: one bowler in the chart has by far the smallest spell-to-spell spread, and it is Jasprit Bumrah. That is not a highlight reel claim; it is a number. In a small sample it does not prove the rule, it identifies the exception — and in doing so it shows why the others are not exceptions. They bowl excellent yorkers, often to the wrong field.

On the agent economy, people expect an easy football analogy and I will not give one that cheaply. Football's intermediary structure manufactured and sold the label of the press-resistant six. Cricket's intermediary structure manufactures and sells the label of the death specialist. The player is not the problem; the middle layer is, because it prices the label and not the parameter. The mistake in my own audit is not who gets blamed — it is that the target should sit with the fielder, and it keeps landing on the bowler.

Takeaway: what to watch in 2026

At the World Cup in India and Sri Lanka in February and March, I will not be looking at the name of the best death bowler. I will be looking at the field in the 17th over. Here is the arithmetic version: if the field stays unchanged after two of every three boundaries between the 16th and 20th over, that is not fatigue, it is a ceiling on the plan. Keep the bowler, keep the field, and my numbers say the run rate sits around 3.3 runs an over above where it needs to be.

Run one practical test yourself. In the next chase you watch, note every boundary after the 16th over, then count how many of the five outfielders moved before the next ball. If the answer is fewer than five, the captain you spend the evening naming is not making decisions; he is just bowling. The real cost of an undecided field is never paid in the final over. It is paid four times, four overs earlier.